Poetic Injustice

Abstract

When J.R. Cash (Johnny Cash) sings that he shot a man in Reno just to watch him die, audiences impressed by the singer's skillful creation and depiction of a nihilistic lyrical subject clap and cheer. When Terrell Doyley (Skengdo) and Joshua Malinga (A.M.) sang broadly similar lyrics at a concert in 2018, London's Metropolitan Police and the Crown Prosecution Service took them to be describing violent acts they had participated in and violent intentions they harbored, and the lyrics were used as the basis for legal proceedings against the singers that resulted in convictions. In this paper, I will argue that Doyley and Malinga's case illustrates a distinctive and important form that epistemic injustice can take. By failing to see their lyrics as speech that involves the exercise of their capacity for imagination, the police and prosecutors treat them as an impoverished sort of epistemic agent. I will call the wrong involved in cases like this one poetic injustice.

1. Introduction

In January 2019, a UK court sentenced London musicians Terrell Doyley (Skengdo) and Joshua Malinga (A.M.) to a suspended nine-month prison term for performing their song ‘Attempted 1.0’ in concert in violation of an August 2018 injunction against producing music videos, uploading music to the internet, or participating in live performances.¹

While Doyley and Malinga were the first to receive a custodial sentence, judicial action against British musicians working in the same genre, called ‘drill’, has taken a variety of other forms. In another case that drew significant media attention, Micah Bedeau, Yonas Girma, and Isaac Marshall, together with two unnamed minors were “given three-year criminal behaviour orders … banning them from mentioning death, injury, or rival postcodes in their songs and requiring them to inform police of any new music videos within 24 hours.”² Under powers granted to UK police in connection with anti-terror legislation, “anyone identified in [drill] videos can be targeted with action including criminal behaviour orders that can prevent them from associating with certain people, entering designated areas, wearing hoods, or using social media.”³

Metropolitan Police Detective Superintendent Mike West provided the following justification for legal action targeting drill artists:⁴

What looks like a music video can actually contain explicit language with gangs threatening each other. There are gestures of violence, with hand signals suggesting they are firing weapons and graphic descriptions of what they would do to each other.

Cressida Dick, who at the time of Doyley and Malinga's case was Commissioner for the Metropolitan Police, claimed that drill rappers “describe stabbings in great detail, joy and excitement,” and provided the following justification of the practice of using lyrics against singers in court during cases involving violent crime: “They say what they're going to do to each other and specifically what they are going to do to whom.”

The song that landed Doyley and Malinga in trouble does contain lyrics that could in principle be taken to describe a stabbing. For example, they sing:⁵

Bellz got chinged, he was tryna go caf’.

And his young G, got cheffed times three.

He was on the lack, tryna get some wings.

Don't nobody know about Uber Eats?

The same song contains lyrics that could be treated as an assertion to the effect that the singer carries a pistol and does not hesitate to use it:

My big .45, it came with fries

I keep her close, my valentine

Her range ain't shit, it's kinda wide

So if you lack, you're gonna die.

I take the reactions quoted above to suggest that what the police and the Crown Prosecution Service hear in lyrics like these are descriptions of actual violent crimes that the artists were involved in, or crimes that they intend or aspire to be involved in. I assume that most philosophers will agree, however, that this is not the interpretive stance we typically take with regard to song lyrics. To have a concrete example to hand, contrast the following text from the 1955 J.R. Cash (Johnny Cash) song ‘Folsom Prison Blues’:

When I was just a baby,

My mother told me ‘Son,

Always be a good boy,

Don't ever play with guns.’

But I shot a man in Reno,

Just to watch him die.

A live 1968 performance of this song inside the walls of California's Folsom State Prison helped cement Cash's standing as one of the most important recording artists in American history and to sell more than 3 million copies of the record on which it was featured. While the live audience can be clearly heard on the record to clap and cheer at the words ‘I shot a man in Reno just to watch him die,’ it is very difficult to believe that many listeners, then or now, would have taken Cash to have been asserting that he shot a man in Reno, from idle curiosity or any other reason.

I will assume that the reason the typical listener does not take Cash's song to implicate him in a crime is that the most natural position to adopt with regard to the speech produced during his performances involves seeing it as speech produced by an agent whose aims and interests are primarily creative or aesthetic, as opposed to testimonial. I assume, furthermore, that many listeners will have no difficulty seeing the speech produced by drill rappers in the same light; presumably part of the reason their cases have received so much attention in the media is that the police reaction seems so shockingly at odds with common sense. That said, it seems clear that a significant number of people, and in particular, people in positions of power, have taken importantly different interpretive perspectives with regard to the work produced by artists like Cash, on the one hand, and Doyley and Malinga, on the other.

Giving a complete explanation of this difference would presumably involve disentangling the complicated intersecting influences of a variety of different forms of social identity prejudice, and indeed, the roles played by the different conventions associated with the different musical genres the artists represent.⁶ I will assume without further discussion, however, that the most significant factor in any plausible explanation is going to be anti-Black racism.⁷

For the rest of the paper, I will focus on the question of how to understand the intuitive wrong that is manifest in the difference in audience reactions to Cash, on the one hand, and Skengdo and A.M., on the other. I will begin, in the next section, by looking at the case in speech act theoretic terms. While I think singers like the ones I discuss are the victims of a problematic form of illocutionary distortion, in sections 3 and 4 I will argue that there is more to the story than this. If we miss the sense in which failing to recognize someone's imaginative activity for what it is is an “undermining, insulting, or otherwise withholding of proper respect for [them] qua subject of knowledge” (Fricker 2007: 20), we miss something important, indeed. If social identity prejudice results in your seeing only testimony in someone's creative output, you do them a form of epistemic injustice I will call poetic injustice.

2. A Speech Act Theoretic Explanation

Langton (1993) and Hornsby (1995) laid the foundation for a large literature on questions involving what we might think of as a form of linguistic justice by pointing out that the illocutionary possibilities that are available to speakers are not evenly distributed. Of course, to some extent, an uneven distribution was a basic feature of the cases Austin (1962) used to introduce speech acts into the philosophical lexicon. But unlike cases in which, say, a police officer ends up with illocutionary possibilities a civilian lacks, the cases Langton and Hornsby focused on were cases in which membership (or assumed membership) in certain social groups that we would not think ought to be differently empowered in fact turn out to be so. So, if men can refuse sex by saying ‘no’, but women cannot, a distinctive moral problem arises.

On the face of things, this sounds like a straightforward description of the case of drill. Comparable locutions produced by Doyley and Malinga on the one hand, and Cash on the other, appear to result in different speech acts – the former's lyrical production is taken to amount to testimony, threat-making, and so on, while the latter's is not. The most obvious explanation of this illocutionary difference involves the difference in the way the singers are racialized.

Because I expect this style of explanation to be so salient for so many readers, and because I think it does in fact go some way towards explaining our intuitions about Doyley and Malinga's case, it will be worth taking some time to look at what can be done with the resources of what is sometimes called applied speech act theory. With the outlines of a speech act theoretic explanation of the case on the table, it will be easier to emphasize the distinctively epistemic facets of the wrong that is involved.

Langton and Hornsby's early contributions developed and explored the phenomenon of ‘illocutionary silencing,’ which occurs when the words a person utters fail to count as the realization of a speech act, in virtue of their failing to meet the ‘uptake condition’ proposed by Austin (1962). While there is considerable disagreement about which acts exactly the condition ought to apply to, if it ought to apply to any at all, and how best to formulate it, on a simple framing that is good enough for present purposes, the idea is that someone's uttering a certain set of words only counts as the realization of a speech act if the audience recognizes the words as so counting. For example, if an actor onstage shouts ‘Fire! Fire!’ in response to a real fire just at the point in a performance at which a fire was meant to have been depicted, the words fail to amount to a warning if the audience thinks they are merely a part of the play.⁸

Something broadly like illocutionary silencing appears to be a feature of Doyley and Malinga's case. Like the actor onstage during a fire, their words, in virtue of the way they are taken up, have a different illocutionary force from the force that would result if those words were uttered by someone else (a firefighter entering the theater from the street, for example). At the same time, however, to say that Doyley and Malinga are silenced seems to put things the wrong way around. Intuitively, the injustice they suffer is due not to their failing to realize an action others would have been able to realize, but to their counting as having realized an action where others would not have.

Langton and Hornsby's early work on silencing has since been developed and refined in ways that can help us capture this intuition. Consider, for example, the notion of ‘discursive injustice’ described by Kukla (2014):

When members of any disadvantaged group face a systematic inability to produce certain kinds of speech acts that they ought, but for their social identity, to be able to produce – and in particular when their attempts result in their actually producing a different kind of speech act that further weakens or problematizes their social position—then we can say they suffer a discursive injustice. (Kukla 2014: 441)

I take Doyley and Malinga's case to involve a problematic form of illocutionary distortion very much like the one at the heart of Kukla's discursive injustice. The duo utter the words “I keep [my .45] close … so if you lack you're gonna die” in a context in which other speakers who uttered the same words would not be taken to be making an assertion (a threat, etc.). They nevertheless count – or at least, are counted by some – as having made an assertion (threat, etc.), and the difference is due to the way in which they are racialized. This ‘queering’ clearly “further weakens or problematizes the speakers’ social positions” (Kukla 2014: 441).

But is there a “certain kind of speech act that” Doyley and Malinga “ought, but for their social identity, to be able to produce” (Kukla 2014: 441)? The answer here is less readily apparent than it would be in the kinds of cases usually discussed in the literature on silencing and illocutionary distortion, but I can imagine two plausible ways in which we might defend an affirmative answer.

The first is to make room in our ontology of speech acts for what we might think of as a kind of act of omission, or as an illocutionary non-act. This is not as strange as it might sound. We often utter words without intending the utterances to count as illocutionary actions. In many cases in which we recognize this kind of activity in others, it seems right to say that the utterances in question do not in fact amount to speech acts. For example, when someone is learning a language, they might be required to pronounce some sentence in order to practice an intonation contour. As a language teacher, you would clearly be wrong to understand such attempts as assertions of whatever content would be conventionally associated with them (or to take words produced in this way to amount to promises, threats, and so on). One way to bring the Doyley and Malinga case under the scope of a Kukla-style treatment, then, might be to say that our ordinary practice when dealing with singers is to treat them as though they realize locutionary, but not illocutionary acts. By taking what in someone else's mouth would have been a non-act to be an assertion (threat, etc.), then, the police and prosecutors do them a discursive injustice.

Alternatively, we might distinguish between two types of apparently assertoric speech acts: ordinary assertions, on the one hand, and ersatz assertions of the sort that in the philosophy of literature are often called ‘fictive utterances,’ on the other. While there is debate about whether the notion of a fictive utterance is really suited to bringing out what is distinctive about fiction, as well as disagreements about how best to formulate the notion itself, philosophers widely agree that declarative sentences that occur in works of fiction do not involve an author's taking on the same kinds of commitments as an assertion would. On one popular treatment, declarative sentences to the effect that ϕ, when they occur in a work of fiction, should be understood not as prescriptions to believe that ϕ, as is often taken to be the case with an assertion, but rather to imagine that ϕ.⁹

If we take fictive utterances to involve a distinctive form of illocutionary action, we can apply Kukla's notion of discursive injustice directly to our case. When Doyley and Malinga sing the words “attempted, [but] should have been a murder” then on this view we would say that their aim is to invite imagination. If the background social conditions are such that poor, young, relatively less famous Black men are not entitled to invite imagination (or at least, not of the sort that might be at stake in a drill song), then on Kukla's picture, their attempts to do so will end up just counting as assertions (threats, etc.). As previously indicated, I will assume that this perversion of the drill artists’ illocutionary aims would best be explained with regard to the way they are racialized, and I take it to be evident that it further problematizes their social position. In addition to the injury of the legal sanctions, the loss of income, and so on, being pushed out of the role of story-teller and into the role of a flat-footed reporter of facts intuitively amounts to a substantial insult.

If discursive injustice were all there were to Doyley and Malinga's case, it would still deserve our attention. The real and hypothetical protagonists that have been discussed so far in the literature on the topic are nearly all speakers who face discursive injustice from ‘the other direction’, as it were. When a woman's attempt to tell a harrowing personal story of sexual violence is treated as erotica (as in Langton 1993), or a minority student's devastating objection to a philosophical proposal is treated as verbal floundering in a seminar (as in Maitra 2009), intuitively, speakers who ought to be taken seriously are not. To our existing repertoire of cases with this broad structure, it is important to add cases that reveal the distinctive sorts of problems that can occur when someone who ought not to be taken seriously – at least at the level of description that is relevant here – nevertheless is.¹⁰

All this said, in my view a speech act theoretic treatment does not nearly exhaust what is philosophically interesting about the police response to Doyley and Malinga. As philosophers working on the phenomenon of epistemic injustice have sometimes emphasized, discursive and epistemic injustices tend to go hand-in-hand, often reinforcing and helping to explain one another.¹¹ For the rest of the paper, my aim will be to show that such is the case where Doyley and Malinga are concerned; I will argue that there is substantial philosophical insight to be had by seeing what happens to them through the lens of epistemic injustice.

3. Epistemic Roles

In her foundational contribution to what has since become a very large and diverse literature, Fricker (2007: 20) defined epistemic injustice as a harm that involves the unwarranted “undermining, insulting, or otherwise withholding of proper respect for the speaker qua subject of knowledge”. Fricker's own presentation of the concept centers on two forms it can take. The first of those – which she calls ‘testimonial injustice’ – is the harm that occurs when the degree of credibility a speaker is afforded is negatively affected by systematic prejudice that targets the speaker's social identity. The other involves what she calls ‘hermeneutical injustice,’ a harm that occurs when people with certain social identities are less able than others to understand or communicate about important aspects of their social experiences in virtue of the uneven way conceptual lacunae are distributed.

While both testimonial and hermeneutical injustice have received significant attention from philosophers working in a broad range of philosophical subfields, important work has also been done on intuitively morally significant facets of our assessments and interactions with one another as epistemic agents that do not fall neatly under either of Fricker's two headings.

Developing a line of thought that I believe can help us understand what goes wrong when the police and prosecutors assess Doyley and Medina's lyrical production, Hookway (2010) and Medina (2011), for example, emphasize that a person's status as a candidate knower or transmitter of knowledge is not the only epistemic status that is important where their identity as an epistemic agent is concerned. Medina points out that we can and indeed often do sensibly ask whether some agent is properly accorded the status of being a producer of knowledge or is seen as someone who is actively engaged in a process aimed at coming to know something. Along related but also importantly different lines, Hookway observes that we can distinguish between cases in which we treat someone as a reliable source of testimony and cases in which we treat someone as a partner in a process of joint inquiry. Although these statuses and the status Fricker discusses are typically co-instantiated, we can imagine circumstances in which they come apart, and in which concomitant questions about epistemic injustice arise.

For example, consider several different ways in which I might interpret a potentially epistemically significant activity that you are engaged in. Suppose I come into the department common room one morning and see you furiously scribbling away at the chalkboard, producing line after line of formal notation. If I take you to be working on memorizing a proof, when you are in fact trying to solve a problem, there is an intuitive sense in which I withhold proper respect for you qua subject of knowledge, even though I do not deny you the status of someone who can know facts about mathematics. I treat you as a candidate knower of mathematical facts, as it were, but not as a candidate discoverer thereof.

At its heart, the failure involved here is a failure to recognize that you are engaged in a certain kind of epistemic activity, a kind of activity that is characteristic of fully-fledged epistemic agents. Without wanting to place undue emphasis on the significance of any particular way of taxonomizing, I think it can be useful to view Hookway's example involving a failure to treat someone as a participant in a joint project in terms that are a bit different from these. While such a failure also involves a misperception of someone's intentions and a misestimation of their sophistication as an epistemic agent, there is a sense in which the Hookway-style failure involves a distinctively relational element.

Suppose, to present our example of your work at the blackboard in a different light, that you are a PhD student in a department where I am a faculty member. Suppose, furthermore, that the particular proof you are working on is addressed at filling a gap in a talk I have recently given. Even if I recognize you as someone who is not merely a candidate knower of facts, but as someone who is engaged in a process of enquiry, it seems like there are at least two importantly different ways in which I might react to seeing the contents you produce. On the one hand, I might see you as a potential collaborator, as the kind of person with whom two-way exchange, whether on equal terms or not, will be potentially fruitful. On the other hand, I might overlook you, as it were, and treat the output you produce as though it were merely more raw material for my own individual process of reasoning.

Intuitively, it seems that overlooking you in this way amounts to undervaluing you, epistemically speaking. As in the first presentation of the example, there is a sense in which it might be right to say that I have failed to see you as someone engaged in or capable of engaging in a certain kind of activity, the activity of joint inquiry. But there is something distinctive to be had by thinking of this failure as a failure to allow you to occupy a certain kind of relational position vis-à-vis myself as an epistemic actor. It seems important to acknowledge, that is, that I have not just missed the fact that you are engaged in a certain kind of activity, but that I have in a sense prevented you from engaging in that activity in the first place, by failing to do my part in meeting you. In addition to failing to see your outstretched hand for what it is, I fail you by not reaching out myself to take it.¹²

In the next section, I will argue that Doyley and Malinga are subject to underminings of each of these broad shapes, and that since the underminings in question are the result of social identity prejudice, they amount to epistemic injustices. Before beginning that argument, however, I want to call attention to a third way in which things can go wrong when we evaluate or engage with another agent epistemically. In addition to failing to recognize your participation in a certain kind of activity and failing to participate in an activity with you that I would have with someone else, I can wrong you by failing to recognize the operation in you of a certain kind of epistemic capacity.

In principle, this kind of failure might happen with regard to a wide range of capacities of the sort that we recognize in ourselves and in those we take to be our epistemic peers. So, for example, to build further on Medina's point, if I fail to see you as the kind of person who could solve a problem, and not merely report a solution, then I fail you even more egregiously than I do if I miss your role as a producer of knowledge on a particular occasion. Given my aim for this paper, I will focus my discussion on one capacity in particular, which I think plays a fundamental role in explaining what goes wrong with the police and prosecutors’ description of Doyley and Malinga's music. The capacity I have in mind is the capacity to imagine.

There is no space here to survey the range of ways in which philosophers have characterized imagination, let alone to evaluate arguments that have been offered for the claim that imagination has a role to play in explaining how we come to have knowledge in one or another specific area. Many philosophers, however, accept that some form of imaginative activity should be counted as a basic component of our epistemic toolkit.¹³ For the purposes of my argument, then, I will follow Shiffrin (2014) in assuming that:

[E]very individual, rational, human agent qua thinker in community with other thinkers has specific interests in … exercising the imagination. [Each] thinker has a fundamental interest in understanding and intellectually exploring non-existent possible and impossible environments. Such mental activities allow agents the ability to conceive of the future and what could be as well as what could have been. Further, the ability to explore the nonexistent and impossible provides an opportunity for the exercise of the philosophical capacities and the other parts of the imagination. (Shiffrin 2014: 86–7)

Failing to see someone as as an imaginer, I submit, is to fail to see them as a fully-fledged epistemic agent. If you are describing a scenario you imagined and I take you to be reporting about facts on the ground, I misrepresent your epistemic activity in a problematically reductive way. Of course, this misrepresentation can lead to bad first order consequences for you, if I take you to be confessing to crimes, for example, or committing yourself to determinate propositions about this or that matter of fact when you are really just exercising your creative spirit. But as in the cases Fricker originally called attention to, in addition to consequences of this sort, I take the fundamental problem here to be a distinctively epistemic one. By missing the fact that you are engaged in an imaginative exploration, I treat you as a less sophisticated sort of agent than you really are. If I fail to recognize you as the kind of epistemic agent that is capable of imagining in the first place, I fail you even more egregiously in this regard.

4. Poetic Injustice and the Case of Drill

On my way of understanding the case, the fundamental wrong the police and prosecutors do Doyley and Malinga is to fail to see them as imaginers. To see how this plays out, consider Cash's ‘Folsom Prison Blues’ again:

When I was just a baby,

My mother told me ‘Son,

Always be a good boy,

Don't ever play with guns.’

But I shot a man in Reno,

Just to watch him die.

To be clear, it is perfectly possible for an interpreter to treat these words as the veridical record of some past sequence of concrete events and to see them at the same time as the reflection of the songwriter's creativity or engagement in an aesthetic project. Given a certain body of facts, after all, there are many ways an agent might choose to report them, and there is room for us to recognize a range of different talents in the delivery. Treating the song as testimony, however, is incompatible with treating the world described as one the singer imagined.

In what follows, I will argue that failing to see someone as an imaginer leads to compounding epistemic wrongs by engendering failures of the sort Medina and Hookway describe, that is, failures to see the person as a producer of knowledge or to allow them to occupy a certain position in an epistemic relationship. Even setting these downstream effects aside, however, there is a straightforward intuitive sense in which we undervalue a songwriter as an epistemic agent simply by failing to recognize the role of imagination in their work in the first place. Even if ‘Folsom Prison Blues’ amounts to nothing more than an idle description of a way a world could be, a vignette held up for our – or the author's own – contemplation without an editorial stance, seeing it in that light means seeing the author exercising what I recently quoted Shiffrin (2014: 86) calling a “fundamental interest” held by “every individual, rational, human agent qua thinker in community with other thinkers.” To treat the song as credible testimony is to cut the singer out of this community and to see them as less than a fully empowered or fully developed epistemic agent.¹⁴

As I indicated previously, it stretches credibility to think that anyone might seriously take Cash's song to involve testimony – normal listeners will immediately recognize that it is a work of fiction, something the singer made up. The contrast with the reaction to Doyley and Malinga's ‘Attempted 1.0’ is stark. When the police and prosecutors hear that song, they hear testimony about crimes the singers were involved in and about what they “[are] going to do to each other and specifically what they are going to do to whom.” I take the difference in these reactions to be philosophically significant because it reveals how an audience's willingness – or ability – to see someone as an imaginer can be affected by social identity prejudice. While it would be hasty to discount the role that prejudices targeting class or assumptions about the singers’ educational background might have to play here, the most obvious explanation of the difference is one based on the way they are racialized. Regardless of exactly how the details shake out, then, I take it that Doyley and Malinga are subjected by the police and prosecutors to a kind of epistemic injustice that I think deserves a name of its own:

Poetic injustice

Poetic injustice is the distinctive subspecies of epistemic injustice that occurs when an agent is undermined, insulted, or otherwise underestimated as an imaginer in virtue of a social identity prejudice.

4.1. Failure to see the imaginer as a producer of knowledge

When we see someone as an imaginer, that perspective tends to entail seeing them as a producer of knowledge as well. By the same token, failing to see someone as an imaginer forecloses on a range of perspectives from which agents typically present to us as engaged in epistemically productive exploration. Poetic injustice, then, will tend to bring injustice of the sort Medina (2011) described in its train.

Recent philosophical work on free speech can be used to flesh out some of the connections between the imagination and the production of knowledge.¹⁵ Philosophers such as Macklem (2007), Gilmore (2011), Shiffrin (2014), and Kendrick (2018) have argued that speech, broadly construed to cover a range of expressive modalities in art, music, and so on, deserves special protection because of the role it plays as a form of external cognition.

According to the formulation developed in Shiffrin (2014), for example, open and reliable channels of communication are important not only so that we encounter truths when proffered by others, but for us to develop our own beliefs and indeed the capacities we use in belief formation and evaluation in the first place. In the same way a painter might realize what their vision for a certain painting was only after experimentally setting brush to canvas, Shiffrin claims that our moral and epistemic development depends on our being free to render our own thoughts public; sometimes it is only by searching for the best way to vocalize a thought, or by setting it down on paper that we are able to take a critical perspective with regard to it, arrive at some consequence it entails, or appreciate how it fits together with some other commitment of ours.

On the way I understand this thread to have been developed in the free speech literature, more focus has been placed on the epistemic significance of externalization itself than on the particular nature of the process by which the externalizations in question are generated. For example, it seems plausible to allow that someone might progress in their moral thinking by building candidate theories on the Mad Libs model, completing sentences like ‘An action is good if …’ by picking words at random from the dictionary and holding the results up for scrutiny in the way Frankfurt (2005) calls typical of a ‘bull session.’ Similarly, we can imagine a painter learning about their personal aesthetic preferences, or about visual or aesthetic space more generally, by programming a computer to randomly drip paint of different colors onto a canvas.

The possible epistemic significance of the ‘random walk’ technique reveals a sense in which someone might count as a producer of knowledge without counting as an imaginer. The reason I call attention to this point, however, is that I take the exception here to prove the rule. Typically, when I see a song, poem, or painting that you produce as the result of your activity as an imaginer – as the result of your exploratory engagement with ways a world might be – I will see the work as a step in a cognitive process aimed at the acquisition of knowledge about matters that are important to you.

To start with the most solipsistic case, if I take some text you produce to be the result of an imaginative exercise, as opposed to an exercise guided only by a concern to faithfully report the facts about some concrete situation, it is natural for me to read it as a step in a process aimed at mapping the contours of some internal vision of yours. By letting your imagination run, you can explore your understanding of the sublime, or the awful, or learn about what words you think look good next to one another, or what you think is funny, and so on.

Seeing you as an imaginer facilitates my being able to see you as someone whose exploration is aimed at less inwardly directed forms of learning as well. By experimenting with the placement of words on the page, the poet gains quite concrete knowledge about the nature of syntactic and phonetic space, knowledge of a sort that is not typically revealed in our ordinary use of language. It is hardly an exaggeration to say that every poem amounts to a discovery about what can be done within the constraints imposed by syntax and semantics, not to mention meter and rhyme, or theme and genre. Parallel points will hold for other forms of written expression and for other media.

Finally, while the question of whether readers can acquire moral knowledge, knowledge about human psychology, or the ability to more reliably empathize with others by engaging with fiction has been much discussed, the question of whether authors learn from the process of producing it has to the best of my (admittedly likely gappy) knowledge received little attention from philosophers.¹⁶ This oversight may be due to the fact that it would be so incredible to doubt that the process of writing The Brothers Karamazov must have involved substantial engagement with fundamental questions about morality, faith, motivation, and so on. But what is true of Dostoevskii strikes me as not obviously less true of less widely acclaimed or systematic authors: one of the basic ways in which we pose moral questions to ourselves, and answer those questions, is by imagining ways a world might be and then attempting to render those worlds three dimensional in speech, song, or art. The same goes for our views about who we and those around us are, about the nature of the universe and our place in it, and so on.

To summarize, then, I take seeing you as an imaginer to conduce my seeing you as a producer of knowledge concerning your own aesthetic sensitivities, concerning the space of possibilities afforded to you by the modalities in which you are productive, and concerning what philosophers would typically consider fundamental questions in ethics, politics, and metaphysics. Importantly, as I indicated earlier, I do not mean to suggest that seeing a text as the veridical record of some series of events is incompatible with seeing the author as a producer in any of these areas. Friend (2012) makes clear, for example, that we do not typically struggle to see the authors of great works of non-fiction as agents engaged in all of these activities.

The point I take to be crucial where poetic injustice is concerned, however, survives this observation: seeing you as a reporter of facts makes it much harder to see you as someone engaged in the project of limning the contours of a certain aesthetic or moral space, while seeing you as an imaginer makes this sort of activity more visible. Where the answer ‘responding to facts’ is plausibly offered in response to the question ‘what is this agent up to in this discourse?,’ alternative answers like ‘imaginatively exploring the generic conventions of drill rap’ or ‘imaginatively reflecting on the nature of contemporary life in London’ will tend to be outcompeted, and sophisticated epistemic activities will thus go overlooked.¹⁷

4.2. Failure to see the imaginer as a guide

It addition to making it harder to see an agent's engagement in the process of producing knowledge, failing to see someone as an imaginer makes it harder for us to appropriately participate with them in certain forms of epistemically significant joint action. We looked previously at a case Hookway (2010) described, in which one person fails to engage another collaboratively. In addition to work we do together as co-equals, it is a familiar and important fact about our epistemic interactions that we sometimes work together in circumstances in which one of us allows another to take the lead.

The following example will help to bring out the structure of the sort of activity I have in mind:

Modal Logic

A philosophy professor is leading a seminar discussion on modal logic for a group of undergraduates. The group's aim is to construct a counter-model for a particular formula. Some proposals are bruited, but uncertainty ensues and the professor says aloud ‘Hmm … I'm having some problems remembering what the properties of S4 were … was it reflexive and symmetric?’

Most students ignore the preface and recognize the leading question for what it is – they see the professor as someone who has structured the discussion intentionally with the aim of helping students to direct their attention in a certain direction as they try to reason their way through a problem. One student, however, in virtue of a social identity prejudice, is amazed to ‘learn’ that the professor does not know the characteristic properties of S4 and posts an angry diatribe on social media about how standards have been watered down to the point where even basic technical knowledge is not required to secure a teaching appointment.

Students with an appropriate view of our hypothetical professor will not see their ‘claim’ to have forgotten the characteristics of S4 as credible, and will instead see it as the prompt that it is. I imagine this setup will be particularly familiar to philosophers – instead of simply depositing our students at the destination, we are trained to use questions in the classroom as a way of constructing a scaffolding for our students’ individual processes of enquiry to develop inside, in the way that the builder of an obstacle course or the curator of a museum exhibition might aim to frame a visitor's experience. Students who fail to engage with our hypothetical professor in the way one would with a guide, then, and instead see them as just another member of a community of epistemic equals thus mis-assess them as an epistemic actor.¹⁸

One of the ramifications of failing to see someone as an imaginer is that doing so removes one of the natural avenues by which we come to see people as guides, and thus by which we come to engage in the kind of shared activity that a guide would lead us in. When we treat a text (speech, composition, etc.) as the record of or a step in an author's imaginative process, it is natural for us to see the author as someone who aims to lead us on a journey ourselves, for us to see them as a kind of epistemic curator who wants to reveal something to us, or to create the conditions for us to come to a discovery.

When I approach a novel (painting, piece of music, etc.) with a view of the author as imaginer in mind, the question ‘why are things set up this way?’ is always present, forcing me at every step to confront someone who is anticipating my engagement, and who actively aims to shape my participation. By making Raskolnikov's crushing poverty and unrelenting illness so vivid and present, Dostoevskii's Crime and Punishment invites the reader to face the choices the character faces as though she were facing them herself. Confronting those choices, though, involves a kind of double recognition of the author as an epistemic agent – in addition to recognizing the depth and remarkable resolution of the author's vision, we come to see him as a kind of teacher, as someone who has invited us in to a world he constructed for a shared purpose. By the end of The Fall, when we have judged the narrator the way he judges the characters he describes, we realize that Camus has played a trick on us, and we come to see the text as a script that invites our engagement with the central moral question the book addresses in much the way Descartes’ Meditations invite us to engage the question of skepticism for ourselves.

While seeing someone as a guide offering to lead others on a path of imaginative exploration is not in principle incompatible with taking them to be speaking credibly,¹⁹ seeing someone as an imaginative auteur helps to generate in us a certain kind of respect, whether we hold their project in high regard or not. When I see you as a credible reporter, I may see you as someone who made careful choices about what to report on and why, choices that serve you as a guide. But I also may not – I can make your activity perfectly intelligible even if I treat you as an automaton built to provide simple reporting about ordinary matters of fact, or as someone who generates lists based on whatever the tide dragged in. If I see you as an imaginer, however, this kind of reading becomes difficult to sustain. Even if I take a dim view of your vision, either with regard to its resolution, its aims, or the way it is realized, simply by engaging with that vision, I have to concede that you are leading me through it. If I miss your attempt to offer guidance by taking you as a mere repeater of facts, then, I fail to afford you a degree of respect commensurate with your sophistication as an epistemic agent, and also place myself in a higher position than is warranted with regard to you.

5. Conclusion

One of the overarching themes of Fricker's work on epistemic injustice is that the first step towards a remedy for a problem is to provide it a name and thus a place within our conceptual scheme. By calling attention to what I take to be the distinctive wrong of poetic injustice, I hope to have revealed a new application for that point, to have brought out some important new facets of our epistemic engagements with one another, and to have shown something about the set of discursive and epistemic capacities we recognize in those we see as fully-fledged epistemic agents.

In addition to depriving them of a source of social esteem, an important basis for self-respect, and an independent source of income, legal action against Doyley and Malinga amounts to a substantial epistemic undermining – perhaps not in their status as knowers simpliciter, but in their status as imaginers, which ramifies where their status as producers of knowledge and people we can be guided by in our own explorations are concerned. By treating their song ‘Attempted 1.0’ as testimony, police and prosecutors do them a distinctive form of epistemic wrong, and provide a cautionary example that we as philosophers ought to pay attention to.²⁰