2. Fricker on Hermeneutical Marginalization

Fricker builds out her account of hermeneutical injustice from a paradigm case, that of Carmita Wood. In 1975, Wood, an employee in the physics department at Cornell, sought help from a colleague, Lin Farley, and her consciousness-raising group. In her memoir of second-wave feminism In Our Time, Susan Brownmiller quotes Farley's account of what happened to Wood. In Farley's words, ‘a distinguished professor seemed unable to keep his hands off her’. For instance, ‘he'd deliberately brush against her breasts while reaching for some papers’. Once, ‘he cornered her in the elevator and planted some unwanted kisses on her mouth’ (Brownmiller Reference Brownmiller1999: 280). Wood found this treatment unbearable and requested a transfer to another department, but this was refused. So, feeling that she was left with no other choice, she quit her job. Needing the money, she applied for unemployment insurance. Yet when asked by a Department of Labor claims investigator why she had quit her job, ‘Wood was at a loss to describe the hateful episodes’ (1999: 280; Brownmiller's own words). She managed only to say that her reasons were ‘personal’ (1999: 280). Unsatisfied by this answer, the investigator denied her insurance claim; Wood appealed this decision. With her appeal outstanding, Wood recounted her experience to Farley and the consciousness-raising group. It rang bells: several students in one of Farley's seminars had had similar experiences. The group decided to hold a speak-out on the matter. But how to advertise it? Someone suggested the neologism sexual harassment. It was quickly agreed that this would do nicely. As one member of the group put it: ‘Sexual harassment! Instantly we agreed. That's what it was’ (1999: 281).

Of concern to us here is Wood's interaction with the claims investigator.Footnote ⁵ Fricker (Reference Fricker2007: 162) points out that Wood had an interest in what had happened to her being intelligible to the investigator. After all, if he'd grasped that Wood had been seriously wronged and thus deserved unemployment insurance he'd have been much more likely to grant her claim (2007: 162). Yet Wood found herself hindered from rendering what had happened to her intelligible to the investigator by a ‘lacuna’ (2007: 150) in what Fricker calls ‘the collective hermeneutical resource’ (2007: 151) – that is, the set of concepts ‘that just about anyone can draw upon’ (2016: 167; see also Goetze Reference Goetze2018: 76). Though Fricker doesn't state this explicitly, I take it that such a lacuna is supposed to consist in having ‘at best ill-fitting meanings to draw on’ in one's efforts to render something about oneself intelligible to someone (2007: 148; see also Fraser Reference Fraser2018: 734).Footnote ⁶ Thus I take Fricker's idea in declaring Wood hindered by a lacuna in the collective hermeneutical resource to be something like the following. The concepts Wood had available to her in the collective hermeneutical resource, such as unwanted-kissing or being-made-to-feel-uncomfortable, were in some sense ill-fitting for the purposes of rendering what had happened to her intelligible to the investigator (perceiving this to be the case would explain why she didn't offer him a description of what had happened to her using such terms); and conversely, a concept absent from the collective hermeneutical resource, namely sexual harassment, would have been in some sense well-fitting for this purpose (2007: 151–3).

This is, I suggest, a highly plausible description of Wood's situation. But it does leave us with the question of how to understand the key notion of a concept's fittingness. Fricker herself says little to address this question (2007: 148). Yet an answer to it will give us a ‘higher-resolution model of hermeneutical injustice’ (Fraser Reference Fraser2018: 735), something we will need when it comes to making sense of what's going on in the otherwise hard-to-decipher Enduring Love case. For this reason I propose to leave behind what Fricker herself has to say for a moment in order to set out Rachel Fraser's compelling answer to this question. Fraser's suggestion is that a concept's fittingness can be understood profitably in terms of which inferences are liable to be drawn from and to claims in which it appears (2018: 735–7). For these purposes we can represent inferences as ‘ordered n-tuples of claims, with the nth entry as the inference output and the others as input(s)’ (2018: 735). An inference introduces a concept C when C isn't exploited in any of the inference's inputs but is exploited in its output. Conversely, an inference eliminates a concept C when C is exploited in at least one of the inference's inputs but isn't exploited in its output (2018: 736). An inference is cognitively accessible to a person B to the degree that drawing it does not require significant cognitive labour of B (2018: 735). And an inference is socially licensed in B's social context to the degree that (i) those in that context are disposed to regard the inference as legitimate and (ii) this disposition is generally recognized (2018: 735–6). The idea is that B will be liable to draw an inference to the extent that it is both cognitively accessible to them and socially licensed in their social context.Footnote ⁷ Putting all this together, we can say that the inferential network associated with a concept C for person B is the set composed of exactly those inferences which (i) either introduce or eliminate C and (ii) are both cognitively accessible to B and socially licensed in B's social context (2018: 736).

Now imagine that person A wants to render something about themselves intelligible to person B. A concept C is ill-fitting for this purpose to the extent that the inferential network B associates with C either (i) includes inferences which it would be against A's interests for B to draw or (ii) doesn't include inferences which it would be in A's interests for B to draw.Footnote ⁸ Conversely, a concept C is well-fitting for this purpose to the extent that the inferential network B associates with C either (i) doesn't include inferences which it would be against A's interests for B to draw or (ii) includes inferences which it would be in A's interests for B to draw. This framework does an excellent job explaining why Wood would have found a concept such as being-made-to-feel-uncomfortable ill-fitting for the purposes of rendering what had happened to her intelligible to the investigator. The idea is that the investigator wasn't liable to infer “Wood was seriously wronged and so deserving of being paid unemployment insurance” from “The professor made me feel uncomfortable.” Likewise, this framework does an excellent job explaining why Wood could well have found the concept sexual harassment well-fitting for the purposes of rendering what had happened to her intelligible to the investigator. The idea is that the investigator could well be liable to infer “Wood was seriously wronged and so deserving of being paid unemployment insurance” from “The professor sexually harassed me.”

Returning now to Fricker, we've seen that on her account a lacuna in the collective hermeneutical resource hindered Wood from rendering what had happened to her intelligible to the investigator. As a result, her interest in what happened to her being intelligible to him went unsatisfied. The non-satisfaction of an interest is a harm (Feinberg Reference Feinberg1984: 33; Fricker Reference Fricker2007: 162). Yet Fricker's intuition is that Wood wasn't merely harmed here, but rather suffered an injustice. Fricker differentiates between a mere harm and an injustice as follows: ‘[f]or something to be an injustice, it must be harmful but also wrongful, whether because discriminatory or because otherwise unfair’ (2007: 151). So Fricker needs to explain why it was discriminatory or otherwise unfair for Wood's interest to go unsatisfied. Here's where hermeneutical marginalization has a role to play. Call participation in those practices which generate and propagate concepts hermeneutical participation (2007: 152). Fricker suggests that ‘[m]ost obvious’ among such hermeneutically powerful practices are those ‘sustained by professions such as journalism, politics, academia, and law’ (2007: 152).Footnote ⁹ Where participation in hermeneutically powerful practices would have value for members of a group, yet members of that group are subordinated within and/or excluded from those practices, we can say that the members of that group are thereby hermeneutically marginalized (2007: 153).Footnote ^10, Footnote ¹¹

We saw above Fricker plausibly suggest that had the concept sexual harassment been available in the collective hermeneutical resource, Wood would have been able to render what had happened to her intelligible to the investigator. Fricker offers an explanation for why this concept was absent from the collective hermeneutical resource which invokes three claims. The first is that women at the time were significantly hermeneutically marginalized (2007: 152). In other words, women were subordinated within and/or excluded from a range of hermeneutically powerful practices. Conversely, men enjoyed a dominant position within those practices. The second is that ‘interpretive efforts are naturally geared to interests, as we try hardest to understand those things it serves us to understand’ (2007: 152).Footnote ¹² The third is that it wouldn't have served men at the time to generate and propagate a conceptualization of sex-based harassment in the workplace as sexual harassment (2007: 151).Footnote ¹³ Fricker's explanation for why the concept sexual harassment was absent from the collective hermeneutical resource that those in a position to generate and propagate the concept sexual harassment hadn't done so because doing so wasn't in their interest. Fricker's plausible suggestion is thus that the lacuna which hindered Wood from rendering what had happened to her intelligible to the investigator resulted from women's hermeneutical marginalization. A further point Fricker makes is that women's hermeneutical marginalization was symptomatic of women's broader relative social powerlessness (2007: 155). Thus when Wood was harmed as a result of women's hermeneutical marginalization, she was harmed in virtue of her membership in a relatively socially powerless group. It seems straightforwardly discriminatory to be harmed in virtue of one's membership of a relatively socially powerless group, and thus in virtue of an aspect of one's social identity (2007: 155). This is the reason Fricker gives for saying that Wood was not merely harmed – as she would have been had her difficulties resulted instead from e.g. epistemically culpable behaviour on her part – but rather suffered a hermeneutical injustice (2007: 153).

By generalizing from this case, Fricker arrives at her account of hermeneutical injustice. Thus on Fricker's account, a person A suffers a hermeneutical injustice iff the following four conditions obtain. First, A has an interest in something about themself, X, being intelligible to someone, B. Second, A's attempt to render X intelligible to B is hindered by a lacuna in the collective hermeneutical resource, as a result of which their interest goes unsatisfied. Third, this lacuna results from the hermeneutical marginalization of some group. Fourth, A is a member of this hermeneutically marginalized group.

At this point, Fricker adds in a complication. She suggests that we can think of what befell Wood as a systematic hermeneutical injustice insofar as the hermeneutical marginalization which gave rise to it was symptomatic of a broader relative social powerlessness (2007: 156). In Fricker's view, not all cases of hermeneutical injustice are like this. She writes that ‘there can be cases of hermeneutical injustice that are not part of the general pattern of social power, and are more of a one-off. They are not systematic but incidental’ (2007: 156). Fricker takes her paradigm example of an incidental hermeneutical injustice from Ian McEwan's novel Enduring Love (2007: 156–8). In the novel the protagonist, Joe, a journalist, is stalked by a man called Jed Parry. Parry suffers from the delusion that Joe is in love with him and is determined to convert Joe to Christianity. When Joe informs the police over the phone that Parry has been ‘harassing’ him, the following conversation ensues:

What emerges from this conversation is that Joe's conceptualization of Parry's behaviour as an instance of harassment is ill-fitting for the purposes of rendering Parry's behaviour intelligible to the police officer. Included in the inferential network associated with harassment for the officer is an inference from “This man is harassing me” to “This man is causing me physical harm, and/or is harming my property, and/or is threatening to do so”. Thus when it transpires that Parry isn't doing any of these things, the officer concludes that Joe isn't being harassed. The other conceptualizations of Parry's behaviour which Joe offers towards the end of the conversation, for instance as the outcome of an obsession, appear no better-fitting. The conceptualization of Parry's behaviour as an attempt at conversion turns out to be particularly ill-fitting, since the inferential network the officer associates with this concept includes an inference from “He's attempting to convert me” to “What he's doing is merely an expression of religious freedom”. Thus it seems that Joe is hindered by a lacuna in the collective hermeneutical resource. The consequent frustration of his attempt at rendering Parry's behaviour intelligible to the officer harms Joe, since in the absence of police intervention Parry's behaviour is allowed to escalate, culminating in a murder attempt.

Fricker's intuition is that Joe wasn't merely harmed here but rather suffered an injustice (2007: 157).Footnote ¹⁴ In Fricker's view, we need to tell a somewhat different story to explain why it was discriminatory or otherwise unfair for Joe to be harmed in this way. Fricker's reason for thinking this is that in contrast to Wood's case, the lacuna in the collective hermeneutical resource ‘has nothing to do with any general social powerlessness or any general subordination as a generator of social meaning, for his social identity is that of the proverbial white, educated, straight man’ (2007: 157). This is a key passage. Fricker is of course right that white, educated, straight men as a group are neither relatively socially powerless in general nor hermeneutically marginalized in particular. Yet it doesn't follow that the lacuna ‘has nothing to do with any general social powerlessness’, since it could have everything to do with the social powerlessness in general and the hermeneutical marginalization in particular of some other group. The inference goes through, then, only on the assumption that in a case of hermeneutical injustice the wronged party must themselves be a member of the group whose hermeneutical marginalization it is that gives rise to the injustice.Footnote ¹⁵

This assumption appears as the fourth condition in Fricker's account of hermeneutical injustice as laid out above, and I will return to challenge it in §4. For now, notice what sort of story this assumption leads Fricker to tell about the ‘incidental’ hermeneutical injustice of which Joe is a victim. Fricker still wants to invoke hermeneutical marginalization to explain why the frustration of Joe's attempt at rendering Parry's behaviour intelligible to the officer is unfair. Given the membership condition, Joe must be a member of the group whose hermeneutical marginalization it is that gives rise to the injustice he faces. Yet no group of which he's a member in virtue of some aspect of his social identity is hermeneutically marginalized. Fricker's solution is to say that the group whose hermeneutical marginalization gives rise to this particular hermeneutical injustice is a group with only one member, namely Joe (2007: 157–8). The idea seemingly is that it is unfair to be harmed as a result of one's subordination within or exclusion from some hermeneutical practice which would have value for one.Footnote ¹⁶

As I will show in §3, Fricker doesn't make much of a case for thinking that Joe really is hermeneutically marginalized. For now the point I want to emphasize is that Fricker explains what's unjust about incidental hermeneutical injustices differently from how she explains what's unjust about systematic hermeneutical injustices. In systematic cases, the group whose hermeneutical marginalization gives rise to the injustice is supposed to be a relatively socially powerless group whose members are those sharing a particular social identity with the wronged party. In these cases, the harm done to the wronged party is supposed to befall them qua someone with that social identity, and so be discriminatory. By contrast, in incidental cases the group whose hermeneutical marginalization gives rise to the injustice is supposed to be a group with just one member, namely the wronged party. In these cases, the harm done to the wronged party is supposed to befall them qua someone subordinated within or excluded from some hermeneutical practice which would have value for them, and so be unfair. Having unpacked what Fricker has to say about hermeneutical marginalization and its role in hermeneutical injustice, I now want to argue that Fricker's picture is significantly flawed.

3. The Problem with Fricker's Picture

The main problem with Fricker's picture is that it can't properly account for Joe's case as an instance of hermeneutical injustice. In particular, it seems clear that Joe isn't hermeneutically marginalized – as the picture requires him to be. That Fricker's picture can't properly account for Joe's case as an instance of hermeneutical injustice has been pointed out before, in a somewhat-neglected paper by Komarine Romdenh-Romluc (Reference Romdenh-Romluc2016: 596–600). In this section I will give some novel reasons for arriving at the same conclusion. In the next section I will go beyond Romdenh-Romluc in offering a solution to the problem.

Fricker says little to suggest that Joe is hermeneutically marginalized, and what little she does say is not very persuasive. Fricker writes only that the ‘trivializing interpretations’ of Parry's behaviour (as not a matter for the police and as merely an expression of religious freedom) put forward by the officer ‘mean that Joe's hermeneutical participation is hindered in respect of a significant, if highly localized, patch of his social experience, and for this reason his case qualifies as a hermeneutical injustice’ (2007: 157–8). Fricker's idea is that in putting forward trivializing interpretations of Parry's behaviour the officer thereby hermeneutically marginalizes Joe. For at least two reasons this isn't very convincing. First, the ‘trivializing interpretations’ Fricker cites are put forward by the officer only towards the end of the conversation. By this point Joe's attempt at rendering Parry's behaviour intelligible to the officer has already been frustrated by the lacuna in the collective hermeneutical resource. Thus, on Fricker's view, Joe's hermeneutical marginalization is bizarrely alleged to take place only after the hermeneutical injustice to which it supposedly gives rise. Second, Fricker earlier defined hermeneutical marginalization in terms of subordination within and/or exclusion from those ‘practices’ in which concepts are generated and propagated (2007: 153). Fricker gave as examples of such practices ‘those sustained by professions such as journalism, politics, academia, and law’ (2007: 152). Yet it is unclear what such practice the officer could be subordinating Joe within or excluding Joe from.

Fricker concedes some doubt as to whether Joe is hermeneutically marginalized in later work, describing him as ‘an educated, white, middle-class man, whose hermeneutical marginalization (if any – it is the vanishingly minimal case) is highly specific, localised to the particular matter at hand’ (2016: 166; my emphasis). With Romdenh-Romluc, I suggest we can go further and say definitively that Joe isn't hermeneutically marginalized. After all, not only is he educated, white, middle-class, cis, and straight. He is in addition a journalist, and thus a participant in one of the practices Fricker suggests is most hermeneutically powerful. ‘As such,’ Romdenh-Romluc writes, ‘Joe is surely well-placed to take part in the cultural discourse about stalking; he does not face obstacles to hermeneutical participation on this issue’ (2016: 599).

Let's take stock. Fricker begins from the intuition that the harm which befalls Joe is unfair, so constituting an injustice. Fricker then offers to explain why the harm which befalls Joe is unfair in terms of Joe himself being hermeneutically marginalized. Yet it seems clear that in fact Joe is not himself hermeneutically marginalized, and so Fricker's proposed explanation fails. Thus as things stand the intuition from which Fricker begins is yet to find a vindication. Hence it may seem tempting simply to give up the idea that Joe suffers an injustice, settling instead for the weaker claim that Joe is merely harmed.Footnote ¹⁷ Yet the intuition that the harm which befalls Joe is unfair strikes me as a strong one, as it does Romdenh-Romluc (Reference Romdenh-Romluc2016: 600). As such, it is worth looking to see whether a different explanation for why the harm which befalls Joe is unfair can be found. In the next section, I show that an explanation can be given once the membership condition is dropped from the account of hermeneutical injustice. Since this explanation seemingly vindicates the intuition from which Fricker begins, those in favour of giving up the idea that Joe suffers an injustice will first need to find a way to debunk this explanation.

4. Solution to the Problem

Though Romdenh-Romluc sees that Fricker's picture is significantly flawed, she doesn't see any good way to repair it, and so leaves her readers in aporia (Romdenh-Romluc Reference Romdenh-Romluc2016: 609). In my view, Romdenh-Romluc's difficulty in seeing a solution stems from her acceptance of the membership condition (2016: 597). In this section I will argue for dropping this condition.

Crucially, dropping this condition allows us to tell a more plausible story about how the hermeneutical injustice Joe suffered came about. The suggestion I want to explore is that the lacuna in the collective hermeneutical resource hindering Joe can be traced back to the hermeneutical marginalization of women who, after all, constitute the majority of stalking victims (Proctor Reference Proctor and Lombard2018: 109; Office for National Statistics 2020). This suggestion can be made a bit more concrete if we ask ourselves why Joe didn't offer a conceptualization of Parry's behaviour as stalking to the officer.

Suppose Joe claimed to the officer that Parry was stalking him. It seems plausible that the officer would have rejected such a claim. For consider the following historical developments in the conceptualization of stalking. The first time the term “stalking” was used in print to describe behaviour other than the hunting of wild animals was in The Washington Post in 1976, where it named the notorious serial-killer Son of Sam's habit of following his female victims around immediately prior to killing them (Kamir Reference Kamir2001: 146–7). The term was then taken up by reviewers to describe the male protagonist's behaviour in films such as John Carpenter's Halloween (1978), Fred Walton's When a Stranger Calls (1979) and Stanley Kubrick's The Shining (1980) (Kamir Reference Kamir2001: 148). These films were early entries in an emerging horror subgenre whose generative story was that of ‘a psychokiller who slashes to death a string of mostly female victims, one by one, until he is subdued or killed, usually by the one girl who has survived’, most commonly referred to as “slasher” but also often as “stalker” films (Clover Reference Clover2015: 21). In 1990 the first law in the US to outlaw stalking was passed in California, in the wake of the murder of actress Rebecca Schaeffer by a male fan (Kamir Reference Kamir2001: 175). This law was rushed through in response to wall-to-wall press coverage of the murder and lobbying from groups representing Hollywood stars, with legislators making ‘no attempt … to investigate and analyze the real social problem of stalking’ (Kamir Reference Kamir2001: 175). Laws modelled on the California legislation were then passed in every US state by the end of 1993 (Kamir Reference Kamir2001: 175). One effect of this wave of legislation was to reinforce in the public consciousness the idea that stalking was principally something that befell prominent public figures. A 2012 parliamentary inquiry found that this idea had also gained significant traction in the UK. For instance, the inquiry noted that the Metropolitan Police had in 2006 set up a dedicated “fixated threat assessment centre” specifically to carry out risk assessments in cases of stalking involving the rich and famous, but rarely carried out similar risk assessments in cases of stalking involving members of the general public (Travis Reference Travis2012).

In light of these various developments, it seems plausible that the inferential network the officer associated with the concept stalking would include some inferences such as “He's a man” to “He's not being stalked”; “He's not a celebrity” to “He's not being stalked”; “His narrative does not conform to common horror movie tropes” to “He's not being stalked”; and “The alleged stalker hasn't shown any signs of being about to murder him” to “He's not being stalked”. The concept stalking would then be an ill-fitting one for Joe, since if he offered the officer a conceptualization of Parry's behaviour as an instance of stalking the officer would likely reject that conceptualization.

Notable is that the various developments in the conceptualization of stalking just discussed were all in the main driven by men. It is worth recalling that male domination of journalism in the 1970s US was a particular focus for second-wave feminist activism (Brownmiller Reference Brownmiller1999: Ch. 6). Moreover, crime-reporting and film-reviewing both stand out as sub-fields of journalism particularly hard for a woman to break into, then as now (on the latter, see Lauzen Reference Lauzen2022). The numerous stalker films of the 1970s and 80s cited by Orit Kamir in her Every breath you take: stalking narratives and the law as particularly influential were all written and directed by men (Kamir Reference Kamir2001: 148; similarly Clover Reference Clover2015: Ch. 1). The California state legislature of the early 1990s, again, had men well in the majority.

Since men are much less likely than women to be stalked, it is plausible that those driving the various developments in the conceptualization of stalking mentioned above weren't much interested in understanding stalking as a real social problem – as indeed Kamir explicitly argues was true of the California legislature. Conversely, since women are much more likely than men to be stalked, it is plausible that they would have been more interested in understanding stalking as a real social problem had they been in a position to drive a different set of developments in the conceptualization of stalking. Thus here we have a story to tell about stalking being an ill-fitting concept for the purposes of rendering Parry's behaviour intelligible to the officer as a result of women's hermeneutical marginalization. Moreover, stalking is the most obvious candidate to be a well-fitting concept with which to render Parry's behaviour intelligible to the officer.Footnote ¹⁸ So the story is in fact one in which women's hermeneutical marginalization gives rise to the lacuna in the collective hermeneutical resource that hinders Joe from rending Parry's behaviour intelligible to the officer. As such, it is a story about Joe as a victim of patriarchy.

This isn't the only plausible story that could be told about this case – it could be supplemented with others. For instance, suppose stalking was ill-fitting for Joe because the officer was liable to infer “He's not being stalked” from “He's a man, and so is the alleged stalker”. It wouldn't be surprising were the officer liable to think along these lines. As already noted, representations of men being stalked were rare in comparison to representations of women being stalked. Even rarer were representations of men being stalked by men. Much more common were representations of men being stalked by women, perhaps the most famous example being Adrian Lyne's Fatal Attraction (1987) (Kamir Reference Kamir2001: 5). The hermeneutical marginalization of gay men could have an explanatory role to play here. Gay men are much more likely to be stalked, and stalked by a man, than straight men are (Office for National Statistics 2020). So it is plausible that had they been in a position to drive a different set of developments in the conceptualization of stalking, gay men would have paid more attention to the problem of men stalking men than the various straight men I've cited in fact did during this period. For our purposes what's relevant about this possible supplementary story is that Joe again isn't a member of the group to whose hermeneutical marginalization it appeals.

At this point someone might wonder: mightn't the hermeneutical marginalization of stalking victims also play a role in bringing about the lacuna in the collective hermeneutical resource that hinders Joe? After all, similar explanations have been thought plausible in other cases of hermeneutical injustice; for instance, Katharine Jenkins attributes the frequent ill-fittingness of the concept rape to the hermeneutical marginalization of rape victims (2017: 197). And Joe himself is a victim of stalking; so were this to play a role in bringing about the lacuna, we could acknowledge him to be a victim of hermeneutical injustice without dropping the membership condition. Admittedly, there is nothing impossible about this suggested explanation. But the point is: suppose it turned out that the hermeneutical marginalization of stalking victims didn't play a role in bringing about the lacuna. Rather, it turned out just to be the hermeneutical marginalization of women, or the hermeneutical marginalization of women in combination with the hermeneutical marginalization of gay men, that did this. Wouldn't we still be inclined to say that Joe was a victim of hermeneutical injustice? I think we would – in which case the membership condition needs to be dropped.

At the same time, if it were shown that the lacuna didn't result from the hermeneutical marginalization of at least some group, we would simply cease to have the intuition that Joe was a victim of hermeneutical injustice. That's because the frustration of his attempt at rendering Parry's behaviour intelligible to the officer would then no longer strike us as unjust, but instead as merely harmful. We still need to invoke hermeneutical marginalization, in other words, to explain why Joe being frustrated in this way was unfair. The hermeneutical marginalization of a group to which Joe doesn't himself belong can do this explanatory work, I suggest, for the reason that it is intuitively unfair to suffer collateral damage as a result of the hermeneutical marginalization of some group to which one does not belong.

So one reason for dropping the membership condition is that it puts us in a better position to account for what happened to Joe as a case of hermeneutical injustice. This constitutes a solution to the main problem plaguing Fricker's picture discussed in §3. An additional reason for dropping the membership condition is the following. Consider the case of a hypothetical man, call him Smith, working at a New England university in the early 1970s. Like Carmita Wood, Smith is sexually harassed at work by his boss, sees no choice but to resign, and applies for unemployment benefits. Again like Wood, Smith is subsequently interviewed by a claims investigator; a lacuna in the collective hermeneutical resource hinders Joe from rendering his boss's behaviour intelligible to this investigator, who consequently denies Smith's application. Since women's hermeneutical marginalization accounts for the lacuna in collective hermeneutical resource that hinders Wood, it seems likely that women's hermeneutical marginalization would go at least a long way towards accounting for the lacuna in the collective hermeneutical resource that hinders Smith.Footnote ¹⁹ In any case, let's stipulate that this lacuna resulted from the hermeneutical marginalization of some group(s) to which Smith himself did not belong. Intuitively, Smith suffers an injustice. Moreover, the similarities with the injustice Wood suffers give us every reason to count this as a hermeneutical injustice.Footnote ²⁰ Yet doing so will require us to drop the membership condition.

Dropping the membership condition leaves us with Fricker's first three conditions. My tentatively held view is that these three conditions are individually necessary and jointly sufficient for a hermeneutical injustice to occur. I will close out this section by considering an objection to this view. Someone could hold that one's interest in intelligibility going unsatisfied as a result of the hermeneutical marginalization of some group to which one does not belong is sometimes unfair – but not always. Such a person could well share my intuitions that Smith and Joe suffered hermeneutical injustices. Still, in their view Fricker's first three conditions aren't jointly sufficient for a hermeneutical injustice to occur. The motivating thought is likely to be that whether it is unfair for one's interest in intelligibility to go unsatisfied depends on whether one is complicit in the hermeneutical marginalization of the relevant group. For instance, suppose it transpired that Joe had gone out of his way as a journalist to prevent women from joining his hermeneutically powerful profession. On the view under consideration, this would show that Joe wasn't a victim of hermeneutical injustice after all. Someone persuaded of this view would say that the non-satisfaction of Joe's interest in intelligibility had turned out not to have been an instance of collateral damage after all, but instead an instance of someone getting their just deserts. If we accept this view, we'll need to add an extra condition to the account of hermeneutical injustice specifying that the wronged party isn't complicit in the hermeneutical marginalization of the relevant group.

Though I feel the pull of this view, I'm ultimately unpersuaded by it. Continue to suppose for the sake of argument that Joe went out of his way to exclude women from the journalistic profession. Obviously enough, his behaving in this way would be seriously wrong. He would deserve to be held to account for it, and this would almost certainly involve the imposition of some sort of social sanction. But we can maintain this and at the same time deny that Joe would deserve the non-satisfaction of his interest in intelligibility. And in fact I think we should deny this latter claim. For one thing, it is hard to imagine how Joe's interest in intelligibility going unsatisfied could play a role in holding him to account. Indeed, it is hard to believe that any path to greater justice begins with involved parties failing to understand significant things about each other. For another, the frustration of Joe's attempt at rendering Parry's behaviour intelligible to the officer has unpredictable and potentially extreme consequences – in the novel, Joe only narrowly avoids being murdered. Finally, it seems to me that much of this latter claim's attraction stems from implicitly subscribing to the retributive logic of an-eye-for-an-eye. One thing wrong with hermeneutically marginalizing a social group is that it renders members of that group vulnerable to having their interests in intelligibility go unsatisfied. This is plausibly what would lead someone to think that a person complicit in the hermeneutical marginalization of some group deserves to have some of their own interests in intelligibility go unsatisfied. Yet an-eye-for-an-eye is widely acknowledged to be a poor guide to justice. These considerations persuade me that even if Joe was himself complicit in the hermeneutical marginalization of women, the non-satisfaction of his interest in intelligibility would be an instance of collateral damage and not of just deserts. Hence my tentatively held view that Fricker's first three conditions are individually necessary and jointly sufficient for an instance of hermeneutical injustice.

5. Rethinking the Systematic/Incidental Distinction

If my suggestion that it was primarily women's hermeneutical marginalization that gave rise to the lacuna in the collective hermeneutical resource that hinders Joe is accepted, we will need to rethink the distinction between systematic and incidental hermeneutical injustices. Recall that Fricker offered what befell Carmita Wood as a paradigm systematic hermeneutical injustice and what befell Joe as a paradigm incidental hermeneutical injustice. I agree with Fricker that there is intuitively an important difference between these two cases. Yet if we're to capture this difference by describing the first as systematic and the second as incidental, the intensions Fricker gives to these two terms will have to be thrown out. For recall that Fricker proposes a hermeneutical injustice is systematic insofar as the hermeneutical marginalization which gives rise to it is symptomatic of a broader relative social powerlessness; otherwise, it is incidental (2007: 156). I've suggested that it was primarily women's hermeneutical marginalization which gave rise to the hermeneutical injustice Joe suffers. Women's hermeneutical marginalization is obviously symptomatic of a broader relative social powerlessness. Thus the hermeneutical marginalization which gave rise to the hermeneutical injustice Joe suffers is symptomatic of a broader relative social powerlessness – just not that of a group to which Joe himself belongs. Yet what befell Joe was supposed to be an incidental hermeneutical injustice.

Fricker comes much closer to capturing what's incidental about Joe's case not in her official view, but in an offhand remark. She notes that Joe ‘suffers the injustice not because of, but rather in spite of, the social type he is’ (2007: 158). I've tried to suggest something like this idea already, in describing the non-satisfaction of Joe's interest in intelligibility as an instance of collateral damage. What's incidental about a case like Joe's, then, is not that the hermeneutical marginalization which gives rise to it isn't symptomatic of a broader relative social powerlessness. Indeed, I'm sceptical that there is any instance of hermeneutical marginalization which isn't in some way symptomatic of a broader relative social powerlessness. Rather, what's incidental about a case like Joe's is that he isn't himself a member of the relevant relatively socially powerless group – he isn't among the targets of the relevant underlying operations of social power. My proposal is thus that a hermeneutical injustice is systematic insofar as the wronged party is themselves a member of the group whose hermeneutical marginalization gave rise to the injustice; otherwise, it is incidental. This proposal successfully recognizes Wood's as a systematic case and Joe's as an incidental case.

An interesting contrast between Fricker's picture and mine emerges when we ask: can incidental hermeneutical injustices be systematically prevented? On Fricker's picture, incidental hermeneutical injustices result from ‘one-off moment[s] of powerlessness’ (2007: 156) and hence are more-or-less unpredictable. Thus it is hard to imagine there being any way to systematically prevent them. By contrast, on my picture both incidental and systematic hermeneutical injustices result from the hermeneutical marginalization of the same groups – for instance, women. Hence both incidental and systematic hermeneutical injustices can be systematically prevented by diminishing and eventually eliminating the hermeneutical marginalization of these groups – something we have a responsibility to do in any case, given the intrinsic badness of hermeneutical marginalization (Romdenh-Romluc Reference Romdenh-Romluc2017: 12; Crerar & Goetze Reference Crerar, Goetze, Townsend, Tietjen, Schmid and Staudigl2022: 100). Notably, in carrying out this responsibility we can systematically prevent incidental hermeneutical injustices while keeping systematic hermeneutical injustices ‘central from our point of view’ as both theorists and political actors (Fricker Reference Fricker2007: 158) – a desirable feature, given that systematic hermeneutical injustices are both more prevalent and play a role in patterns of oppression. The upshot is that on my picture we can be somewhat more optimistic about our prospects for eventually eliminating incidental hermeneutical injustices than we can be on Fricker's.

6. Hermeneutical Injustice is Widespread

If I am right that a person can suffer a hermeneutical injustice without themselves being a member of the group whose hermeneutical marginalization gives rise to it, it seems likely that somewhat more people suffer hermeneutical injustices than Fricker realizes. After all, recall Smith. On Fricker's picture, Smith can't have suffered a hermeneutical injustice because he isn't himself a member of the relevant hermeneutically marginalized group(s). On my picture, by contrast, Smith did suffer a hermeneutical injustice. Since this accords with what I expect will be a widely shared intuition about the case, I take this to be another advantage of the picture I develop.

At this point, an anonymous reviewer raises the following worry. Consider Joe's case one last time. Suppose Joe came to you after his telephone conversation with the police officer and told you what had happened. It is unclear how identifying what had happened to him as a case of hermeneutical injustice would provide you with practical guidance for how to help Joe. If I am right that there are more cases of hermeneutical injustice resembling Joe's than Fricker realizes, then this multiplies the number of such scenarios – which looks like a problem. The first thing to say in response is that this is why, in common with the rest of the literature on hermeneutical injustice, in my brief comments on what to do about incidental hermeneutical injustices I've placed the emphasis on prevention rather than restitution (see e.g. Fricker Reference Fricker2007: 169–75; Barnes Reference Barnes2016: 183; Hull Reference Hull2017: 585; Jenkins Reference Jenkins2017: 201–2; Goetze Reference Goetze2018: 85–6). In general, what to do about hermeneutical injustices before they arise seems like a much more tractable problem. Second, suppose that some of the additional hermeneutical injustices I think Fricker misses did for some reason prove particularly difficult to do something about. It seems to me acknowledging both these injustices and the difficulty in doing anything about them would be preferable to not acknowledging them at all – not least because acknowledging an injustice is also a way of acknowledging the victim of that injustice. Finally, saying much more here would take me outside the scope of this paper. I have primarily been concerned with how incidental hermeneutical injustices arise, and what is unfair about them. What can be done about them is a separate question, worthy of a standalone paper. As such, this paper should be taken as a spur for further work.

I want to close by reconsidering Laura Beeby's relevant suggestion that the professor who harassed Carmita Wood ‘might be both a perpetrator of sexual harassment and a victim of hermeneutical injustice’ (2011: 483). In Epistemic Injustice, Fricker gives two reasons for thinking that the professor can't possibly have been a victim of hermeneutical injustice. First is that the professor was not himself a member of any hermeneutically marginalized group (2007: 153). Given the argument of §4, this doesn't stand up as a reason for thinking that the professor can't possibly have been a victim of hermeneutical injustice. Second is that the professor didn't have an interest in his behaviour towards Wood being intelligible to him (2007: 151). On this point I disagree, though in a way I recognize not everyone will accept.Footnote ²¹ I find intuitive the Aristotelian idea that moral goodness is necessary – though not sufficient – for human flourishing in any meaningful sense (Tessman Reference Tessman2005: 11; Aristotle Reference Ross2009: NE 1).Footnote ²² Not understanding that you're seriously wronging someone is a significant barrier to achieving moral goodness. Thus I think the professor did have an interest in his behaviour towards Wood being intelligible to him – call this a moral interest.

So it seems to me at least possible that the professor was a victim of hermeneutical injustice. But was he, in fact, such a victim? Of this I'm not so sure. Suppose he really didn't understand that he was seriously wronging Wood; that this was thanks to a lacuna in the collective hermeneutical resource; and that this lacuna resulted from women's hermeneutical marginalization. If all this was the case, then: yes, it seems to me that the professor suffered a hermeneutical injustice. But whether all this was the case is highly disputable. In particular, it is far from obvious that the professor didn't understand he was seriously wronging Wood – at least for all Brownmiller tells us about him (1999: 280–1; likewise the account given in Strebeigh Reference Strebeigh2009: 220–3).

In allowing that the professor might have been both a perpetrator of sexual harassment and a victim of hermeneutical injustice, I don't at all mean to excuse his behaviour.Footnote ²³ Nor do I mean to engage in what Kate Manne calls himpathy, which paradigmatically takes the form of showing ‘excessive sympathy … toward male perpetrators of sexual violence’ (Manne Reference Manne2018: 197). I mean only to draw attention to one way in which the politically useful feminist claim that patriarchy is bad for men as well as for women might sometimes be true.Footnote ²⁴ In her ethically rich Burdened Virtues: Virtue Ethics for Liberatory Struggles, Lisa Tessman talks of moral damage occurring when the ‘unconducive conditions of oppression’ hinder someone from cultivating a morally good self (2005: 4). Being hermeneutically marginalized is one way in which a group can be oppressed.Footnote ²⁵ And not understanding when one is seriously wronging another is a significant barrier to cultivating a morally good self. Thus supposing women's hermeneutical marginalization resulted in the professor really not understanding that he was seriously wronging Wood, this would be an instance of moral damage. Allowing that the professor might have been a victim of hermeneutical injustice commits us to thinking that his being morally damaged in this way would be not merely harmful but wrongful. Yet this needn't diminish our determination to condemn him as a perpetrator of sexual harassment and so a perpetuator of patriarchy. Rather, we should think of this commitment as serving ‘to increase the breadth of [our] complaint about systems of oppression’ (Tessman Reference Tessman2005: 5).