Welcome to what promises to be an engaging discussion of Aidan McGlynn’s “Objects or Others? Epistemic Agency and the Primary Harm of Testimonial Injustice.” The paper was recently published in Ethical Theory and Moral Practice and may be found here. What follows is a critical précis by Lauren Leydon-Hardy. Please join the discussion!
Take it away, Lauren:
What is the primary harm of testimonial injustice?
A Critical Précis of Aidan McGlynn’s, “Objects or Others? Epistemic Agency and the Primary Harm of Testimonial Injustice”
In Fricker’s initial framework for testimonial injustice she characterized the phenomenon as the attribution of a credibility deficit owing to an identity prejudicial attitude on the part of the hearer, leading to a failure of uptake for the proffered testimony. She argued that the central harm of testimonial injustice—the primary, non-contingent harm—was that testimonial injustice amounted to a kind of epistemic objectification, wherein speakers were treated as objects to be observed, rather than persons to be interacted and collaborated with. In the intervening years objections have proliferated on all sides, having to do with both the schematization of the phenomenon and the identification of the locus (and nature) of its attendant harm(s). Aidan’s paper, here, focuses on the latter.
The argument against characterizing the primary harm of testimonial injustice as a form of objectification comes out of a suite of counterexamples. If we have cases of testimonial injustice without objectification, then objectification would seem not to be its primary harm. On the basis of several purportedly tough cases some have proposed that what is at issue in cases of testimonial injustice is, instead, a kind of ‘othering’. Othering accounts hold that testimonial injustice needn’t involve treating someone as inert or object-like, or in failing to recognize their epistemic agency. Instead the harm lies in only partially recognizing the speaker’s epistemic agency.
What unifies the tough cases is that they involve testimonial injustices in which the speaker’s epistemic agency is acknowledged, at least inasmuch as they are identified as knowers, and so as (potential) participants in the knowledge economy. Why do the tough cases spell trouble for Fricker’s objectification thesis? Because according to Fricker ‘epistemic objectification’ is a matter of being treated as ‘epistemically inert’, where epistemic inertia is understood as being “devoid of epistemic agency”. (5) So, the primary harm of testimonial injustice is epistemic objectification, and one is epistemically objectified when one is treated as epistemically inert, which is to say, when one is treated as lacking epistemic agency. The tough cases are tough because we should both have the intuition that the speakers have epistemic agency (and so are not epistemically objectified) while also being victims of testimonial injustice.
Here is a brief taxonomy of the kind of cases that would seem to motivate the rejection of the objectification account:
- Exploitation: Testimonial injustices that arise from treating a speaker’s epistemic agency as (merely) instrumentally valuable to the hearer
- Excess: Testimonial injustices arising from attributions of credibility excess
- Eschewal: Testimonial injustices in cases where the hearer recognizes the testifier as a knower, but where other (perhaps prejudicial) attitudes lead the hearer to reject the testimony
Exploitation cases involve treating a speaker as “an informant, and not merely as a source of information.” (4) The idea here is that a speaker can legitimately participate in an exchange and yet fail to be recognized as “having an epistemic agenda of their own.” (4) This is because the exchange concerns only the epistemic agenda of the inquirer, with the speaker serving as an epistemic subordinate. In these cases, it’s not that the speaker is being treated as a mere source of information, to be extracted from, but that they are being treated as an informant, tasked with pursuing the epistemic agenda of an exploiter. The speaker is thus exploited in their capacity as an epistemic agent, rather than having their epistemic agency denied, pace Fricker.
Excess cases involve the attribution of identity-based credibility excesses, as when one is assumed to have some knowledge or expertise in virtue of being x. These cases suggest that the reasons we are believed matter to us, perhaps even as much as that we receive our due credibility in the first place. Like Exploitation cases, Excess cases are understood not to involve the denial of epistemic agency in that they are cases in which the victim of testimonial injustice is treated as a source of knowledge.
Eschewal cases involve recognizing the speaker as a knower but downgrading their credibility assignment on the basis of their alleged lack of trustworthiness; as Frankfurt taught us, in order to lie you must first take yourself to know the truth. Unlike Exploitation and Excess cases, Eschewal victims are not treated as a source of knowledge, but for reasons other than a presumed lack of knowledge. For that reason, Eschewal cases are also understood to be examples of testimonial injustice that do not posit a lack of epistemic agency.
What precisely is meant by ‘epistemic agency’ here is, admittedly, a bit opaque to me. The concept of epistemic agency is not uncontroversial; some people regard it as functionally meaningless, others as hopelessly onerous. I am not one of those people—I think that the concept is good and useful (essential, even) for understanding certain phenomena. However, it often seems to be a quasi-technical stand-in for ‘being believed’ or ‘having knowledge’. If this is all that is meant by ‘being treated as an epistemic agent’ I find myself rather sympathetic to the charge that the notion just isn’t all that philosophically illuminating. This is not a mark against Aidan’s project here, of course, merely an invitation to think further about what precisely it means to be ‘treated as an epistemic agent’ and, relatedly, what it means to have one’s epistemic agency disrupted. I digress.
The first part of Aidan’s argument is to show us why the tough cases aren’t so tough after all. The idea is that, given a suitably enriched conception of epistemic objectification, the putative tension here can be dissolved. This is because the tension depends on Fricker’s selective application of Nussbaum’s account of objectification, viz. the denial of agency. But, as Nussbaum has shown us, denial of agency is just one of seven ways of objectifying another. Other modes of objectification include, for example, treating an individual as fungible, as an instrument, or as violable. Aidan contends that embracing the full breadth of Nussbaum’s account of objectification, countenancing the plurality of ways in which a person can be objectified, allows us to bring the tough cases back in line with the objectification account. Exploitation can be explained as objectification through instrumentality, treating a knower as an epistemic tool for the objectifier’s purposes. Excess can be explained as objectification by treating someone as epistemically fungible, as in “All x’s know about y.” If all x’s know y, then it doesn’t really matter which x you ask. Something similar could be said for familiar cases of Eschewal, including the case of Tom Robinson in To Kill a Mockingbird. On the basis of a racist stereotype, the jury believes not that Robinson doesn’t know, but that he’s knowingly lying. But the same explanation that leads the jury to believe that Robinson is lying would lead the jury to believe that any Black man was lying. That is, they don’t think that Tom Robinson lies; they think all Black men lie. Interestingly, this isn’t exactly the way that Aidan goes in explaining Eschewal cases; in a previous paper he argues that Tom Robinson’s case is a combination of instrumentality and fungibility, but his real emphasis there is on instrumentality. He writes,
Robinson’s attempted testimony is treated as material in the service of the epistemic projects of others, and his own intention to communicate his knowledge of what happened is systematically thwarted. (McGlynn, 2019: 13)
Yet a third explanation might be given in terms of ‘denial of subjectivity,’ another mode of objectification identified by Nussbaum. Suppose that the objectifier downgrades the testifier’s credibility for the reason that the content of their testimony does not reflect the worldview of the objectifier. If their testimony is met with this kind of imaginative resistance such that the testimony is construed as mendacious, this might be an example of the objectifier “treating the object as something whose experiences and feelings (if any) need not be taken into account.” (McGlynn, 2020: 3)
It strikes me as a virtue of the enriched objectification account that it shows this much dexterity and nuance in modeling what are quite disparate cases. The various modes of objectification don’t seem to be offering competing explanations. Rather, it seems that at least in some cases we will want to tease out multiple modes of objectification operating simultaneously. And, plausibly, a one-size fits all explanation, even in a single partitioning of the cases, would seem too idealized. It seems right that not all eschewing cases, for example, will call for the same objectifying analysis. The enriched objectification account thus seems apt to describe the complexities of the epistemic dealings we’re most interested in. One wonders, though, what this implies for the experience of different kinds of testimonial injustices. For cases that are properly described as involving multiple forms of objectification—we might call these cases of compound-objectification—is the objectifying harm different, or more significant? In any case, this result seems like a real payoff for the view.
The second part of Aidan’s argument is to show us why it is not the case that the othering account fairs just as well. He focuses first on Pohlhaus’s account in terms of derivatization. The basic idea of derivatization is that one epistemically others a testifier by failing to recognize them as an agent in their own right, instead interpreting them, as it were, derivatively; through the lens of some other person or entity. In this way the testifier is not able to project themselves, instead reflecting whomever they are interpreted through, or in terms of.
Aidan in fact takes on four arguments that Pohlhaus gives for the derivatization account, but for the sake of space and time I’ll focus on just one before turning to Aidan’s extension of the objectification account, in the final section.
Pohlhaus contends that an analysis of the harm of testimonial injustice in terms of objectification leads us to focus on the role of the individual to the exclusion of (or, at least, with an underemphasis on) the social, or interpersonal dynamic of the phenomenon. But, the interpersonal nature of our epistemic lives is crucial to our ability to flourish epistemically. For this reason, it would seem, Pohlhaus will think that any explanation of the primary harm of testimonial injustice that fails to countenance the sociality of our epistemic lives, will miss out on some crucial element of the ways in which we can be harmed epistemically.
In response, Aidan objects that Pohlhaus has not motivated the central premise, that the objectification account leaves us “stuck at the agential level” and therefore without an explanation of the interpersonal harms that accrue to victims of testimonial injustice. You can agree (as Aidan does) with Pohlhaus about the important connection between sociality and epistemic health, but why think that the objectification account is silent on this? I share Aidan’s sense that, absent some further argument, it is not clear why we should think that the objectification account suffers this shortcoming. I also share Pohlhaus’s sense that it is worth pausing to meditate on how, precisely, to countenance the way that sociality figures in our thinking about the harms of testimonial injustice (and epistemic misconduct, more generally). Let me try to say a bit about what I think Pohlhaus might be onto here.
Pohlhaus characterizes epistemic agency as ‘one’s ability to pursue epistemic projects’. When she’s arguing that the primary harm of testimonial injustice needs to include some element of our interpersonal lives, our sociality, she invokes a familiar example from The Talented Mr. Ripley. Here’s the relevant bit from Aidan’s paper:
Pohlhaus is surely right to draw attention to the ways in which testimonial injustice can lead to obstacles to one’s ability to pursue epistemic projects that require cooperation from those we’re trying to share our knowledge with. In Pohlhaus’s example (2014: 110–1), Marge Sherwood is unable to pursue various lines of inquiry concerning her fiancé’s disappearance due to Herbert Greenleaf dismissing her well-founded suspicions concerning Ripley, rather than finding ways to help her discover more. (9, emphasis added)
I think this example, or this gloss on the example, undersells the important thing that Pohlhaus is drilling down on. It suggests that what is lost—the epistemic good at issue in the distinctively interpersonal dimension of the harm of testimonial injustice—is something like, other people helping you get evidence, or helping you to settle a question. I think it’s deeper than that. I mean, it is that, but it is more than that.
Refusing to collaborate, cooperate, etc. with someone not only frustrates their current effort to participate in the knowledge economy, but it can also function as an erosion of their epistemic agency, more globally. Marge could say to hell with Greenleaf and continue on in her epistemic project, unassisted. If that aspect of epistemic collaboration were all that were lost in cases of testimonial injustice, it wouldn’t be especially interesting, or seem especially relevant to describing the primary harm. But Marge isn’t just looking for extra epistemic labour, she needs something subtler: she needs not to doubt herself, not to feel like she’s losing her grip, out to sea. Note that Greenleaf could have agreed to help Marge search for evidence while adopting a mocking attitude about their mutual epistemic project. It is not hard to imagine Greenleaf taking up the project with her in a way that still gives rise to the epistemic harm attendant to the original case. Conversely, he could have declined to take up the project without harming her epistemically—all tied up with my own business, but good luck finding Dickie! In that case, he would be declining to contribute his epistemic labour, and yet we wouldn’t expect to see the epistemic harms that emerge in the original case: the sense of helplessness and bewilderment, a kind of alienation from one’s own epistemic life.
Those deeper, subtler ways in which our epistemic health depends on our sociality include a broad, and important suite of attitudes and interactions that signal to us in ways big and small that we are navigating the world properly, learning the right things, interpreting them appropriately, exchanging information of all kinds, and so forth. So, I think it’s a mistake to relegate the role that sociality and interconnectedness play in our epistemic health to a shared division of epistemic labour, at least so construed.
In any case, and for all that is deep and interesting about locating the connection between our sociality and our epistemic health in the primary harm of testimonial injustice (and, as I say, in other forms of epistemic misconduct), it is still unclear why the objectification account is necessarily silent on this. And, in fact, I think we see quite clearly that the objectification account can capture that connection vividly in the final section of the paper, in Aidan’s discussion of violation as a form of objectification. I’ll turn to this now.
In the final sections of the paper, Aidan wants to make the case that the objectification account can do more than withstand the objections leveled against it; it can also do some nifty stuff that the othering account cannot. He makes two arguments here: 1) that the objectification account has a cleaner story to tell about Excess cases, specifically in terms of fungibility, and 2) that the enriched objectification account offers explanatory resources that fully outstrip the othering account. Again, for space and time, I’ll focus my comments here on the latter.
Among the modes of objectification identified by Nussbaum is violatability, which involves regarding something as ‘permissible to break up, smash, break into’. (12, quoting Nussbaum) Analogized to the epistemic, Aidan suggests:
One way to treat someone as lacking boundary-integrity, in their role as an epistemic subject, is to act towards them as if it’s permissible to interfere relatively directly with their belief about the world, without their consent. (13)
He ‘speculatively suggests’ that gaslighting might be an example of this sort of thing. Citing Abramson’s phenomenal 2014 paper, he highlights her discussion of gaslighting as a strategy that aims at more than just getting someone to regard themselves as mistaken, or in error, but quite fundamentally out of depth. As Abramson tells us, gaslighting involves,
charging someone not simply with being wrong or mistaken, but being in no condition to judge whether she is wrong or mistaken. The accusations are about the target’s basic rational competence—her ability to get facts right, to deliberate, her basic evaluative competencies and ability to react appropriately: her independent standing as a deliberator and moral agent.” (Abramson, 2014: 8, bolded emphasis added)
Aidan’s point here is that cases of testimonial injustice that involve gaslighting (as some, but not all, do) meet the epistemic standard of violatability, involving the transgression of epistemic boundaries in the way that can ‘break up’ or ‘smash’ someone’s epistemic life. Given the way in which being gaslit can leave an individual estranged from even her most basic epistemic competencies, Aidan suggests that it can plausibly be described as a form of objectification by violation.
Where this explanation is straightforward for the enriched objectification account, it is less clear how the othering account can capture this phenomenon. The revised derivatization account reproduced from Emmalon Davis (p. 12) describes othering as a matter of prejudicially assessing the epistemic competencies of speakers in a manner that fails to appropriately countenance her subjectivity. Gaslighting cases, however, are not plausibly a matter of a prejudicial assessment of a target’s epistemic competencies. Indeed, the effectiveness of a gaslighting campaign will depend in part on a thoroughly accurate assessment of those competencies.
The central goal of Aidan’s paper is to defend a particular account of the primary harm of testimonial injustice, so it is worth noting, in closing, that his discussion of gaslighting comes in the service of that project. That is, the argument here is not that the primary epistemic harm of gaslighting is objectification (even, specifically, objectification viz. violatability). It is rather that some cases of gaslighting involve testimonial injustice, and that the primary harm of testimonial injustice is objectification, including, in some cases, violability. This leaves open that there may be other, perhaps more central, epistemic harms at work in gaslighting. This, too, is a virtue of Aidan’s argument here. As he notes, Abramson highlights the special significance of the particular mechanisms of manipulation in gaslighting, including their reliance on the particular social norms that so often animate the relationships in which gaslighting takes place. For example, norms of love, intimacy, dependence, expertise, etc. Veronica Ivy has likewise theorized the special role that norms of allyship play in certain other examples of gaslighting. Plausibly, then, these crucial aspects of the particular kind of manipulation at work in gaslighting will redound to the epistemic harms at its center.
Aidan’s paper marks a significant intervention in this literature. Social and feminist epistemology has much to glean from this paper in particular, and from Aidan’s broader project of theorizing epistemic objectification, more generally.
Thanks so much to Aidan and to the blog for inviting me to engage with such a rich and exciting new paper! This has been a ton of fun and super edifying, and I’m looking forward to the discussion!