Contribution
The paper that introduces testimonial injustice. It comes to the concept from the epistemology of testimony. A hearer’s spontaneous critical reception of testimony is explained by a socially trained sensibility, on the model of an ethical sensibility. Since that sensibility must use social stereotypes, prejudice can enter it, and prejudice is the source of the injustice. It then names the virtue that corrects it, and argues that the virtue’s availability is historically contingent.
Key insight
Credibility judgements are perceptual and socially trained, not inferred. That is why prejudice in them is so hard to see and so ordinary. A hearer can wrong a speaker while sincerely trying to judge well, and whether he is to blame depends on whether the concepts he would need to catch himself exist yet.
Assumptions
- Testimonial knowledge needs critical reception. The uncritical model is “justificationally lax”, and the inferential model is “intellectualism in a tight corner” (pp. 2–3).
- McDowell’s “doxastic responsibility”, a responsiveness to reasons that need not be inference, is taken as the target to explain (pp. 4–6).
- Zagzebski’s definition of virtue: a motivational component plus reliable success (p. 21).
- No exact credibility ratio. “There is no such precise probability available at all” (n. 11).
Key results
- Testimonial sensibility (pp. 7–12). Ethical sensibility is a rational capacity that “comprises virtues; which is inculcated in the subject through a process of socialization; and which permits of ongoing correction”. Its testimonial analogue is “learned and internalised through a process of epistemic socialization”. It is informed by “common cultural stereotypes of intellectual authority”, which are “positively needed to oil the mechanism” and “fertile ground for prejudice”. Where stereotypes are “empirically unfounded”, the sensibility “will be both epistemically and ethically defective”.
- Testimonial injustice (introduction, pp. 12–13). “A speaker receives the wrong degree of credibility from his hearer owing to a certain sort of unintended prejudice on the hearer’s part.” Prejudice can produce “credibility excess” or “credibility deficit”. More usually it “will surreptitiously raise or lower the hearer’s degree of belief”, rather than flip belief.
- The cases (pp. 14–18). Tom Robinson’s jury “really do find him guilty”, so their failure is “at once … a gross epistemic failure and an appalling ethical failure”, and it is culpable. Greenleaf’s failure with Marge (“there’s female intuition, and then there are facts”) is offered as plausibly non-culpable. Ripley “cynically exploits the gender attitudes of the day”.
- Reflexivity (pp. 18–19). The hearers fail to adjust for the difference made “not only that he is black but equally that they are white”. The virtue is “Reflexive Critical Openness”: the hearer “reliably succeeds in correcting for the way testimonial performance can be prejudiced by the inter-relation of the hearer’s social identity and the speaker’s social identity”.
- Historicism (pp. 20–22). “A setting in which there is little critical awareness of gender is a setting in which no one is in a position to possess the virtue.” Greenleaf was “always at fault” but “not culpably at fault until the requisite critical consciousness of gender became available to them”. The argument goes through Williams’s “sound deliberative route”. The motive to truth is present, but the route is missing because “the critical concepts … are genuinely not yet available”.
- Oppression (p. 23). “Where it is systematic then we should recognize it as a face of oppression.” The virtue’s “anti-oppressive power remains hostage to the broader social structures”.
Claims
| id | claim | strength | support |
|---|---|---|---|
| C1 | Testimonial knowledge is gained through a socially trained sensibility, critical but non-inferential | moderate | argued as the exit from an impasse, §§1–3 |
| C2 | Testimonial injustice is a credibility misjudgement caused by the hearer’s unintended prejudice | moderate | defined, with two literary cases, §4 |
| C3 | The corrective virtue is reflexive: it corrects for the relation between hearer’s and speaker’s identities | moderate | argued from the cases, §4 |
| C4 | A hearer cannot be culpable for lacking the virtue before the needed critical concepts exist | moderate | argued via Williams’s internal reasons, §5 |
| C5 | Systematic testimonial injustice is a form of epistemic oppression | weak | stated, with Bartky and Fanon, §5 |
Concepts
- Testimonial sensibility. The socially trained capacity that governs spontaneous, critical uptake of others’ word.
- Credibility excess and deficit. More or less credibility than the speaker rationally deserves.
- Testimonial injustice. The wrong degree of credibility owing to the hearer’s unintended prejudice. The paper treats deficit.
- Reflexive Critical Openness. The virtue that corrects for prejudice arising from the relation between the two parties’ social identities. In the book it becomes “testimonial justice”.
Connections
The paper responds to McDowell’s “Knowledge by Hearsay”, Coady’s Testimony and Elizabeth Fricker’s anti-reductionism debate. It draws on MacIntyre, Zagzebski, Hookway, Karen Jones (“distrust their own distrust”) and Williams. Its two cases are reused in the book and in the 2017 chapter (NOTE-853), where Ripley’s part is reclassified as “intentional gaslighting”.
Bearing on the record
- THEORY-317. This paper confirms its testimonial definition, “prejudice on the hearer’s part”, and adds “unintended”. It speaks of stereotypes of social types and of social identity, but does not use “identity prejudice”. It also allows credibility excess as prejudicial dysfunction. The précis account is silent on that, and does not need it.
- QUESTION-072, part 1. The paper’s model of credibility as a trained sensibility fits the question’s case. Whoever files is the hearer, and the filing habit is a sensibility shaped by what has been filed before. Testimonial injustice then requires two things: that the habit rest on an unfounded stereotype about whose moves are worth filing, and that it cause a deficit for a speaker. The historicism result bears on blame, not on whether the wrong occurs. A filer can do the injustice before anyone “is in a position to know better”. THEORY-327 states the condition.
- Nothing here is ML practice.
Limitations
- The cases are literary, and the culpability verdicts are offered as interpretations (“on this (perhaps rather charitable) interpretation”).
- Credibility excess is named and set aside.
- The paper is the earliest of the three and predates the book’s terminology, so its definitions are the first versions.
Open questions
- Whether a sensibility with no prejudice can still produce systematic deficits by statistical stereotype that is empirically “founded” but self-fulfilling. The paper counts founded stereotypes as legitimate.