Epistemic Injustice

In The Talented Mr Ripley, Marge Sherwood tells Herbert Greenleaf that she believes Tom Ripley murdered his son. She is right. Greenleaf dismisses her with a line that Miranda Fricker made one of the central cases of her 2007 book: there is female intuition, and then there are facts. Marge's evidence is good. Her reasoning is sound. She is not believed, and the reason she is not believed has nothing to do with either.

Something has gone wrong here that is more than a failure of prudence on Greenleaf's part. He has made an error, certainly, since he has failed to acquire a true belief that was on offer. But he has also wronged Marge, and wronged her in a particular capacity. Fricker's claim is that this wrong is distinctively epistemic, that it has a structure worth analysing, and that the analysis requires epistemology to stop treating knowers as interchangeable.

The social turn

The epistemology of the preceding articles is largely individualist. The knower is a single mind confronting evidence, and other people appear only as sources of testimony whose reliability is to be assessed like that of any instrument. Social epistemology does not reject that picture so much as complete it. Testimony was the first place where it became clear that the individualist picture was inadequate, since so much knowledge is irreducibly collaborative. Disagreement was the second. This article is the third, and it adds something neither of the others contained: the thought that the assessment of other people as sources is itself an activity that can be done well or badly, justly or unjustly.

Alvin Goldman's version of social epistemology is veritistic, concerned with which social practices tend to produce true belief. Fricker's is ethical as well as epistemic, concerned with what the participants in those practices owe one another. The two are compatible, and the cases Fricker discusses are ones in which both kinds of failure occur together.

Two kinds of wrong

Testimonial injustice

The structure has three parts, and each matters.

There is a credibility deficit. The speaker is believed less than they should be. Innocent credibility deficits are everywhere, since hearers often have no way to tell how reliable a stranger is, and the deficit alone is not an injustice.

The deficit is caused by prejudice. The hearer's estimate of the speaker's reliability is distorted by an association between the speaker's social identity and incompetence or insincerity, an association that is resistant to counter-evidence in a way that marks it as prejudice rather than error.

The prejudice is identity-based and systematic. It is not an idiosyncratic dislike but a pattern that follows members of a group across different contexts: the courtroom, the clinic, the classroom, the workplace. Fricker's second central case is Tom Robinson in To Kill a Mockingbird, whose testimony is disbelieved by a jury because he is Black, in circumstances where the evidence clearly supports him.

What is the wrong, exactly? Fricker argues that it is a harm to the speaker in their capacity as a giver of knowledge, and that capacity is essential to being treated as a rational agent. To be systematically disbelieved is to be excluded from one of the basic relations of human life, the sharing of what one knows, and to have one's standing as a reasoner denied.

Hermeneutical injustice

The second form is subtler and in some ways more troubling, because no individual need be at fault.

Fricker's central example is the experience of sexual harassment before the concept had a name. Carmita Wood, whose case helped bring the term into use in the 1970s, left a job after persistent unwanted advances from a supervisor. When she applied for unemployment benefit, she had no way to describe the reason she left that the system would recognise. The experience was real, it was harmful, and she lacked the words to render it intelligible, not because of any failure on her part but because the collective stock of concepts did not contain the one she needed.

What makes this an injustice rather than a misfortune is the explanation of the gap. The concept was missing because the people who had the experience were systematically underrepresented in the institutions where shared meanings get made, in law, in medicine, in journalism, in the academy. The gap was not random. It fell exactly where the experiences of a marginalised group were concentrated.

Once the term existed, something changed that had not merely to do with vocabulary. Women who encountered it recognised their own experiences, found they were not alone, and acquired the ability to make a claim that others could understand. The concept did not just describe their situation; it made a range of their experiences available to them as knowledge.

Extending the framework

The literature since 2007 has expanded Fricker's two categories considerably.

Kristie Dotson identified testimonial smothering, in which speakers truncate their own testimony because they anticipate that it will not be received well, and contributory injustice, in which a marginalised group has developed its own interpretive resources and dominant hearers refuse to learn them. The second extends hermeneutical injustice: the gap is not in the concepts available but in the willingness of the powerful to use the concepts the marginalised have already produced.

José Medina has argued that credibility excesses, receiving more credibility than one's evidence warrants, are also part of the problem, since credibility is comparative and an excess for some is structurally connected to a deficit for others. An expert witness whose authority is inflated because of their social standing makes it harder for a less credentialled witness to be heard.

Work in medicine has documented patterns in which patients' reports of their own symptoms are discounted, and the literature on epistemic injustice has become a tool for analysing diagnostic delay and the treatment of chronic pain.

Where the intermediate tier leaves things

Twelve articles have taken the beginner tier's vocabulary and put it under pressure.

The analysis of knowledge failed under Gettier's counterexamples, and the repairs showed a pattern: each identified a real defect and each was met by a case that exhibited the failure without the defect. The structural theories of justification were set against the regress and none prevailed. The internalism dispute revealed two jobs the concept of justification was being asked to do. Reliabilism and virtue epistemology offered externalist accounts with their own hard problems, and the modal conditions of safety and sensitivity turned the anti-luck intuition into something testable. Closure turned out to be the fulcrum on which scepticism turns. The a priori survived Quine in modest form. Induction remained unjustified. Bayesian methods replaced all-or-nothing belief with degrees and made everything more precise without settling the old questions. Finally, testimony, disagreement, and injustice showed that the individual knower of the tradition was always embedded in a social world, and that the embedding has norms.

The advanced tier takes these threads into current research. Contextualism takes up where closure left off. Knowledge-first epistemology takes the failure of analysis as data rather than defeat. Disjunctivism attacks the internalist and externalist dilemma directly. The formal articles push the Bayesian framework to its limits, and the last articles return to disagreement as a problem about higher-order evidence.

Next

The advanced tier begins with contextualism, which returns to closure and scepticism with a claim about the semantics of the word "know" itself.