9
Epistemic Injustice and the Preservation of
Ignorance
A1iranda Fn :ckir
Ignorance is not always bad; far from it . Looking at the issue in its most
general aspect there is the obviou s point that for finite beings, massive
ignorance is a precondition ofha\·ing an epistemically functional lifr, for
cognitive overload is an epistemic liabilin ·. Then.: is an imkfinite , indeed
infinite , number of things that we d o not have the slightest need to
know · the number of hairs on your h..:ad at midnight on your next
birthday , for instance. Furthermore, we activdy need 1101 to know most
of them (or not to spend time and energy investigating them ) in order to
conserve cognitive capacity for tho se things that we do need to know.
Less abstractly, there is also the point that then.: are many things it would
be morally and /o r prudentially bad to kn cm· - intimatt: details that are
none of our business; techniques of criminality; methods of rekindling
old ethnic hatreds in a population. These points are familiar from
debate s about 'the value question ' in relation to knowledge . l
Furthermore, as Cynthia Townle y ha s argued, many forms of epistemic
cooperation and many of the dispositions involved in epistemic virtues
generally depend crucially upon our leaving some useless or harmful
things unknown by passively or actively preserving others' ignorance of
things they need not or should not know (Townley 2011). In short, good
epistemic practice is necessarily highly selective in all sorts of ways.
What matters is that we know what we need to know, expanding outwards to the broader aim of knowing and telling what we should know
and tell, given our purposes and broadly ethical obligations all things
considered. Good epistemic conduct needs to be understood as the
maintenance of appropriate balances of knowledge and ignorance, in
oneself and also in relation to others.
This opening reflection on the epistemic value of ignorance and its
place in the epistemic economy directs our attention to the basic norma-
1
Sec Sosa's example of 'trivial' knowledgt." (2002, p. 156); and Zagzebski 's examples of
prudentially and morally 'bad' knowledge (2 003, p . 21).
lbO
Epistemic Injustice and the Preservation of Ignorance
161
tive ambivalence in our use of the term. 2 'Ignorance' may refer simply to
an epistemically innocent absence of knowledge (this absence being
advantageous or disadvantageous, as the case may be, without any reflection on the conduct of the epistemic subject in question); or alternatively
it may refer to some kind of cognitive failure, which might be nonculpable (perhaps the result of misleading evidence) or which might, on
the other hand, represent a blameworthy failure to put the requisite effort
or skill into knowing something one ought to know.
This paper will focus on those forms of culpable and non-culpable
ignorance that are created or preserved by one or another kind of epistemic injustice that I have elsewhere labelled 'testimonial injustice' and
'hermeneutical injustice'. 1 I shall discuss the first only briefly, for it is
more specifically in relation to hermeneutical injustice that new and
complex issues have recently been raised concerning various different
forms of ignorance that can be involved in this phenomenon.
In particular I hope to say something useful about the place of 'wilful' 1
or motivated ignorance, and to thereby contribute to recent debates in
which the phenomenon of hermeneutical injustice has been related to
what Charles Mills has termed 'white ignorance'. 5 Ultimately I sha 11
argue that the phenomenon 1\1ills characterises on the whole picks out
a different kind of ignorance from any that is involved in hermeneutical
injustice. But I shall also argue that the two categories can overlap.
1
Preserving Patterns of Social Ignorance: Testimonial
Injustice and Hermeneutical Marginalisation
When the level of credibility attributed to a speaker's word is reduced by
prejudice operative in the hearer's judgement, the speaker ,uffers
a testimonial injustice. Despite the specific label, the speech act in which
his word is expressed need not be strictly that of testimony or telling, but
might equally be the airing of an opinion, suggestion or relevant possibility. Furthermore, as Christopher Hookway has suggested, it might even
be occasioned by the asking of a question that is designed to contribute to
some shared inquiry. 6 The prejudice driving any case of testimonial
injustice may or may not be a belief, and it operates specifically in the
2
For debate about what ignorance is necessarily ignorance of, see for imtanct' the t·xchang,·
between Pierre Le Morven (2010, 2011) and Rik Peels (201 la).
'See Fricker (2007, 2013b).
1
The term 'willful ignorance' is from Gaile Pohlhaus (2012).
' See Mills 12007, 2015'
'' I am grateful ro Chris Hookway for this point that someone who puts a quesr1,m a,
a contribution to collective inquiry (perhaps in the classroom) might find her ,Juestinn
passed over due to prejudice (Hookway 20 I 0). I hope I may ultimately he allcw,ed rhi, as
a limiting case of testimonial injustice, even though it concerns a speech acr rhar i, nnr 111
162
Miranda Fricker
hearer's judgement of crt1dibility, where the judgement may be unreflective
and spontaneous - a matter of ingrained habit. (The trained quasiperceptual dispositions governing such judgements I have elsewhere
labelled the hearer's 'testimonial sensibility '.) The influence of prejudice
in judgements of credibility can make itself felt regardless of the hearer's
beliefs, indeed in spite of them, for prejudice can operate unconsciously
7
or, as we have now learned to sav, at the level of 'implicit bias'.
Testimonial injustice's obvious connection to ignorance is that in cases
where the speaker knows thatp and the prejudice operative in the hearer's
credibility judgement prevents her learning that p from the speaker, other
things equal she thereby stays ignorant of p .
Testimonial injustice not only blocks the flow of knowledge, it also
blocks the flow of evidence, doubts, critical ideas and other epistemic
inputs that are conducive to knowledge . The free circulation of these
epistemic goods is conducive to knowledge not only in the direct sense
that ready-made items of knowledge may themselves be transmitted, but
also in the indirect sense that such items tend also to constitute reasons to
believe other things, so that they may have the epistemic power to convert
other of the hearers beliefs into knowledge. The obstructions that testimonial injustice introduces into the circulation of such epistemic items is
therefore not only bad for the person whose word is prejudicially downgraded; it is epistemically bad for the hearer, and for the epistemic system
quite generally . An epistemic system characterised by testimonial injustice is a system in which ignorance will repeatedly prevail over potentially
shared knowledge, despite speakers' best efforts. \Vhere a speaker knows
something the hearer does not (and where the level of credibility deficit is
such that the hearer does not accept what she is told) the hearer's ignorance is conserved. Alternatively, where the speaker is offering evidence
with a (positive or negative) bearing on something the hearer already
7
assertion . The label ' 1es1imonial injustice' was always explicitly in1ended capaciousl y, to
include not only the broad class of tellings bu1 also cases where a speaker 'expresses
a personal opinion to a hearer, or airs a value judgemenI, or 1ries out a new idea or
hypothesis on a given audience' (Fricker 2007, p . 60 ). Hookway's case of the studeni's
relevant question admittedly stretches my charac1erisation; but provided one can regard
the asking of such a question as potentially vulnerable to a prejudicial credibility dejicli, then
it seems more or less to belong to the same category. I would certainly acknowledge that
this requires us to take 'credibility' in its everyday sense, as covering the wide range of
respects in which what someone says may be taken as more or less authoritative . Such
a colloquial construal is supported by the fact that the object of any credibility judgement
includes nor only whai is said but also 1he speaker. Ai any rate, I hope these considerations
provide enough commonality co keep 1he diverse possibili1ies sufficiently unified under the
care gory '1estimonial injustice'.
There is a fast growing philosophical literature drawing upon empirical work in psychology on implici1 bias . See, for inscance, Holroyd (2012), Saul (2013), Gendler (2014),
Nagel (2014) , Leslie (forthcoming) and Browns1ein and Saul (eds .) (2016) .
Er1stemic Injustice and the Preservation of Ignorance
163
believes but does not know, then the hearer misses out on reasons which
(if positive ) might render her belief knowledge or at least lend it greater
justificatory weight; or which (if negative) might disabuse her of a false
belief, or at least reveal it as less well supported than it had seemed. Either
way, an opp ortunity for epistemic improvement is lost, and ignorance
prevails .
A further, more buried, form of epistemic damage caused by testimonial injustice is that, where it is persistent and socially patterned (as
anything driven by prejudice is likely to be), it will tend to create or
increase hen11e11cut1cal
marg-inalisacion. That is to say, it will tend to create
and sustain a situation in which some social groups have less than a fair
crack at contributing to the shared pool of concepts and interpretive
tropes that we use to make generally shareable sense of our social experiences . ·we might gloss this idea of a pool of concepts and interpretive
tropes as 'shared social meanings', where the idea is that while this pool
will surely not exhaust all the various up and running sets of social meanings that are being used locally by this or that group in a given society, the
shared pool (elsewhere I have called this the 'collective hermeneutical
resource ') contains only meanings that just about anyone can draw
upon and expect those meanings to be understood across social space
by just about anyone else. The collective hermeneutical resource contains
those concepts and conceptualisations that are held in common.
This means that being a member of a social group that does not
contribute on an equal footing with other groups to that shared interpretive resource (a position ofhermeneutical marginalisation) puts one at
an unfairly increased risk of having social experiences that one needs,
perhaps urgently, to understand and/or communicate to certain powerful
social others - to a teacher, an employer, a police officer, a jury - but
which cannot be made mutual sense of in the shared terms available .
We are only now, for instance, entering a historical moment in the West at
which it is increasingly possible for a young person originally assigned as
'male' to be able to say to a parent , teacher or friend that he has always felt
himself to be a girl in 'the wrong body ' and hope to be understood as
expressing an intelligible experience . Increasingly the various concepts
and conceptions of how sex, gender, sexual orientation and other deep
identity affiliations may be organised and reorganised in an individual's
experienced identity - notably the concept of trans together with its less
established counterpart cis - are gradually entering the shared henneneutical resource instead of staying local to the trans community . Still now,
where a trans woman might attempt to describe her experience of gender
identity to a social other who does not share the relevant concepts, she is
unlikely to be able to make herself much understood, and this is where her
remaining hermeneutical marginalisation will manifest itself in the unfair
deficit of intelligibilitv that constitutes a hermeneutical
injustice. Like
te.,timonial injustices , thi-, kind nf hermt?neuric::il injustice preserves
ignorance, for that which remains insufficient!\ · intclligihle w the relevant
social other cannot be passed on tn them as knowledge
Here we see how closelv the t\vo kinds of epistt:mic injustice are related :
testimonial injustice can create or sustain hem1t.:neutical marginalisation
by blocking the flow of reports, ideas and perspective-, that would help
generate richer and more diversified shared hcrmeneutical resources that
all can draw on in their social understandings. whether of their own or of
c,thers' experiences. Therefore the broad patterns nf testimonial injustice - most likely patterns created by the operation of negative identity
prejudices, inasmuch as these are the chief systematic prejudices · will
tend to reproduce themselves as pattern,; of henneneutical
marginalisation, and it is these that give rise rn systematic hermeneutical iniusrices.
Thus we can see how the preservation of hearer-ignorance
that is the
likely effect of any instance of testimonial injustice can contribute directly
to the hermeneutically
marginalised po
Purchase answer to see full
attachment