bs_bs_banner

Categorical Injustice

Ásta Sveinsdóttir

Introduction

Some of the most exciting research in philosophy in the last couple of de-
cades has been work at the intersection of theoretical and practical philosophy
from a feminist perspective. The work I present here is in that vein, as it lies
at the intersection of metaphysics, feminist philosophy, and social philosophy.
In my recent Categories We Live By (Ásta 2018), I presented an account of the
construction of social categories of individuals. These include sexes, genders,
and races, but also any other category defined by a social property, such as ref-
ugees and single mothers. In this essay, I will address a specific way in which
such social construction can involve an injustice that is distinctly metaphysical.
I name this sort of injustice “categorical injustice” and it occurs when agents are
systematically thwarted in their attempts at performing actions by how they are
socially constructed.

The Conferralist Account of Social Categories

The central idea in my theory of social categories is that individuals get con-
ferred onto them a social status in contexts and this status consists in constraints
and enablements on their behavior. These constraints and enablements are what
results in categorial injustice, given certain conditions. I offer a theory of how
these constraints and enablements are put in place and the mechanism by which
they result in an injustice. But this is looking a bit ahead. Let us get a sense of the
conferralist framework first.
We all have various features, and some of these features have social signif-
icance in the contexts we travel. The features we have vary. Some of them are
physical features, others relational, some even themselves social. For example, I
am 168 cm tall, have moss green eyes, breasts, and broad shoulders. I am also the
daughter of Þóra Kristjánsdóttir and Sveinn Einarsson, an Icelander, and a native
speaker of Icelandic. Some features only have social significance in very specific
contexts (having a short pinky finger, perhaps), others matter in most contexts
(sex assignment, perhaps). The key question is: what is it for a feature to have
social significance in a context? The intuitive idea behind the theory I offer is that

JOURNAL of SOCIAL PHILOSOPHY, Vol. 0 No. 0, Spring 2019, 1–15.

DOI: 10.1111/josp.12318
© 2019 Wiley Periodicals, Inc.
2  Ásta

a feature has social significance or meaning if people treat you differently when
they take you to have the feature. I make this a bit more precise by saying that a
feature has social significance in a context in which a person gets conferred onto
them a social status if they are taken to possess the feature. So, for example, being
female is socially significant in a context in which people taken to be female get
conferred onto them a social status. The person may in fact not have the feature,
they only need to be taken to have it. The social status consists in constraints and
enablements on the individual’s actions in the context, which means that certain
actions are available to them and others not, as compared to if they were not taken
to have the feature. For example, if you are taken to be female at a summer party
you may not be invited to operate the grill.
I make a distinction between two sorts of categories, two sorts of contexts,
and two sorts of features: institutional and what I call “communal.” Institutional
contexts are contexts governed by a set of rules or laws—for example laws
governing driving in California or rules governing the activities of the Dolphin
Swimming and Rowing Club in San Francisco. The entities that confer status on
a person in those institutional contexts have the requisite institutional authority to
confer the status in question. For example, a license to drive in California is con-
ferred by an official of the Department of Motor Vehicles upon judging the per-
son to meet the relevant prerequisites. A person is conferred the status of Member
of the Dolphin Club by the President of the club upon being taken to meet the
membership requirements and having been elected by the members present at the
monthly members’ meeting. People have the status as long as they remain in the
context and it does not get revoked.1 
The profile of institutional properties is like this:

Conferred property: P
• 
Who: a person or entity or group in institutional authority
• 
What: their explicit conferral by means of a speech act or other public act
• 
When: under the appropriate circumstances (in the presence of witnesses, at a
• 
particular place, etc)
Base property: the property the authorities are attempting to track in the confer-
• 
ral. The individual need not have the property; they just need to be taken to
have it.

I call the conferring of an institutional status on an individual an act of clas-

sifying that individual. Classifying an individual is an act of placing them in an
institutional order.
The other sort of feature, the communal, is of more interest to the project
of this paper, although institutional and communal conferrals can interact in a
variety of ways.
A communal feature is also a social status conferred upon a person in a con-
text on the basis of them being taken to have a feature that is socially salient in
the context. The status consists in constraints and enablements, just like in the
 Categorical Injustice   3

institutional contexts, but these constraints are not deontic. They do not concern
rights, privileges, and obligations, but power, sway, and nondeontic restrictions.
For example, being tall may be a socially important property in a context, and
people taken to be tall deferred to in making decisions for people in the context
or allowed to speak more than others. Being knowledgeable (or not) about a par-
ticular subject matter may also be a socially significant property such that if you
are taken not to be knowledgeable no one will listen to what you have to say. The
base property for the conferral of a communal property can vary, as in the insti-
tutional case. Consider, for example, a party where people are interested in who
is and is not married, which makes that a salient property in the context. Perhaps
they are looking for someone to marry and they do not want to get entangled with
people who already are (in a different context that may be precisely what they
want). Alex is legally married, but he does not carry a wedding ring and acts as
if he is not. At the party, Alex gets conferred upon him the status eligible by the
other people at the party. This is so even though he is already married. At the
party, he has the communal status of eligible and various people flirt with him be-
cause he has that status in the context. Let us suppose now, that a friend of Alex’s,
Sami, shows up at the party and is enraged at Alex’s acting as if he is not legally
married. Sami tells some people at the party that Alex is already married and after
many a whisper Alex loses his communal status as eligible in the context. People
may even feel hurt or betrayed by his acting as if he were not legally married.
The profile for a communal property is thus the following:

Conferred property: P
• 
Who: a person or entity or group with standing
• 
What: their conferral, explicit or implicit, by means of attitudes and behavior
• 
When: in a particular context
• 
Base property: the property the conferrers are attempting to track in the confer-
• 
ral, consciously or unconsciously. The individual need not have the property;
they just need to be taken to have it.

The conferral of a status in the communal case is an act of placing a person

in a context-specific communal order. It is important to note that what we have
here are two distinct properties: being legally married and being presumed to
be legally married and the corresponding being legally not married and being
presumed to be legally not married, which we have labeled “eligible.” Being le-
gally married and being legally not married are institutional properties and Alex
has the former if he has had that status conferred upon him at a prior date. Being
presumed to be legally married and being presumed to be not legally married are
communal features.2  They are social statuses that get conferred upon Alex on the
basis of his being taken to have the relevant institutional statuses. They consist
in constraints and enablements on his behavior and with them come norms for
behavior for Alex and for others engaging with him. For example, it would not
be appropriate for Alex to be flirting outrageously with people if he is presumed
4  Ásta

to be married; similarly would it not be appropriate for others to flirt excessively

with him.3 
Let us now turn to the content of the constraints and enablements and from
where they derive.

Placement in Categories

In the case of the institutional placement of individuals, which I have termed

“classification,” the constraints and enablements are deontic. These are rights and
privileges, duties, and so on that come with the particular roles assigned and are
encoded in laws, rules, or regulations. Their particular content was agreed upon
by the relevant parties when the laws, rules, or regulations were instituted. The
story in the communal case is more complicated. Here, there are no laws or rules
that describe the constraints and enablements that come with the assigned status
or role. Where does the content of these constraints and enablements come from?
I consider them to have ideological sources and our habits, practices, as well
as narratives and the like contribute to them. I mean to use the notion of ideol-
ogy in a neutral way, as not inherently bad, although it often is (Geuss 1981).
The epistemic components of an ideology are especially important here as
they include stories, concepts, narratives, and assumptions that we use to make
sense of the world around us, and often to justify why the world is as it appears.
Generalizations about people and things play a large role here, including stereo-
types and prejudice. What is of particular interest to us here is assumptions about
groups of people or stereotypes about groups. These assumptions are part of the
ideologies we live with and set the intelligibility conditions for the actions of the
people we interact with.
I am going to follow Erin Beeghly (2015) in thinking of stereotypes and
stereotyping in a descriptive way. Then a stereotype is a set of features associ-
ated with a group and stereotyping is attributing one or more of the stereotypical
features to a person who you take to be a member of the group.4  There are other
uses of stereotypes and stereotyping, but this one is the one which is most relevant
to the construction of social categories. It is important, however, that the act of
stereotyping, as I am using it, need not be conscious. In the cases I am interested
in it is not a conscious act of judging a person to have certain features on the basis
of judging them to belong to some group.
The social map in the particular context consists in several parts: what roles
there are to play in the context,5  who plays what role, and what are the expecta-
tions of each role. What roles there are to play is partly determined by the material
and institutional circumstances. For instance, a car is stuck in the snow in the
Sierras and the people want to get it unstuck and get back on their way. Given
the material and technological circumstances,6  there are two roles to play: driver
and pusher. But who is to play what role? That can be partly determined by in-
stitutional and material factors, too, but stereotypes can play a role there also, as
who is seen to “fit” each role will be partly a function of what features the people
 Categorical Injustice   5

are taken to have, and that, in turn, is influenced by the use of stereotypes. In our
example there are four people, three of whom are licensed to drive. One of the
three is a woman, one is a very large man, the third is a young boy, and the fourth
a frail man. As is clear, these are roles they assume relative to each other.
The question is who of the four people is going to drive the car and who
push? Here, we can imagine a stereotype kicking in: men are strong and good at
pushing cars, women are weak and not good at pushing cars. The woman ends
up driving the car while the two men and the boy push the car, even though the
woman is stronger than the frail man and he also has a license to drive. In the
negotiation of who plays what role in this case, stereotypes regarding men and
women kick in in the assignment.
The expectations of each role are the norms that people use to guide their own
action in the role and use to police each other’s behavior in their respective roles.
Associated with each role are thus two sets of rules, if you will: constitutive rules
and regulative norms. The constitutive rules outline what sort of behavior counts
as acting in the role at all; the regulative norms outline the standards for acting
well in the role (Cf. Burman 2007 and Witt 2011). Stereotypes can contribute
to the content of both sorts of rules. In our example, because gender stereotypes
played a role in the assignment of roles, the roles are themselves inflected by gen-
der. How do we spell out how they are inflected? My preferred way of capturing
that is to say that we have a further partitioning of the roles: there is the role of
woman driver and the role of man pusher and boy pusher. The norms governing
woman driver could be different from those of a man driver in that context, and
the norms governing large man pusher, frail man pusher, and boy pusher differ.
There are other ways to think about that. For example, we could say that for the
roles to be inflected by gender is for the woman to play both roles at once, driver
and woman. When there is a conflict in what one should do, we could say that one
trumps the other, depending on context, or even always. Charlotte Witt (2011)
suggests this picture. We then play multiple roles, but gender is a mega role that
trumps all others. I do not think we need to settle which route is better here.
What precisely is the content of the associated norms and from where do
they derive? The answer to that will vary from context to context. When we enter
a communal context such as a cocktail party we bring with us social maps from
the other contexts we have traveled in (see Ásta 2018, Ch. 6). These social maps
correspond to a set of assumptions about one’s own place in a social encounter as
well as that of others in that encounter. Here, Hegel’s ideas (1807) about subjec-
tification and objectification inspire. Just as for Hegel one consciousness forms
a conception of itself and the other it encounters and acts as if those conceptions
are valid, so people entering an encounter with other people bring with them sets
of assumptions about their own role in that encounter and that of the others they
meet. These sets of assumptions need not be conscious, although sometimes they
are.7  The social roles with associated norms for behavior correspond to a social
map. These social maps have locations which are the available roles to play in
the context and with each location comes norms for behavior for the occupant
6  Ásta

and for others. The social locations are fully intersectional in that each feature
that is socially significant in the context inflects the constraints and enablements
for the occupant of that location.8  The relationship between the constraints and
enablements and the associated norms is akin to the relationship between the
constitutive rules for a game such as chess and the norms for playing well. If you
flout the constitutive rules, you are not playing at all; if you flout the norms, you
are simply playing badly.9 
The constraints and enablements define what you can do or not do in the con-
text because they set the intelligibility frame for a person’s actions. This is key.
Actions that require the hermeneutical participation of others for success may be
blocked by this very intelligibility frame, as we shall see shortly.
So my suggestion is that stereotypes and stereotyping, as part of ideologies,
contribute in various ways to the formations of our intersectional social identi-
ties, but then also to the constraints and enablements we have on our actions in
the various contexts we find ourselves in. And as a social category is defined by
the constraints and enablements on the behavior of its members, stereotypes and
stereotyping contributes to the construction of social categories in these ways as
well.
Stereotypes and group-membership do not dictate every context, however.
Personal relationships and knowledge of each other can change what features
are salient in a context. The knowledge that the woman is the strongest of them
all, for example, can override the operation of the stereotype of women as weak.
Likewise the knowledge that the frail man is the best driver (and the other two
are terrible) can make salient not only the license to operate the car, but ability
to operate it.
Stereotypes are most likely to be operating in the contexts where people do
not know each other very well, or at all, and are judging on the basis of the social
groups they take people to belong to and the features they take them to have.
Knowing each other is, of course, not merely a matter of spending time together,
but of treating each other as individuals epistemically, as opposed to mere in-
stances of groups.

Action Impossible

When a person gets conferred a status certain actions become possible that
were not possible before and some may no longer be possible. For example, when
a person gets conferred the institutional status of instructor for a class, they are
entitled to set assignments, grade them, and report the grades to university au-
thorities at the end of the term. Some actions that may have been possible in the
past, are so no longer, on pain of losing the status, such as asking a student in
the class out on a date. Other actions get blocked out as unintelligible. Consider,
for example, the philosophy professor who appears to be asking a student if the
student will permit the professor to take the class. The range of available interpre-
tations include that there is some misunderstanding, that the professor is having
 Categorical Injustice   7

a mental health episode, and that the professor is mocking the student, but the
interpretation where the professor is earnestly asking to be permitted to the class
is blocked out.
I am interested in the cases where the status of a person in a context makes
actions either impossible or unintelligible. And I characterize categorical injus-
tice as a type of injustice where an individual is institutionally entitled to perform
an action but their action is blocked by their placement in a category. There is
a certain mismatch between what the person is entitled to do and what they are
able to do, given their placement. The paradigm examples involve a mismatch
between an institutional entitlement and communal status.10 
Consider the following scenario:11 
An engineer oversees a large construction project and has the institutional
authority to assign work. The whole group shows up for the first day of work.
Tom shows up late and, when he enters the room, the engineer says to Tom “you
will take section C.” Tom says “we’ll see about that. I’ll discuss it with the engi-
neer.” He does not realize that the engineer meant to be giving an order, because
the engineer is a woman. A variation of this is such that he understands very well
that he has been given an order, but he does not heed it, as he does not respect the
authority of the engineer because they are a woman.
Let us zero in on the first sort of case. Do we want to say that the engineer
gave an order, even though Tom did not understand it as such or do we want to
say that the uptake partly determines the nature of the action?12  I am inclined
to say that the order was issued but the lack of uptake made it unsuccessful. As
Austin, Searle, Bach, and others, stress various things can go awry with speech
acts, but not all of them make it the case that we do not perform the intended
speech act at all. And here an order is analogous with other speech acts such as
promising. If you promise to return my car, but do not return it, you have failed
in your obligation, as opposed to not having undertaken the obligation after all.
Similarly, if the engineer orders the worker to perform a certain task and the
worker does not perform it, then the worker has failed in their obligation. So there
are two sorts of cases: when the worker understands it as an order and does not
follow it and when the worker does not understand it as an order at all.
The former case is a straightforward case of failing in your obligation. What
about the case when the worker does not understand it as an order? Compare with
the case when Aba lends Belo money and Belo says “I’ll pay you back.” Aba
knows of Belo’s reputation for never paying people back and does not think they
will ever see the money. Sure enough, Belo does not pay back the money; in fact,
he had never intended to do so. He made an insincere gesture and Aba took it to
be insincere. However, it is clear that the type of gesture that was insincere and
taken to be insincere was a promise.13 
A variation on the case above that is perhaps closer to the case of the engi-
neer’s order we want to understand is a case where Belo undertakes the obligation
sincerely but is not believed (perhaps because of their track record). But such a
case is clear cut: Belo makes a promise, even though they are not believed. If they
8  Ásta

brought the money back a week later and said “that’s the money I promised to pay
you back,” Aba would be pleasantly surprised, not because they got a windfall
out of the blue, but rather because Belo had been sincere when Aba thought they
were not.
Let us return to our case of the engineer and the order. It may matter less
whether we want to say that what we have is an unsuccessful order or that the
engineer does not manage to issue an order at all. In that way we may avoid taking
a stand on some theoretical issues regarding the mechanics of speech acts. What
is clear is that the engineer does not manage to issue a successful order. The lack
of success is, however, not a fluke; it is a part of a pattern. Can we shed light on
what prevents the worker from understanding the order as an order? That is a part
of the project of this paper.
The key idea is that for many actions we perform, they are only successful if
they are understood by the people we are engaging with as we intend them. Many
speech acts are only successful if the audience interprets them as they are in-
tended. For example, an order that is not understood by the recipient as an order is
“unhappy”14  to use Austin’s language. I want to offer an account of how actions
can systematically be made unhappy by the intended interpretation of the action
being blocked by how the performer is socially constructed.
Ideas from the philosophy of language can help us to articulate how our ac-
tions can be systematically thwarted in a particular way, in particular the Gricean
notion of common ground, as developed by Robert Stalnaker, and David Lewis’s
idea of keeping score in a language game.
On Stalnaker’s view, the common ground in a conversational context is a set
of propositions or beliefs that are taken for granted by the participants in the con-
versation. Each conversational move changes that set as propositions are added or
subtracted. None of this need be conscious.
Consider a context with only three people, Alma, Ben, and Chris. Alma is
a Latina woman, Ben is an African-American man, and Chris is a genderqueer
Caucasian-American person. Ben and Chris are both philosophy professors and
Alma cleans their offices and is a nanny for Ben’s kids. The three of them are
having a conversation in the hallway in the philosophy department about Trump’s
immigration policies. At the beginning of the conversation the three participants
place each other on locations on a social map. Each location comes with con-
straints and enablements for the participant but these constraints and enablements
also offer the frame of intelligibility for their speech. These are the assumptions
taken for granted in the conversation, or the Stalnakerian common ground. The
common ground sets the intelligibility and permissibility conditions for the
conversational moves. For example, if Alma appears to be saying something
supportive of Trump’s immigration policies, Ben and Chris assume they have
misunderstood her, given that her identity in the context includes being a Latina.
The Stalnakerian common ground blocks it as an interpretation of what Alma
is saying. I am suggesting that a similar thing happens in the more general case
of action contexts. Where interlocutors’ interpretation of the intended action is
 Categorical Injustice   9

required for the action to be successful, the intended interpretation can be blocked
for systematic and unjust reasons.
Making use of an analogous idea of a common ground for action contexts,
we can then say that by placing a person on a social map the subjects also place
certain assumptions in the common ground, that is, the set of shared assumptions
in the context.
The placing of people on a social map is not always a simple affair that hap-
pens without struggle. Often the people in the encounter bring incompatible social
maps and some negotiation happens before people settle into their roles. Some
context are even too short for anyone to settle into any role. There are attempts at
placing each other onto social maps, but it is contested from start to finish. The
unfortunate aspect of the metaphor of a social map is that it suggests that we apply
a fixed map that remains static through the encounter, but a better way to think
of it is such that each action move in the context changes the shared assumptions
(the common ground) a little bit (cf. Lewis 1979; Stalnaker 1999). The social map
is then a dynamic entity, changing ever so slightly with each move in the context.
In each and every context we travel, there are certain features that are socially
salient and people taken to have those features get conferred onto them an inster-
sectional status consisting in constraints and enablements on their action because
being placed in a category is to have certain assumptions made about them such
that potential actions become unintelligible or impossible. Certain other actions
also become impermissible and performing them risks sanctions (although they
are intelligible and possible to perform).

Categorical Injustice

I have outlined above how placement in social categories frames the interac-
tion between people in contexts, how it sets the frame of intelligibility and pos-
sibility for their actions, as well as permissibility. It is now time to focus on the
cases that result in an injustice.
The main idea is this: To perform an action successfully, having some spe-
cific features is needed. While some person is entitled to perform the action, they
are thwarted in their effort to perform it because they are placed in a category
and the assumption about members of that category is that they lack one of the
requisite features. Because the intented interpretation of the action is required for
the action to be successful this assumption blocks the intended interpretation and
the person cannot perform the action successfully.
More formally: To perform A successfully, having features F1-Fn is needed.
S is entitled to perform A but is thwarted in their effort to do so because they are
placed in category C and the assumption about members of C is that they are not
Fi, for some i. That assumption blocks the interpretation of S’s action and as the
intended action is required to successfully perform the action, S is not able to
perform A successfully.
10  Ásta

This is the cognitional version of categorical injustice. There will be another

case where the intended interpretation is not blocked, but instead not recognized
and the interlocutor resists the agent. That is certainly an injustice, but requires a
different analysis. I term that one the “recognitional” one.
Let us apply this to the two engineering cases above.

Case 1: The Worker Does not Understand the Action as an Order

To perform A requires the authority to do so. E has the institutional authority.

However, E is taken to be a member of group G and the stereotype of group G is
that they do not have the institutional authority to perform A. W takes E not to
have the authority to perform A and the interpretation of the action where E is is-
suing an order is blocked. This is the cognitional version of categorical injustice.
Compare with the second sort of case.

Case 2: The Worker Understands it as an Order

In the case where the worker understands the engineer as giving an order,
but does not heed it, requires a different sort of analysis. Here, it is not that the
interpretation of the action as an order is blocked by the social construction of
the engineer, but rather that the engineer’s communal status as a woman blocks
her being recognized as having the institutional authority that she has. For it is
not enough to have the authority to issue the order, one’s authority needs to be
recognized. And here the features associated with being a woman include not
having the authority to issue an order, and the engineer’s attempt at issuing a
successful order is thwarted. The recognitional version of categorical injustice is
more formally thus:
To perform A requires the institutional authority to do so. E has the institu-
tional authority. However, W takes E to be a member of group G and not merit
such an institutional authority due to that group membership. Thus W does not
recognize E’s authority to perform A and resists it.
What is unjust about categorical injustice? The injustice lies in the fact that
the agent is entitled to perform the action but is thwarted in their effort because
of how they are classified or placed. It is a distinct sort of injustice perpetrated
against a person through socially constructing them as a member of a social group
where that social construction undercuts their ability to perform the action they are
entitled to. There is thus a mismatch between what a person is institutionally en-
titled to do and what they are able to do, given how they are socially constructed.
Note that a person can be subject to categorical injustice but thereby saved
from committing some morally heinous act. Consider a military officer who or-
ders her men to kill innocent children and suffers recognitional categorical in-
justice, but is saved from bearing the responsibility for the death of innocent
children.
 Categorical Injustice   11

Categorial Injustice and Ontic Injustice

Katharine Jenkins has recently (2016; forthcoming) isolated a certain injus-

tice that involves socially constructing a person as a member of a certain group
which she terms “ontic injustice.” Is categorical injustice a case of ontic injustice?
On Jenkins’s account being socially constructed, whether institutionally or
communally, comes with constraints and enablements, and these constraints are
unjust when there is a mismatch between what a person is morally entitled to do
and what they are able to do, given how they are socially constructed.
In the cases of categorical injustice, moreover, the mismatch is between what
a person is institutionally entitled to do and what they are able to do, given how
they are socially constructed.
So, despite the fact that both ontic injustice and categorical injustice involve
a phenomenon where the social construction of an individual as a member of a
group is implicated in an injustice, these two ways of highlighting how the social
construction of people can involve an injustice complement each other, rather
than being rival theories.

The Role in Discursive and Testimonial Injustice

I have offered a theory of how the social construction of individuals results

in an injustice that draws directly on the conferralist story of the construction
of social categories. The various accounts of epistemic and discursive injustice
focus on specific harms that are epistemic or discursive and some offer their own
theories of what mechanisms, that might be thought metaphysical, contribute to
the injustice while others leave that open. I isolate a way in which a specific ac-
tion is thwarted which can help with an understanding of when a certain epistemic
or discursive action is thwarted. There can, however, be harms and injustices that
do not involve thwarted action and categorical injustice is not meant to capture
those. In this way, categorial injustice can play a role in cases of epistemic and
discursive injustice, although an analysis of such cases of injustice may not be
exhausted by attending to categorical injustice. Here below, I highlight how cat-
egorical injustice plays a role in some paradigm cases of epistemic and discursive
injustice. To be sure, a theorist of those cases might reject my proposal and prefer
another one because of the theoretical commitments that categorical injustice re-
lies on, that is, the conferralist framework. An alternative framework would then
have to be found to account for the distinctly metaphysical mechanisms.

Fricker

On Miranda Fricker’s (2007) view group-based prejudice regarding a speaker

undercuts the success of speaker’s testimony, even though the person is entitled to
testify. However, to testify one must have credibility, which involves competence
and trustworthiness. The prejudice regarding the group is that its members are
12  Ásta

untrustworthy or incompetent or both, and that is how the successful testimony of

the speaker is thwarted. It is impossible for the speaker to testify and be believed.
Testimonial injustice is thus clearly a case of the following phenomenon: To
perform action A, having features F1 (competence) and F2 (trustworthiness) is
needed. Speaker Sis thwarted in their effort to perform A because they belong to
a group G and the prejudice about group G is that they are not-F1 and/or not-F2.
This is a case of cognitional categorial injustice.

Dotson

What about Kristie  Dotson’s (2011)  cases of testimonial quieting? Testimonial

quieting may seem to fit nicely the general framework: In order to speak, an agent
needs to be given credibility, that is, competence and trustworthiness. And be-
cause of stereotypes about black women, where they are taken to be incompe-
tent and/or untrustworthy, their attempts to speak successfully are thwarted, even
though they are entitled to perform that action: they are not seen to be capable
of being speakers so there is no uptake when they attempt to speak. Testimonial
quieting thus fits with the general framework offered.15  There can be cognitional
and recognitional versions of categorial injustice at work in these sorts of cases.

Langton

Rae Langton’s (2009) paradigm case of illocutionary disablement involves

a person’s attempt at refusing sex but her refusal is silenced. The silencing in
question is not locutionary: she is not gagged or her mike turned off. She utters
the word “No” and attempts to thereby refuse sex, but her word is not taken up
as signifying refusal. Does this case fit in our framework? Yes, here is how: the
person belongs to group women and associated with the group in this context are
stereotypes about what women are like and what they want, as well as norms for
how to act in this context.
Women in sexual contexts are taken to have some features F1 and F2. For
example: women in a sexual context are taken to want sex, yet not want to offer
themselves, but rather be taken by the man. If Langton is right, then the main-
stream heterosexual porn industry perpetuates that stereotype. In order for the
woman to be able to perform the action of refusal she has to be able to report on
her desires. In this case, any speech act that entails that she does not desire sex
goes against the stereotype of what women are like in the context and is disabled
(blocked as an interpretation). More formally: a person is taken to belong to group
G and the stereotype of members of the group is that they have feature F. Any
potential speech act that entails that the person does not have feature F is disabled.
This is a case of cognitional categorical injustice.
The woman is not able to exercise her capability as a speaker and be heard
for what she means in the context. She can only assume the role of prop in the
sexual game that is the man’s fulfilling his desire. Anything she attempts to say or
 Categorical Injustice   13

do in the context gets interpreted through the interpretative framework of the porn
roles. Her attempt to refuse is merely an encouraging move in the courting ritual.

Kukla

Rebecca Kukla’s (2014) paradigm case involves the phenomenon that a per-

son is thwarted in her attempt to play a certain role by her group membership, but
instead of illocutionary disablement we have a different unintended illocutionary
force. A woman foreman has the institutional authority to issue an order but does
not have the adequate communal standing to issue it and is instead taken to be
issuing a request. Here, we can say that in order to issue orders a person needs
to be taken to have authority over the receiver. A woman, in virtue of her group
membership, is taken to have features incompatible with having authority and
thus her attempt to exercise that authority is thwarted. Her communal status in the
context involves being a woman and the features associated with being a woman
undercut her ability to exercise her institutional role and issue an order. Instead,
she is taken to issue a request. The illocutionary force of her utterance is different
from what she intended. This case fits nicely in the framework. Because of how
the foreman is socially constructed, she is unable to perform certain actions she
is institutionally entitled to perform, and instead is taken to perform other (unin-
tended) actions. This is a case of cognitional categorical injustice.

Conclusion

I have offered a theory of how social construction can result in an injustice

that is distinctly metaphysical. I have labeled this injustice “categorical injustice”
and isolated two versions of it, cognitional and recognitional. I have compared
categorical injustice to a form of metaphysical injustice recently theorized by
Katharine Jenkins as well as discussed how categorical injustice can play a role in
discursive and epistemic injustice.
I would like to thank Erin Beeghly, Jules Holroyd, and Katharine Jenkins for
helpful comments on earlier drafts. Naturally, I am the only one responsible for
the views herein and any remaining errors.

Notes
1
We can exit and reenter institutional contexts (think about entering and exiting a jurisdiction where it
is illegal to be gay), but perhaps for certain contexts it may make more sense to say that they are
latent or inactive and can get activated. Consider Dolphin Club members who are not swimming
in the San Francisco Bay or conversing with other Dolphins, but merely walking around town, or
when California licensed drivers are blissfully not driving around California (or stuck in Califor-
nia traffic), but rather hiking in the Sierras or picking blueberries in Maine.
2
I address the concern regarding the proliferation of properties in Chapter 2 of Ásta 2018.
3
The base property for the conferral of a communal property need not be institutional, but could be,
say, physical. We then would have two properties: for example, having a large nose (base prop-
14  Ásta

erty) and being presumed to have a large nose (communal status).

4
It is compatible with this picture that the generalization of the group is correct. Also, a stereotype can
involve a normative judgment such that people of group G are lesser than people of group H or
that people of group G have a lower moral status than others.
5
I am using roles and statuses interchangeably in this essay.
6
This includes a lot of factors and my characterization is not exhaustive. Consider, for example, the
weather. If there were a freak thawing of the snow, no one would have to push anything. Also in
case of a freak deep freeze, no one could push.
7
These are akin to psychological schemata (see Valian 1997), but need not be conscious. When con-
scious, they form part of the explicit ideology.
8
How does each significant feature contribute to the constraints and enablements on the person’s ac-
tions in the context? Ann Cudd (2006) has suggested that the contribution of aspects of identity
not be thought of as additive, but on a vector model, and that might be a promising way to be
responsive to intersectional worries about additive models of oppression (cf. Crenshaw 1989).
It is, however, an empirical question what the constraints and enablements are in each context.
9
Given that in the communal case, there are no deontic rules, it may be hard to see how this distinction
can be upheld in such a case. Here, I think Charlotte Witt’s work on social normativity is helpful.
We then distinguish between being subject to certain social norms at all versus how well we do
with regard to those norms.
10
I do not rule out other sorts of cases.
11
I consider this to be a variation of an example discussed by Rebecca Kukla (2014).
12
This may make us favor one more than another account of what determines the illocutionary act. For
a recent overview of options see (Fogal, Harris, and Moss 2018).
13
Someone might want to say that what we have here is not a case of promise at all, but rather an
acknowledgment of receipt of the money by uttering the words “I’ll pay you back,” much in the
same way that saying “how are you?” is a greeting and not a question and “let’s do lunch” is not
an invitation to lunch. If we acknowledge that possibility there is a certain tension that arises, but
perhaps it is resolved by saying that an insincere promise functions like an acknowledgement of
receipt. I do not think it matters for our purposes here how we resolve the tension.
14
I think it may be up for debate what sort of unhappiness is involved so I would not delve further into
Austin’s own classifications.
15
How about smothering? The complicating factor is that the person carrying out the smothering is
the victim herself. Here, the victim cannot testify properly because she takes her interlocutor
to associate certain features with her because of her group membership. In this case, the victim
refrains from assuming the role of testifier because she thinks it not only hopeless but also risky.
This sort of case is certainly different from the standard case where the person in questions at-
tempts to take on role R but is thwarted in their effort to play it. It nevertheless is a case where
group stereotypes interfere with a person’s ability to play a certain role, but the victim herself is
the agent of the injustice, if by proxy.

References

Ásta. 2018. Categories We Live By. New York: Oxford University Press.
Beeghly, Erin. 2015. “What is a Stereotype? What is Stereotyping?” Hypatia 30 (4): 675–91.
Burman, Åsa. 2007. Power and Social Ontology. Lund, Sweden: Bokbox Publications. Published
under the name “Åsa Andersson”.
Crenshaw, Kimberle. 1989. " Demarginalizing the Intersection of Race and Sex: A Black Feminist
Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics." University of
Chicago Legal Forum, Article 8. Available at: https​://chica​gounb​ound.uchic​ago.edu/uclf/vol19​
89/iss1/8
Cudd, Ann. 2006. Analyzing Oppression. Oxford: Oxford University Press.
Dotson, Kristie. 2011. “Tracking Epistemic Violence, Tracking Practices of Silencing.” Hypatia 26
(2): 236–56.
 Categorical Injustice   15

Fogal, Daniel, Harris, Daniel W. and Moss, Matt. 2018. New Work on Speech Acts. Oxford: Oxford
University Press.
Fricker, Miranda. 2007. Epistemic Injustice. Oxford: Oxford University Press.
Geuss, Raymond. 1981. The Idea of a Critical Theory: Habermas and the Frankfurt School.
Cambridge: Cambridge University Press.
Hegel, G. W. F. 1807. Phänomenologie des Geistes. Bamberg and Würzburg, Germany: Joseph
Anton Goebhardt.
Jenkins, Katharine. 2016. Ontic Injustice. PhD thesis, University of Sheffield.
   . forthcoming. “Ontic Injustice.” Journal of the American Philosophical Association.
Kukla, Rebecca. 2014. “Performative Force, Convention, and Discursive Injustice.” Hypatia 29 (2):
440–57.
Langton, Rae. 2009. Sexual Solipsism. Oxford: Oxford University Press.
Lewis, David. 1979. “Scorekeeping in a Language Game.” Journal of Philosophical Logic 8 (1):
339–59.
Stalnaker, Robert. 1999. Context and Content. Oxford: Clarendon Press.
Valian, Virginia. 1997. Why So Slow? Cambridge, MA: MIT Press.
Witt, Charlotte. 2011. The Metaphysics of Gender. Oxford: Oxford University Press.