The Epistemic Role of Outlaw Emotions

  • Laura Silva (University of Geneva)


Outlaw emotions are emotions that stand in tension with one’s wider belief system, often allowing epistemic insight one may have otherwise lacked. Outlaw emotions are thought to play crucial epistemic roles under conditions of oppression. Although the crucial epistemic value of these emotions is widely acknowledged, specific accounts of their epistemic role(s) remain largely programmatic. There are two dominant accounts of the epistemic role of emotions: The Motivational View and the Justificatory View. Philosophers of emotion assume that these dominant ways of accounting for the epistemic role(s) of emotions in general are equipped to account for the epistemic role(s) of outlaw emotions. I argue that this is not the case. I consider and dismiss two responses that could be made on behalf of the most promising account, the Justificatory View, in light of my argument, before sketching an alternative account that should be favoured.

Click the links below to download an audio version of this article's content:

Keywords: outlaw emotions, epistemology, emotion, epistemic perceptualism

How to Cite:

Silva, L., (2021) “The Epistemic Role of Outlaw Emotions”, Ergo an Open Access Journal of Philosophy 8: 23. doi:



Published on
13 Dec 2021
Peer Reviewed

1. Introduction

Outlaw emotions are emotions that stand in tension with a large set of an agent’s beliefs. They are recalcitrant emotions, as they conflict with the agent’s evaluative judgements, but they are not merely recalcitrant. In typical cases of recalcitrant emotion, such as fear of a dog that one believes is not dangerous, there need be only one belief that the emotion conflicts with.1 In outlaw emotion cases, although there is typically a belief with which the emotion conflicts, making the emotion recalcitrant, the emotion also stands in tension with a large set of further beliefs, often clashing with an agent’s wider belief system. This is what makes this class of emotions distinctive. Outlaw emotions are prevalent under conditions of oppression.2 Consider the following two cases:

Discontent Housewife:

A woman who has been taught that a ‘woman’s place is in the home’ may be driven to question this maxim precisely in light of her persistent dissatisfactions and repeated urges to flee from the responsibilities and limitations which structure her domestic life.… if her highest principles themselves also include notions of ‘appropriate’ sex roles, duties to others and the importance of self-sacrifice as an ideal of femininity, then there is not much available among her highest principles to afford an independent standpoint for assessing the maxim about woman’s place. Her frustration, grief, and depression, and the motivations to change her life which spring from these sources, may be her only reliable guides. (Friedman 1986: 31)


Raquel is a woman living under conditions of gender oppression in which the concept of sexual harassment is not available. Raquel is on a night out with a group of girlfriends. She feels someone squeeze part of her body and turns to see a man she has never met before. The man smiles at Raquel in acknowledgement of having been the person to touch her, and walks away. Some of Raquel’s friends feel excited by what just happened, and encourage her to go talk to the man. Others aren’t excited but jealous for not having been the ones squeezed. Raquel believes she should be flattered by the attention she has received, as well as proud to have been the one approached, and indeed she does feel a mixture of these emotions, but she also feels uneasy and angry. Raquel forms the belief that what the man did was not ok, based on her negative emotional responses, despite this belief going against her wider set of beliefs and emotional responses, as well as those of her peers.

Jaggar coined the term ‘outlaw emotion’ to capture those emotions ‘distinguished by their incompatibility with the dominant perceptions and values’ (1989: 166) that agents have internalized.3 At least sometimes outlaw emotions will be better guides to truth than those beliefs with which they stand in tension. This seems to be what is going on in the example cases above. The agents in these cases are getting something right about reality, and they are doing so despite having internalized oppressive ideology.4 These example cases can be seen to involve appropriate, or fitting,5 recalcitrant emotions that occur under conditions of oppression.6 Feminist philosophers have argued that outlaw emotions occurring under conditions of oppression often play crucial and radical epistemic roles by responding to reasons the agent may have otherwise not tracked, given that reasoning would tend to favour the conflicting oppressive beliefs (Friedman 1986; Narayan 1988; Jaggar 1989; Campbell 1994). Raquel’s belief that ‘what the man did is not ok’ seems justified, and it is her outlaw anger that seems to provide this justification. I focus on such cases throughout, and use the term ‘outlaw emotion’ to refer to ‘outlaw emotions that occur under conditions of oppression’ for simplicity. I will refer to the beliefs that outlaw emotions stand in potentially justificatory relations to as ‘outlaw beliefs’ throughout.

First personal experience of outlaw emotions can be disorientating as agents often lack the resources to name or interpret their experience. In cases where these emotions are more readily made sense of, agents will often experience cognitive or affective conflict, given that these emotions stand in tension with their wider set of beliefs. I will take for granted in what follows that outlaw emotions form a sufficiently cohesive class of emotions to permit inquiry into their epistemic role. Additionally, following most discussion on emotions in general, I focus on occurrent and intentional cases of outlaw emotions such as the ones illustrated in the above cases.

We need not be committed to the implausible view that outlaw emotions always track reasons, nor that they always do so better than our belief systems, so as to accept their epistemic value. We need only hold that sometimes emotions track reasons better than our belief systems do. To deny this weaker claim would be to deny the role of consciousness raising, where collective efforts to make sense of emotional experiences shared across groups led to critical progress (Fricker 1991; 2007; Thompson 2006). It would also be to deny the common phenomenon of feeling a certain way, while not knowing exactly why, and later coming to realize one’s feelings were justified. Indeed, the thought that emotions sometimes track reasons better than our belief systems do is prevalent within the philosophy of emotion, as well as the philosophy of action, where emotions are common motivators of inverse akrasia (see MacIntyre 1990; Arpaly 2000; Tappolet 2016).

In what follows, I take for granted that outlaw emotions are epistemically important. This point is, by and large, uncontentious. Specific accounts of how outlaw emotions play epistemic roles remain largely programmatic in the work of those who highlight their epistemic value. In contemporary philosophy of emotion, on the other hand, there is active debate over the epistemic roles played by emotions in general, but outlaw emotions have not received sustained attention. Due to this, it is often assumed that the dominant ways of accounting for the epistemic roles of emotions in general will be able to account for outlaw emotion cases as well. The specific assumption in the literature is that existing accounts can capture outlaw beliefs, such as ‘what the man did was not ok’ in Harassment, as justified (Deonna & Teroni 2012a; Tappolet 2014; 2018; Sreenivasan 2018; Arpaly 2018).7 We will see that the epistemic role envisioned can be more or less direct, depending on which view of the epistemic role of emotions is favoured. I argue that existing accounts, which I call the Motivational and the Justificatory view respectively, struggle to deliver the relevant propositions, outlaw beliefs, as justified in outlaw emotion cases. This should worry us as, contrary to claims in the literature, we are left without an account that can capture beliefs such as Raquel’s in Harassment as justified. I end by sketching an alternative view that is promising with regards to its ability to remedy this.

2. Motivation and Justification

In Discontent Housewife, the outlaw emotions are ‘persistent’ and ‘repeated’ urges that motivate the agent to question her oppressive beliefs. Outlaw emotions seem here to be motivating the agent to assess her current beliefs about a woman’s place in the home. This assessment likely involves the outlaw emotions motivating inquiry into why one is experiencing the relevant emotion. In Harassment too, Raquel’s anger will likely motivate her to think about why she has responded so differently to the man’s actions, as compared to her friends. Raquel’s anger, in fueling a search for its reasons, is liable to challenge her conflicting emotional dispositions and wider set of oppressive beliefs. This motivational epistemic role can be summarized as follows:

Motivational View (MV): Emotions motivate agents to search for the reasons for their emotions.

Brady (2013) has been MV’s most vocal advocate. He holds that in capturing and consuming our attention, emotions motivate search for reasons that could justify our emotional responses. Those reasons that are uncovered can also provide justification for evaluative beliefs about the world. This is an epistemic role that should not be underestimated. By fueling reflection and inquiry that uncovers motivations and reasons for valuing certain actions and aims, our emotions are likely to play key roles in gaining evaluative knowledge and understanding.8 In outlaw emotion cases this motivational role seems paramount, for there is much to uncover, and many obstacles to doing so given the internalization of oppressive ideology. Having strong and persistent motivations to question one’s beliefs and search for the reasons for one’s emotions seems crucial. On MV, outlaw emotions motivate much needed inquiry, and it is this inquiry that uncovers reasons that can then provide justification for outlaw beliefs, such as Raquel’s belief that ‘what the man did was not ok’. Outlaw emotions contribute in this crucial yet indirect manner to the justification of outlaw beliefs on MV.

A separate view does not deny this motivational role, but holds that emotions also play a more direct epistemic role. Emotions can themselves, on this view, justify evaluative beliefs. I call this the Justificatory View. In becoming angry, Raquel may well be motivated to inquire into the reasons for her emotion, but simply by experiencing it, she is experiencing the man’s actions as problematic, even if she doesn’t understand why. Raquel and her friends perceive the event in different ways given their distinct emotional responses. That the event emotionally strikes Raquel as wrong is sufficient to defeasibly justify the belief that she has been wronged, according to some (Johnston 2001; de Sousa 2007; Döring 2007; Tappolet 2016; Cowan 2016). This view can be summarized as follows:

Justificatory View (JV): Emotions provide immediate defeasible propositional justification for evaluative beliefs.9

The evaluative beliefs we are concerned with are those which have a similar content to the outlaw emotions. Typically, these beliefs have the form ‘x is F’ where x is the object of the emotion, and F is some evaluative property attributed to the object. The propositions in question spell out in conceptual terms what we take to feature in the representational content of the emotion.10 JV is most often endorsed by Perceptual Theorists, who take analogies with perception to be central to understanding emotions (Johnston 2001; de Sousa 2007; Döring 2007; Tappolet 2016; Cowan 2016).11 An epistemic analogy with perception is typically appealed to whereby emotions, like perceptual experiences (on standard accounts), are themselves sufficient to defeasibly justify relevant propositions. Emotions, on this view, stand to the relevant evaluative beliefs, as perceptual experiences stand to the relevant empirical beliefs. That you perceive a ball to be red gives you defeasible reason to believe that the ball is red, much as your experience of fear gives you defeasible reason to believe the object of your fear is dangerous.

It is likely that in outlaw emotion cases, where conceptual resources relevant to one’s experience are typically underdeveloped, the content of the evaluative beliefs will be somewhat vaguer than in other, non-outlaw, cases where thick evaluative concepts are available. Further inquiry into the representational content of these emotions is, I think, called for. I cannot pursue this inquiry here. For current purposes I take the content of outlaw emotions to be sufficiently determinate such as to at least stand in potentially justificatory relations to propositions with vague evaluative content such as ‘what that man did was not ok’, as opposed to ‘that man sexually harassed me’. It is plausible that a given outlaw emotion can defeasibly justify a range of closely related evaluative beliefs. In Raquel’s case, for example, her anger could defeasibly justify the belief that ‘that man has harmed me’, or ‘what that man did was not nice’, and variations on these.

JV comes in two main varieties, Reliabilism and Phenomenalism.12 For Reliabilism, it is the claimed reliability of emotions in generating true beliefs that accounts for their justification conferring capacities (see Pelser 2014), while for Phenomenalism, it is the phenomenology of emotional experiences that account for their justification conferring capacities (see Tappolet 2016). JV can therefore come in versions that vary in their alignment with epistemic externalism (Reliabilism) or internalism (Phenomenalism). Whichever variety one prefers, JV seems particularly well placed to account for a direct epistemic role in outlaw emotion cases. If emotions provide immediate defeasible justification, no further beliefs, for example, regarding the reasons for one’s emotion, or the reliability of one’s emotional experience, are required for the justification of an outlaw belief. This fits outlaw emotion cases well, as these are cases where the agent may feel confused by the very experience of the outlaw emotion, not understanding why they are feeling this way. Indeed this view has been advocated as ideally suited to account for outlaw emotion cases where non-emotional access to one’s reasons is typically lacking (see Tappolet 2014).13

The motivational and the justificatory views capture the two main ways of thinking about the epistemic role of emotions in general,14 and are similar to the roles mentioned by Jaggar (1989) concerning outlaw emotions specifically:

The most obvious way in which feminist and other outlaw emotions can help in developing alternatives to prevailing conceptions of reality is by motivating new investigations.… As well as motivating critical research, outlaw emotions may also enable us to perceive the world differently from its portrayal in conventional descriptions. (1989: 167, emphasis my own)

But whether and how exactly outlaw emotions might play these epistemic roles has not been examined. MV and JV have been asymmetrically related, that is, proponents of the motivational view typically deny that emotions play justificatory roles, while proponents of the justificatory view typically do not deny that emotions also play a motivational epistemic role.15 I begin by arguing that although the motivational view captures an important epistemic role that outlaw emotions play, it only manages to secure these emotions a limited, and indirect, role, such that the view struggles to deliver justified outlaw beliefs in outlaw emotion cases. The ways in which MV falls short invites the justificatory view to account for the direct epistemic role outlaw emotions seem to play. As the justificatory view appears ideally equipped to account for outlaw emotion cases, I will spend more time dissecting the obstacles it faces. I argue that on JV, as it currently stands, the justification of outlaw beliefs is always defeated. I consider two lines of response to my argument on behalf of the justificatory view, before sketching a more promising alternative view that preserves JV’s main insight.

3. Limitations of the Motivational View

That emotions often motivate agents to inquire is not contentious. But this motivational epistemic role falls short of an account of the epistemic role assumed of outlaw emotions for two main reasons: inquiry in outlaw emotion cases is both compromised, as well as particularly demanding. We will see that, due to this, the motivational view has a hard time delivering on the assumption that outlaw beliefs are justified. This would mean that Raquel’s belief in Harassment, that ‘what the man did was not ok’, is unjustified.

First, outlaw emotion cases are ones where inquiry will be compromised by the internalization of oppressive ideology. Outlaw emotions can seem alien to the emoting individual themself, and their best efforts at making sense of their experience may fail. If the agent felt an emotion unrelated to dominant ideology this need not be the case, but because in outlaw emotion cases the emotion concerns exactly the domain over which the agent is susceptible to oppressive ideology, their ability to reflectively understand their emotion, at least on a short time scale, is compromised. Emotion-motivated inquiry is therefore less likely to yield reliable results in outlaw emotion cases. As Jaggar writes:

When unconventional emotional responses are experienced by isolated individuals, those concerned may be confused, unable to name their experience; they may even doubt their own sanity. Women may come to believe that they are ‘emotionally disturbed’ and that the embarrassment or fear aroused in them by male sexual innuendo is prudery or paranoia. (1989: 166)

Second, when inquiry in outlaw emotion cases is not entirely blocked, we can expect it to be far more laboursome than in other scenarios. We can expect outlaw emotion-motivated inquiry to not only require a longer time span to yield potential results, but to also rely on factors beyond the emoting agent and their immediate environment. To this effect Jaggar adds that:

When certain emotions are shared or validated by others, however, the basis exists for forming a subculture defined by perceptions, norms, and values that systematically oppose the prevailing perceptions, norms, and values. (1989: 166)

The capacity for outlaw emotion-motivated inquiry to yield epistemic fruits is likely to depend on the agent interacting with others that experience similar emotions, where this interaction may involve long and difficult work untangling shared experience. As MV currently construes inquiry primarily as an individual endeavor (Brady 2013), as it stands, the view falls short at accounting for the relevant motivational epistemic role outlaw emotions play. Although Raquel’s anger may have a crucial motivational role to play in starting inquiry then, this inquiry is unlikely to deliver epistemic fruits in the short-term, due to inquiry being compromised by oppressive ideology, or in the long-run, so long as MV makes no reference to interaction with other agents as a feature of the inquiry envisioned. On the motivational view then, the epistemic role of outlaw emotions is important, but it is significantly limited by features of such cases.

More problematically, however, even if MV were modified in an attempt to remedy this, the view would still be hard pressed to deliver on the assumption that outlaw beliefs are justified in outlaw emotion cases. Recall that proponents of MV typically deny JV by holding that emotions only play motivational epistemic roles, and cannot, themselves, justify evaluative beliefs. MV demands that the agent access the reasons for her emotion non-emotionally for these reasons to themselves play justificatory roles. On a motivational view, the justification of Raquel’s belief that ‘what the man did was not ok’ is dependent on the potential fruits of a compromised and particularly laboursome inquiry. Only if, and when, the results of this inquiry are available to Raquel, will her outlaw belief be, on this view, justified.16

This overly demanding account of the justification is one of MV’s greatest shortcomings. JV, on the other hand, grants emotions a direct justificatory role, as emotions can be taken at face value in the justification of beliefs. I will devote more attention to JV for this reason, as this makes it seemingly ideally placed to account for outlaw emotion cases. For now, it should be clear that while MV grants outlaw emotions a crucial epistemic role in triggering much-needed inquiry, this inquiry will be especially compromised and laboursome in ways that MV, as it stands, is ill-equipped to accommodate. Moreover, as MV grants emotions only an indirect role in the justification of beliefs, outlaw beliefs will remain unjustified less the emotion-motivated inquiry is successful. A view that allows an immediate epistemic role of outlaw emotions, such that need for inquiry is bypassed, seems called for.

4. A Problem for the Justificatory View

JV seems to deliver where MV falls short. It grants emotions themselves justificatory force, such that no search for reasons, nor any further beliefs, or inferences, are needed for the defeasible justification of outlaw beliefs. This view has strong intuitive appeal, for we tend to quickly form beliefs based on our emotions, and take their content for granted until disconfirming evidence is made apparent. The undemanding nature of JV makes it seem particularly well-suited to account for outlaw emotion cases where agents typically lack access to the reasons for their emotions. Indeed, proponents of JV have argued that it is a virtue of their view that they can accommodate outlaw emotion cases such as Friedman’s (1986) Discontent Housewife (Tappolet 2014). Recall that on JV nothing besides the emotion is required to ensure this justificatory relation. All that is required for the justification of Raquel’s belief that ‘what the man did was not ok’, is that Raquel experience outlaw anger, and that no defeaters be at play.

JV’s appeal in accounting for outlaw emotion cases is, I argue, unfortunately merely apparent, due to an underappreciated feature of the view. JV says that outlaw beliefs will not be justified if there are defeaters present, as these would defeat the justification conferred by the emotion. But in outlaw emotion cases, agents have internalized a vast range of beliefs that stand in tension with the emotion. These beliefs are likely to include both rebutting as well as undercutting defeaters (Pollock 1986). Patriarchal societies are often characterized by a dismissal of emotions as inherently irrational and tied to the subjugated feminine (Jaggar 1989; Fricker 1991). The two example cases above involve gender oppression, making it likely that the agents portrayed in them believe their emotions to be unreliable means of acquiring knowledge. These beliefs could undercut the justification conferred by outlaw emotions. While the presence of these undercutting defeaters is contingent on the particular modes of oppression at play in a given case, the presence of rebutting defeaters seems to be built into all outlaw emotion cases. Recall that Raquel, for example, has internalized beliefs that she should be flattered, she believes the man’s actions are desirable. She therefore believes that the man’s actions are (more than) ok. This oppressive belief conflicts with her outlaw belief that ‘what the man did was not ok’, and is liable to act as a rebutting defeater against its justification. This worry can be summarized as follows:

  • P1: JV claims that emotions are a source of immediate defeasible propositional justification for evaluative beliefs.

  • P2: Outlaw emotion cases involve rebutting defeaters.

  • P3: If P1 and P2, the justification of outlaw beliefs is always defeated on JV.

  • __________________________________________

  • C: The justification of outlaw beliefs is always defeated on JV.

The following options are available to a proponent of JV in response to this argument:

  • 1) Drop JV’s defeatism commitment.

  • 2) Drop JV’s immediacy component.

  • 3) Accept the conclusion.

  • 4) Reject P2.

I take the forth option to be the most promising one. Option 1 is too permissive. Emotional justification would on such a move not admit of defeaters. This is an undesirable option as it secures emotions a stronger epistemic role than we typically grant other sources of epistemic justification. Option 2, on the other hand, is too demanding. This option would mean that emotions can only justify evaluative beliefs when the agent holds separate (likely justified) beliefs, about the positive epistemic status of the reasons on which the emotion is based, for example, or about the reliability of their emotion. This would not only rule out outlaw emotions cases from involving direct justificatory roles, but any case where the agent lacked these further beliefs. In ruling such cases out, option 2 would risk denying the justificatory force of emotions altogether, as a purely emotional route for justification is blocked. The proponent of JV could accept the conclusion of my argument instead (option 3). They could rest content with the limits of their view and highlight that at least JV manages to secure outlaw emotions a justificatory role, that is, these emotions still provide prima facie justification to outlaw beliefs, despite the ultima facie justification of these beliefs being defeated by the conflicting oppressive beliefs. This would be an unsatisfying bullet to bite for proponents of JV, as outlaw emotion cases would always involve defeated outlaw beliefs. Indeed, proponents of JV have been reluctant to accept this concessionary conclusion (see Tappolet 2018). Accepting the conclusion of my argument would in any case reveal JV to be far less suited to accounting for outlaw emotion cases than it has been assumed. Option 4 involves denying the premise that outlaw emotion cases involve rebutting defeaters. I take this to be the most promising option.17 To deny the conclusion of my argument then, a proponent of JV would need to deny that the conflicting oppressive beliefs in outlaw emotion cases are defeaters (P2). The main way of doing so, I take it, is to endorse the view that defeaters must have the positive epistemic status of being themselves justified, and then argue that the would-be defeating beliefs in outlaw emotion cases cannot act as defeaters because they are unjustified.

This move comes in internalist and externalist varieties. Internalists typically take justification to depend only on facts about an agent’s non-factive mental states, while externalism denies this. Externalists take justification to depend at least partly on how things stand beyond the individual’s mental states.18 We are concerned with whether moves to deny the positive epistemic status of the would-be defeaters in outlaw emotion cases go through to allow a denial of P2. I will argue that moves open to the internalist and the externalist alike face serious difficulties.19

4.1. Denying P2: The Internalist

An internalist response to my argument can be developed in two main ways. Either by adding a counterfactual internalist condition such that the conflicting belief comes out as unjustified, or by making a distinction between the reasons an agent has and the reasons an agent takes themselves to have. I will outline each response in turn and argue that they both face serious difficulties in delivering a denial of P2. As such, the externalist move will emerge as preferable.

The first internalist move involves endorsing the condition that the oppressive propositions believed in outlaw emotion cases can only act as defeaters if they are ones that the agent would endorse had they had time to reflect. In other words, only beliefs that would have survived reflection are justified and hence can act as defeaters. The thought is that Raquel and the discontent housewife believe oppressive propositions that would not survive careful reflection. This move takes the would-be-defeating belief to be one whose epistemic status the agent herself is in a position to undermine. This is a problematic assumption.

Recall, firstly, that Raquel and the discontent housewife both likely lack access to the reasons for their emotions. Reflection is likely to favour the conflicting oppressive proposition given that they inhabit, and have internalized, oppressive ideology. It is therefore unlikely that agents in outlaw emotion cases are in a position to reflectively undermine their oppressive beliefs. This means that the oppressive beliefs in outlaw emotion cases likely remain justified on this move. At the very least, this move will restrict the cases in which outlaw beliefs remain undefeated to the subset of cases in which the agent could have reflectively undermined the relevant beliefs, resulting in a drastic restriction of those cases that JV would accommodate.

To secure more than this, the counterfactual condition would have to be much stronger than one that invokes reflection, it would have to say something like this: Only conflicting beliefs that do not survive reflection, personal development, research, and collective consciousness raising efforts can defeat the justification of outlaw beliefs. This is likely far too demanding a counterfactual condition to endorse. First, invoking such a strong condition might seem ad hoc. That is, in so far as this move says that oppressive beliefs don’t count as defeaters if they would be undermined by a radically enlightened agent, the move invokes an idealized counterfactual condition designed to deliver the desired outcome: that oppressive beliefs be undermined. Most importantly, however, the move is actually ill-equipped to deliver its desired outcome. An internalist move of this sort assumes that the strong counterfactual condition is a scenario in which the would-be-defeater would in fact be undermined by the enlightened agent. This need not be the case. It is possible that many of the oppressive propositions in outlaw emotion cases turn out to be internally justified after critical feminist reflection. The positive value of being a housewife, for example, might still be endorsed, perhaps for distinct reasons; but the proposition itself could remain justified in the counterfactual scenario, as it may not be disavowed by the ideally enlightened agent. It seems, therefore, that many oppressive propositions will remain justified, and hence apt to defeat the justification of outlaw beliefs, even on such a strong counterfactual condition. Finally, such a move seems to imply that any belief that one later disavows at t2 is not justified at t1, a consequence we may not sit well with.

A second internalist move would be to hold that in outlaw emotion cases, the agent really does have more (internal) reasons in favour of the emotion, despite this not being apparent to the agent herself. The thought is that the agent has more reason for the emotion, and hence the outlaw belief, than for the conflicting oppressive belief. The agent is just mistaken about what reasons she takes herself to have, such that she believes an internally unjustified proposition. As the oppressive proposition is internally unjustified it can’t act as a defeater against the emotion, and therefore the emotion can justify the associated belief in outlaw emotion cases.20 Discontent Housewife would, on this view, involve the housewife having most internal reason to be angered about her situation—given that she values her own independence and has professional ambitions, for example. But these reasons, that she has, are not accessible to her. Perhaps they have been psychologically repressed by her hostile social environment, such that she makes a mistake about what reasons she takes herself to have.

An initial problem with this move is that it seems to face similar obstacles to the one just considered, for which reasons count as ones the agent has but lacks access to? Intuitive candidates seem to be exactly those reasons that one could have reasoned to, and over time realized that one has (see MacIntyre 1990). This move will therefore face the same worries as the one above. In addition to these however, this move claims that in outlaw emotion cases the agent really has, at the moment of experiencing the outlaw emotion, more reason in favour of the emotion than in favour of the conflicting oppressive belief. This doesn’t seem right. Of course, the agent must have some internal reason for the emotion, otherwise the occurrence of the emotion in that individual is entirely mysterious. But it would be a mistake to use this trivial fact to infer that the agent has more reason for her emotion than for the conflicting oppressive proposition, at least in so far as this is meant as a general account of what is going on in outlaw emotion cases. Outlaw emotions are exactly emotions that go against the agent’s web of internalized oppressive beliefs about the world. The conflicting oppressive belief is one component of the internalized worldview, and it therefore better coheres with the other beliefs, emotions, and intentions that stand in line with that framework. Such cases are best described as ones where the agent has most internal reason in favour of the oppressive conflicting proposition.21 This second move therefore seems to deny the initial set up of outlaw emotion cases.

The internalist moves available to a proponent of JV seeking a denial of P2 then, either risk denying basic features of outlaw emotion cases, or result in a drastic restriction of those cases in which outlaw beliefs will remain undefeated (the subset of cases in which the internalist conditions are satisfied). Outlaw emotion cases seem to be best characterized as ones where considerations from the agent’s own point of view favour the oppressive proposition as opposed to the emotion, and hence where the oppressive propositions come out as internally justified. It seems unlikely then that an internalist story will be able to deliver on a denial of P2.

4.2. Denying P2: The Externalist

The discontent housewife believes that her place is in the home and holds a number of beliefs and intentions in line with this, as well as dispositions to feel emotions such as pride when she succeeds at fulfilling this role. What justifies her outlaw belief must be something about how things stand beyond the emoting agent. Indeed, it has been argued that externalism is the preferred epistemological framework for dealing with knowledge acquisition under conditions of oppression, as only facts external to the agent can deliver justified beliefs that go against oppressive ideology Srinivasan (2020).

The external justification move takes the conflicting oppressive beliefs in outlaw emotion cases not to be defeaters on account of being externally unjustified. The thought is that living under conditions of oppression puts one systematically out of touch with truths, and compromises one’s belief-forming capacities relative to a certain domain, such that one is not reliably connected to facts within that domain Srinivasan (2020). Propositions of the sort ‘a women’s place is in the home’ are false and the product of the systematically misleading conditions of oppression. The processes that lead to belief in such propositions are unreliable, making these beliefs unjustified. The conflicting oppressive beliefs therefore cannot, according to this move, act as defeaters against the justification of outlaw beliefs. P2 is denied and the conclusion of my argument against JV is blocked.

Although this is intuitively appealing, the externalist move is actually harder to establish than one might think. First, what seems to be driving the force of this move is a commitment to the falsity of the oppressive propositions, rather than any problem with the mechanism by which belief in them is arrived at. In arriving at the oppressive beliefs, is the agent not treating the evidence available to her adequately, in a fashion that we would think reliably generates true beliefs had the beliefs been true? Is the agent not reasoning correctly from the testimony of otherwise reliable agents, and from her otherwise reliable empirical observation? It seems that the issue is not the reliability of her belief forming processes themselves, but the fact that the beliefs generated tend to be false. This is what gives us reason to think the belief forming processes involved must be unreliable.

This seems to get something right, after all it is false ideology that outlaw emotions are thought to provide evidence against, and people living under conditions of oppression get things systematically wrong by believing all sorts of insidious falsities. But isn’t there a sense in which the proposition ‘a women’s place is in the home’ gets things right about the society in question? Similarly, isn’t there something true about propositions like, for example, ‘women are submissive’ and ‘women shave’ when believed within the context of patriarchal societies? Indeed, isn’t it exactly one of the features of such oppressive societies that stereotypes become self-fulling and self-perpetuating? Such considerations suggest that oppressive propositions can often be both justified and true. This calls into question the claim that oppressed individuals are unreliably connected to facts within the relevant domain. If oppressed individuals are reliably connected to oppressive facts, it seems plausible that, in being formed by the same belief-generating mechanisms, many false oppressive propositions may turn out to be justified despite their falsity. If this is the case, then it would seem that even the externalist attempt at denying P2 may fail, since the conflicting oppressive beliefs may still be justified and act as defeaters against the justification of outlaw beliefs.

Let me say a bit more on this possibility. The prevalence of self-fulling oppressive facts is widely acknowledged, and argued for, in feminist philosophy (Frye 1983; MacKinnon 1987; Langton 2009; Haslanger 2011). MacKinnon, for example, writes that oppressive beliefs ‘become proven, in part because the world actually arranges itself to affirm what the powerful want to see’ (1987: 164). Similarly, for Langton:

When oppression is systematic enough, there is nothing accidental about a correlation between beliefs that women are servile, and women’s servility; and the connection between beliefs and the truth of those beliefs can be as reliable as one could wish. (2009: 303)

The thought is that oppressive propositions can be true and justified. This doesn’t mean that these beliefs are unproblematic, however, just that their failing is not semantic or epistemic. Langton takes self-fulling oppressive beliefs to not only be instrumentally harmful, but to constitute harms themselves. In a similar vein, Haslanger (2011) argues that generics like ‘women are submissive’ can be true, and beliefs based on them justified. Haslanger writes that:

Under conditions of male dominance, women are, in fact, more submissive than men. This is a true generalization and those who live under male dominance are justified in believing it. (2011: 182)

Haslanger locates the problem with oppressive generics not in their semantic truth value, as they adequately capture the way the social world happens to be arranged, but in their propensity to generate beliefs with normative or essentialist content. The descriptive generics should be resisted not for lacking epistemic credentials but because they tend to pragmatically imply, as opposed to logically imply, essentialist and normative propositions which are both false and morally problematic. The descriptive proposition that ‘women are submissive’ might be true in patriarchal societies then, but that women should be, or are by nature, submissive is false. These normative and essentialist oppressive propositions are nonetheless, troublingly, justified, if formed by reliable belief-generating processes.22 For if many oppressive descriptive beliefs are true and justified, and the same belief-generating mechanisms are relied upon in the generation of related, normative, evaluative or essentialist, oppressive beliefs, then these are plausibly justified despite their falsity. This means that Raquel, and the discontent housewife, may be justified in believing normative and evaluative propositions such as ‘it is good for women to be in the home’, and ‘the man’s actions are flattering’ or ‘ok’, whether or not these are true. Once we take seriously the idea that reality is structured by oppression, the externalist move open to proponents of JV seems unlikely to succeed.

In sum, the problem with the externalist move is that it assumes that oppressed agents are unreliably related to truths within a relevant domain. Raquel and the discontent housewife, on the externalist move, are unreliable knowers regarding gender relations. But if, as many feminist philosophers have argued, many oppressive propositions are in fact true, then oppressed agents do not seem to be unreliable knowers regarding the domain in question. Raquel, for example, is likely to hold a wide range of descriptive oppressive beliefs that are both justified and true (including for example the belief that ‘unknown men squeeze women on nights out’). Raquel’s oppressive belief that ‘what the man did was ok’ is an evaluative or normative one however, and is false. But despite its falsity, Raquel’s oppressive belief may still be justified, as the state of affairs in the world provide evidence for it, and this evaluative belief was formed via the same belief-formation process that reliably generated a wide range of true descriptive beliefs regarding gender relations. The evaluative beliefs with which outlaw emotions conflict then, although false, may well be justified. If this is the case, these beliefs can defeat the justification of outlaw beliefs. This leaves P2 standing and JV unable to deny that outlaw beliefs are defeated in outlaw emotion cases.

5. Sketch of an Alternative

I take the central intuition of the Justificatory View to get something right about outlaw emotion cases, namely that outlaw emotions provide justification for evaluative propositions. Outlaw emotions do not just lead indirectly to the (potential) justification of outlaw beliefs by motivating inquiry, as MV would have it. Outlaw emotions seem to themselves be capable of defeasibly justifying evaluative, albeit often vague, beliefs. I have argued that the Justificatory View, as it stands, struggles to account for a satisfying justificatory role in outlaw emotion cases. I will sketch an alternative justificatory view that can deliver on a denial of P2. Given space constraints this will be a rough sketch, but it will, I think, prove the most promising direction to pursue in providing an account of the justificatory role of outlaw emotions.

When we take ideology seriously, I have suggested that it isn’t easy to establish that the would-be defeater in outlaw emotion cases is unjustified. My sketch of an alternative way of denying P2 will not hinge on denying the would-be-defeater its epistemic standing as justified. To begin with, I will assume a form of externalism of the sort arrived at above. An externalism that takes ideological propositions to often be true and justified is one that properly maps a troubling feature of life and knowledge under conditions of oppression. In line with this, and externalism in general, I will assume some form of epistemic reliabilism, where a belief is justified if it is produced by a process that reliably generates true beliefs. According to my view, emotion-based beliefs, as well as beliefs arrived at non-emotionally, via authoritative testimony or deliberation, for example, are reliably formed, as these processes are reliable at generating true beliefs.23 The question is why the conflicting oppressive belief doesn’t defeat the justification of the outlaw belief.

My alternative story for why this is so involves combining the insights of reliabilism and feminist standpoint epistemology.24 Being informed by social epistemology seems prima facie promising in dealing with outlaw emotion cases given that JV and MV failed to incorporate social components into their accounts. Here I sketch how such an account might be developed using insights from standpoint epistemology in particular. A consideration of standpoint epistemology lies beyond the scope of this paper.25 For our purposes we will take the main insight from standpoint epistemology to be the claim that oppressed individuals are in a position of epistemic privilege regarding their own oppression. The claim is not that being a member of an oppressed group is necessary or sufficient to gaining knowledge about the relevant domain of oppression, but rather that oppressed groups are better positioned to gain knowledge about their oppression given their particular social standing.26 This is because social positions make a difference to what reasons agents have, as well as their dispositions to recognize reasons as reasons (Toole 2019).

What seems natural to suggest then, is that one’s social standing makes one a more reliable truth tracker relative to a certain domain. So, women will more reliably track the reality of society’s gender relations, while people of colour will more reliably track truths regarding racial relations. Reliability here is comparative, the oppressed are more reliable, as compared to those who are not oppressed, at tracking truths relative to a relevant domain.27 Kitcher has made a similar point and says that ‘reliabilists should insist that some standpoints are better or worse than others with respect to certain types of propositions’ (1994: 125). 28 This gives us the following reliability claim about social positions:

Standpoint Reliability (SR): Oppressed individuals are more reliable at generating true beliefs regarding particular domain(s) of oppression.

Outlaw emotions are one relevant type of experience that can generate beliefs. Jaggar notes that ‘people who experience conventionally unacceptable, or what I call ‘outlaw’ emotions often are subordinated individuals who pay a disproportionately high price for maintaining the status quo’ (1989: 166). Outlaw emotions are sensitive to reasons that oppressed agents are in a privileged position to access. This is partly because one’s dispositions to experience particular emotions depend on one’s social standing. Social positions will, therefore, dispose agents to be sensitive to particular reasons. One type of belief-generating process by which oppressed individuals are more reliable at generating true beliefs (SR) then, is emotional. As Kitcher says ‘the claim that a particular standpoint is preferable to others can thus be recast in terms of the relative reliability of the processes that different standpoints make available’ (1994: 124). This gives us a second claim:

Standpoint Emotions (SE): Occupying a particular social position makes one more emotionally sensitive to particular reasons.

Outlaw emotions are one of the belief-generating processes made more readily available by occupying a particular social position. This is because doing so will dispose one to greater sensitivity to certain reasons for emotions (SE). Emotions are, however, not the only belief-generating processes that social positions make available. A doxastic version of SE is likely true as well. That is, occupying a particular social position makes one more sensitive to particular epistemic reasons too. It is plausible that members of oppressed groups are also more reliable at generating true beliefs about their oppression (compared to dominant groups) through other means, such as reasoning, for example. Do we have reason to think that emotions constitute a particularly reliable belief-forming process, within agents that occupy privileged social positions? Indeed, we might.

Briefly, emotional processes are typically taken to be informationally encapsulated (de Sousa 1987; Faucher & Tappolet 2006; Majeed 2019). Recalcitrant emotions are a common feature of everyday life, where our emotional experiences seem to ignore, or at least take greater time to respond to, conflicting evidence. This uncontroversial feature of emotions makes them particularly good at picking up on reasons in the presence of conflicting information, beliefs, and emotions. This means that emotions can often aid reasoning by responding to reasons despite what the agent already believes (see Majeed 2019). This seems particularly useful in outlaw emotion cases where the agent holds a number of beliefs that conflict with the outlaw emotion. This informational encapsulation plausibly grants emotions a more reliable connection to true beliefs in outlaw emotion cases, as they are a belief-generating mechanism that is particularly immune to the agent’s wider set of beliefs. This gives us:

Emotional Reliability (ER): Emotions are often more reliable than non-emotional processes at generating true beliefs under conditions of oppression.

Beliefs based on emotions, then, are likely generated by a more informationally encapsulated process than beliefs arrived at through reasoning, or based on testimony. We have seen that one intuitive way of making sense of SR is to cash out the reliability of social positions in terms of the reliability of the belief-generating-processes they make available. Emotions constitute one such belief generating process. This gave us SE, the claim that social positions make us more prone to track particular reasons through emotional experiences. ER states something further, that within the processes made available by a privileged position, emotions may constitute a particularly reliable belief-generating process given that they respond to reasons in a manner that is more encapsulated from an agent’s wider set of beliefs than other belief-generating mechanisms.

The outlaw emotion cases that we have been concerned with involve agents that, according to SR, occupy privileged positions relative to domains of oppression. Discontent Housewife and Harassment both involve women living under conditions of gender oppression. According to SE, they are therefore agents whose social position will have played a role in disposing them to experience outlaw emotions. Furthermore, given ER, their emotions are often more reliable at generating true beliefs than other belief generating processes. The conflicting oppressive beliefs in these cases are, however, also generated by processes available to agents occupying privileged position, after all Raquel and the discontent housewife hold those oppressive beliefs despite being, following SR, in a position of epistemic privilege.

Why doesn’t the oppressive belief defeat the justification of the outlaw belief in such cases? Appeal to ER alone is insufficient in answering this question and delivering a rejection of P2. First, ER doesn’t differentiate between those emotions that go against prevailing ideology and those that do not. Raquel’s conflicting oppressive belief may have been generated by feelings of pride or flattery. If this is the case, Raquel’s oppressive belief and her outlaw belief alike, will be equally reliably formed, and we have made no headway denying that the oppressive belief is a defeater. Furthermore, a denial of P2 based solely on ER would likely say that because emotions are often more reliable under conditions of oppression, they are immune to defeaters. This is an undesirable outcome however, as it would block beliefs formed emotionally from ever being defeated by beliefs that are formed via other processes, reasoning, or perception, for example.

I propose that we need to make reference to the content of the beliefs in question, rather than merely the processes that generate them, in a denial of P2. Specifically, we need to make reference to which social positions beliefs in particular propositions most typically arise in. Following SR, women are more reliable knowers than men regarding gender oppression. So, women are more likely than men to form the outlaw belief that Raquel forms. I propose to deny P2 by appealing to the fact that belief in the conflicting oppressive proposition occurs disproportionately in members of groups that lack epistemic privilege, as compared to those that have epistemic privilege relative to that domain. The thought is that the conflicting beliefs in outlaw emotion cases are not defeaters because they are less reliably connected to truth than the outlaw beliefs. Outlaw beliefs are more reliable because they have propositional content that is more likely to feature in the beliefs of those who occupy a position of epistemic privilege relative to that domain. The claim is that only beliefs that are justified and whose propositional content features disproportionately amongst those that have privilege relative to the relevant domain can defeat the justification of outlaw beliefs.

This is different to the way in which the externalist move considered above attempted to deny P2. The externalist move above denied oppressive beliefs justification. On my sketch the oppressive conflicting proposition is typically justified. The conflicting belief is, however, less reliable than the outlaw belief and therefore cannot defeat it. But this is not due to the belief having been formed by a defective process. This is because the proposition believed is more likely to be shared by agents that are less reliably connected to truth, due to not occupying a position of epistemic privilege relative to the domain of oppression in question. The sketched account is also different from the internalist way of denying P2, where appeal to reasoning is made. On my sketch the agent need not have the ability to herself undermine the would-be defeater, as features external to the agent’s reasoning capacities determine the justification of her beliefs. Structural features of society underlie why oppressive beliefs do not defeat the justification of outlaw beliefs.

Note that, on this account, members of dominant/privileged groups can also experience outlaw emotions that provide immediate justification for outlaw beliefs that remain undefeated by conflicting oppressive beliefs. Membership of an epistemically privileged group is not itself a condition that must be satisfied on my account. Additionally, the account excludes racist or sexist emotions from providing justification to oppressive evaluative propositions. Imagine a scenario where egalitarian principles are dominant, and an agent experiences, for example, an emotion of disgust towards a particular social group that conflicts with their explicitly egalitarian beliefs. In such cases, the conflicting egalitarian beliefs act as defeaters against the justification of beliefs based on this disgust, for they more adequately reflect the beliefs that characterize epistemically privileged positions relative to the relevant domain.29

6. Conclusion

I have argued that existing accounts of the epistemic role of emotions are ill-equipped, as they currently stand, to account for the epistemic role of outlaw emotions,30 and I have provided a sketch that points in what I take to be the most promising direction to remedy this.31 We saw that an adequate account of the motivational epistemic role of outlaw emotions is likely to require inclusion of insights from social epistemology. A modified account of this sort would, however, still find it difficult to deliver on the assumption that outlaw beliefs are justified. JV, on the other hand, seemed particularly well suited to account for a more direct justificatory role than MV allowed, but the view could not block oppressive beliefs from defeating the justification of outlaw beliefs in outlaw emotion cases. I provided an initial sketch of a view that takes steps towards remedying this. For those proponents of JV happy to accept the view I arrive at, my sketched alternative will count as a specific version of their view. For those married to perceptual analogies or internalist commitments, however, my alternative may count as a rival view to theirs.

In conclusion, we have seen that the assumption in the philosophy of emotion literature that existing accounts of the epistemology of emotion are well-equipped to deliver justified outlaw beliefs in outlaw emotion cases faces unacknowledged obstacles. Although I have not argued for a desideratum claim here, if it were a desideratum of any epistemology of emotions that they deliver justified outlaw beliefs, my argument would suggest recommendations for the epistemology of emotions beyond outlaw emotion cases.32 First, it would suggest that reliabilist versions of JV hold greater plausibility than phenomenalist versions. If externalist accounts of JV are better placed to account for outlaw emotion cases, and accounting for these cases is a desideratum of an epistemology of emotions, then we would have reason to favour externalist accounts of the epistemology of emotions in general. Second, I mentioned that JV is most commonly endorsed by perceptual theorists of emotion. The moves to save JV open to a proponent working broadly within a perceptual framework faced serious problems (as seen in Section 4). Appeal to a standpoint epistemology, I have argued, promises to fare better. This move to a social epistemology framework, if held in conjunction with a desideratum claim, would suggest that a departure from perceptual analogies in the epistemology of emotions in general might be called for.

Many details of the proposed positive account remain to be fleshed out, and problems inherited from adopting a reliabilist, and standpoint, epistemology will have to be addressed.33 Even if my alternative proves unsuccessful, I have brought forth an unacknowledged problem in the epistemology of emotion and mapped the main moves that can be made in response to it. This invites proponents of rival, and sympathetic, views alike to make moves towards the provision of an adequate account of the epistemic role of outlaw emotions. The provision of such an account has until now not featured high on the research agenda. I hope to have convinced the reader that it should.


For feedback and discussion, I would like to thank Santiago Echeverri, Lucy O’Brien, Robert Simpson, Giada Fratantonio, Lucy Campell, Charles Jansen, Bence Nanay and his research group at the CPP, the Critical Emotion Theory Network, as well as audiences at Glasgow University and the 94th Joint Session of the Aristotelian Society. I would also like to sincerely thank two anonymous referees at this journal.


  1. On emotional recalcitrance see Brady (2009) and Döring (2015).
  2. I understand oppression as a social injustice, that is to say that it is perpetrated through social institutions, practices, and norms. Certain social groups are systematically and unjustifiably disadvantaged by oppression, while other groups benefit from it.
  3. Although Jaggar (1989) is concerned with outlaw emotions occurring under conditions of oppression, the term seems to encompass a wider class of emotions. Outlaw emotions can occur in agents that have internalized non-oppressive ideology (emotions that conflict with egalitarian values for example), as well as in agents who hold a large set of beliefs we might think aren’t characteristic of a particular ideology (a self-purported and devoted ‘nature-lover’ who feels frustrated and bored in the great outdoors, for example, arguably experiences outlaw emotions). In the present paper I follow the philosophical literature in focusing on outlaw emotions that occur under conditions of oppression.
  4. By ‘internalization’ I mean the psychological phenomenon occurring when a person comes to believe prejudices and biases regarding identity groups, even when these are groups to which they themselves belong (see David & Derthick 2014). Agents that have internalized such beliefs will be disposed to experience emotions in line with them. Outlaw emotions escape this trend.
  5. The thought that emotions are responsive to reasons is widespread in moral philosophy (Skorupski 2010; Raz 2011; Scanlon 2014), feminist philosophy (Frye 1983; Jaggar 1989; Fricker 1991; Lorde 1981; Jones 2003; Bell 2009) and philosophy of emotion (D’Arms & Jacobson 2000; Solomon 2003; Deonna & Teroni 2012b; Tappolet 2016). It is therefore common to think of emotions as amenable to normative assessment, such that some emotions are appropriate while others are not. Emotions are appropriate when there are reasons for them. Assessments of appropriateness equivocate over two distinct types of normative assessment however (D’Arms & Jacobson 2000). An emotion can be appropriate in the sense that it is accurate, or fitting, with respect to an evaluative state of affairs, but the very same emotion can be inappropriate in a moral, or all things considered sense (for example, it might be inappropriate to laugh during an academic talk, even though your friend’s whispered comment was funny). Whether there are reasons for the emotion in the fittingness or accuracy sense is therefore typically separate from whether social or moral considerations count for or against the emotion. In what follows, I am concerned with the fittingness, or accuracy, sense of appropriateness, such that by ‘reasons for emotions’ I mean those considerations that make an emotion fitting, rather than those that might bear on the prudential or moral status of the emotion. The reasons that bear on the fittingness or accuracy of emotions are given by the formal object, or evaluative property, characteristic of each emotion type. Reasons for danger make fear fitting while reasons for offence make anger fitting, for example (see Kenny 1963; Teroni 2007).
  6. Note that outlaw emotions occurring under conditions of oppression can be in fact unfitting. For example, if Raquel was squeezed accidently (in an excusable way, perhaps by a visually impaired man groping around in the dark), her anger would arguably be unfitting despite conflicting with sexist ideology. Similarly, a woman living under gender oppression can feel an outlaw emotion of pride at having behaved in a manner that she follows convention in believing condemnable of her gender, but if this behaviour is not in fact praise-worthy her pride will be unfitting (for example, if a female academic feels pride at having shut down a junior colleague in a public and aggressive manner). My argument, much like those views I critique, does not assume that emotional fittingness is a condition on emotions being able to play epistemic roles (Raquel’s unfitting anger may lead to the justification of a false belief, for example).
  7. I follow others in focusing on propositional justification throughout (Cowan 2016; Carter 2019). For ease of exposition and consideration of example cases, what follows is put in terms that might at times seem to refer to doxastic justification. However, it is propositional justification that we are concerned with throughout. (See also Footnote 9).
  8. Brady (2013) takes the epistemic goal of emotions to be understanding as opposed to knowledge, where understanding involves grasping reasons, connections and causes, while knowledge need not. Brady (2013) is one of the Justificatory View’s strongest opponents, as he argues against the view that emotions provide immediate justification for evaluative beliefs. On Brady’s view, emotions contribute to the justification of evaluative beliefs indirectly, by motivating inquiry that uncovers reasons for emotions. It is these reasons, rather than the emotion itself, that can provide justification for the relevant evaluative beliefs on Brady’s account. Crucially however, Brady thinks that by uncovering reasons for emotions, the agent does not merely gain justification for beliefs (contributing to knowledge), but grasps explanatory and coherence relations characteristic of the more valuable epistemic goal of understanding.
  9. JV claims that emotions provide immediate defeasible justification for evaluative propositions. It should be noted however that most proponents of JV also take emotions to at least sometimes provide doxastic justification for beliefs (Cowan 2016; Tappolet 2016). As propositional justification is typically thought to be required for doxastic justification, focus on the former seems warranted, and we will be concerned with propositional justification throughout.
  10. Most philosophers of emotion take emotions to have evaluative content of some sort, be it conceptual evaluative content for judgement theorists or non-conceptual evaluative content for perceptual theorists. Deonna and Teroni’s (2012b) Attitudinal theory is proposed as deviating from this trend, but it has been argued that even their view does not escape postulating evaluative content of emotions (see Rossi & Tappolet 2019).
  11. What I call the Justificatory View has been called ‘Epistemic Perceptualism’ due to its connection with Perceptual theories of emotion (Cowan 2016; Carter 2019), but has also gone by ‘justification thesis’ (Pelser 2014), the ‘justification view’ (Mitchell 2017), and ‘epistemic dogmatism’ (Brogaard & Elijah 2016).
  12. See Cowan (2016).
  13. Objections have been raised against JV (Brady 2013; Brogaard & Elijah 2016), and responses to them issued (Pelser 2014; Cowan 2016; Mitchell 2017; Tappolet 2018; Carter 2019). JV remains a popular and dominant view. I will not be concerned with defending JV here. I am granting that it enjoys at least moderate success in accounting for the epistemic role of emotions in general. My question is whether the view can, as some of its proponents claim, deliver justified outlaw beliefs in outlaw emotion cases. The problems that I will argue JV faces are distinct from all previous objections that have been raised against the view.
  14. There is an important, indirect, epistemic role that emotions can play under conditions of oppression which will not be my focus. By observing where and when one’s emotions are denied ‘uptake’, or are not taken seriously, as claims about the world, an agent can map their oppression. According to Frye (1983), cartographies of oppression can be developed through such observation. For example: if a woman’s anger is taken seriously when it pertains to issues in the kitchen, or other household concerns, but not taken seriously with regard to political concerns, or in the bedroom, these facts inform the emoting agent about the nature of their oppression.
  15. This is a claim about how the views relate to one another in the literature on the epistemology of emotion. There is of course nothing inherent to the motivational view, as I’ve stated it, that excludes emotions from also playing a justificatory epistemic role.
  16. The prospects of including an internalist counterfactual condition on justification are discussed in Section 4.1.
  17. Indeed, it seems to be the one favoured by Tappolet (2018: 536).
  18. Note that here I am concerned with the justification of mental state defeaters, that is, with whether the oppressive propositions in outlaw emotion cases are justified. Despite mental state defeaters typically being the purview of internalist, as opposed to externalist, epistemologies, endorsement of defeating conditions on justification (that include mental state defeaters) is widespread amongst externalists (Alston 1988; Bergmann 1997; Greco 2010).
  19. Strictly speaking these will be mixed or hybrid internalist-externalist accounts as mental state defeaters are granted. What makes the externalist move externalist is that the justification conferring features that grant mental state defeaters positive epistemic status are externalist. The relevant options of response from a proponent of JV are then either of purely internalist variety, or of hybrid internalist-externalist variety. I refer to the latter as the ‘externalist response’ for simplicity.
  20. This move is analogous to the one proposed by MacIntyre (1990) in arguing for rational akrasia. Put simply, she argues that akrasia can be rational when in acting against our best judgement we are actually acting in accordance with what we have most internal reason to do. Our best judgment, on her account, gets our reasons wrong because we don’t have full internal access to them. Tappolet (2016) adopts such an internalist move to account for emotion-based rational akrasia. It is likely she would endorse an analogous story in the epistemic case. It is, in any case, clear that Tappolet would attempt one of the justificatory moves I outline in Section 4, as she claims that outlaw emotion cases do not involve defeaters because the would-be defeaters are unjustified (see Tappolet 2018: 536).
  21. A similar problem would likely be faced by a mental state internalist attempt to deny that outlaw emotion cases involve defeaters. Mental state internalism holds that whether a proposition is justified depends on mental states the agent has but need not have access to. In outlaw emotion cases the agent has a larger set of mental states that stand in support of the oppressive proposition than of the outlaw emotion (including perceptual experiences, emotions, beliefs based on these, beliefs based on testimony, beliefs arising from reasoning from the oppressive information available etc.). For a mental state internalist move to successfully deny P2 it would have to overcome this.
  22. For Langton oppressive beliefs ‘can be self-verifying or self-fulfilling … if a belief is self-verifying, it provides evidence for itself (given a certain context). If it is self-fulfilling, it makes itself true (again, given a certain context). Beliefs that are self-verifying may be false, but justified: the evidence may be misleading, but being evidence, it supports the belief, even if the belief is false’ (2009: 300).
  23. Any reliabilist account of the epistemology of emotions will have to provide an argument for the reliability of emotions. I do not have space to do so here. See Brun Doğuoğlu, and Kuenzle (2008), Pelser (2014) and Sreenivasan (2018) for arguments to this effect.
  24. That reliabilism and standpoint epistemology are particularly compatible has been noted by Kitcher (1994), Michaelian (2008) and Srinivasan (2020).
  25. On standpoint epistemology see Hartsock (1983), Harding (2004), and Toole (2019).
  26. What specific features of one’s social position grant epistemic privilege? Marxist standpoint epistemologists take women’s contact with the material conditions of reproductive labour to underlie their epistemic privilege regarding gender oppression. Most contemporary standpoint epistemologies have, however, moved away from strict materialist readings centered around labour and ground the epistemic privilege of groups in a wider set of relations they bear to surrounding social conditions and how these structure their experience (Fricker 1999; Dotson 2012; Toole 2019). These features characterize social positions from which standpoints can be more readily achieved. Standpoints themselves are collective identities or consciousnesses that constitute political achievements arrived at through collective struggle.
  27. There are, of course, many ways in which oppressed groups are epistemically disadvantaged. The internalization of oppressive beliefs about one’s own identity group can breed false consciousness and complicity, and oppressed groups suffer hermeneutical injustices (Fricker 2007), whereby a lack of shared resources for social interpretation compromises their ability to make sense of their own oppression (compromising their ability to make sense of their outlaw emotions). These real and widespread epistemic disadvantages are not incompatible with standpoint epistemology’s main insight. It is consistent to grant both that oppressed groups are epistemically oppressed in a myriad of ways, while also granting their epistemic privilege, compared to dominant groups, in being more capable of achieving progressive standpoints with regards to their own oppression.
  28. Kitcher uses the word ‘standpoint’ where I mean ‘social position’, see Footnote 26. I label the claims of my sketched alternative view ‘standpoint’ to indicate that they are informed by a standpoint epistemology, rather than because they refer to standpoints themselves.
  29. What about unfitting outlaw emotions? On my sketch, an unfitting outlaw emotion can provide justification for false evaluative beliefs so long as the conflicting beliefs aren’t more reliably formed. This means that if Raquel was actually squeezed accidentally, by a visually impaired man, her anger might still provide justification to the evaluative belief that what he did was not ok, as this proposition features disproportionately in the beliefs of groups with epistemic privilege relative to a certain domain, and would therefore not be defeated by the conflicting belief. Raquel’s belief would in this case be justified despite its falsity. This shouldn’t give us cause for worry however, as it just means that unfitting outlaw emotions, much like unfitting everyday emotions, can sometimes provide justification for false beliefs. Unfitting emotions of all sorts can, on JV and MV as well, sometimes provide (or uncover) justification for false beliefs.
  30. I have been concerned with whether existing accounts of the epistemology of emotions can deliver on the assumption, present in this literature, that outlaw beliefs are justified. There may of course be further, perhaps crucial, epistemic roles that outlaw emotions play beyond providing (according to JV) or uncovering (according to MV) justification for outlaw beliefs. What these further epistemic roles might be is a topic for future work (see Footnote 14).
  31. I have focused on outlaw emotions occurring under conditions of oppression. What about those outlaw emotions that occur under different conditions? I have suggested above that my sketch looks promising regarding its ability to cope with inverted cases (where oppressive outlaw emotions conflict with egalitarian ideology). Working out the specifics of this is a topic for future work. What about cases that are not directly related to ideology, such as a devoted nature-lover who experiences negative outlaw emotions whenever they are in the great outdoors? I think it is likely that these cases will be more easily accommodated by existing externalist frameworks than cases that involve structural ideology. This is because cases that do not involve ideology will be unlikely to involve self-fulfilling externally justified beliefs with which the outlaw emotion conflicts. Further investigation into these widespread and interesting cases is a topic for future work.
  32. Note that we might not think outlaw emotions call for the same type of epistemic account as non-outlaw emotions, such that endorsement of a desideratum claim does not immediately follow from my argument that existing accounts struggle to accommodate them.
  33. My sketch, much like existing accounts, does not provide an account for how emoting agents themselves might distinguish between fitting and unfitting outlaw emotions. Indeed, we saw that this capacity is likely to be compromised in such cases. Additionally, much like the existing accounts critiqued, my sketch does not distinguish between which token emotions are fitting or unfitting from a third-personal perspective either. I have made use of evocative example cases where the outlaw emotions are fitting throughout the paper, but my proposal, much like those I critique, does not presuppose that emotions must be fitting so as to play epistemic roles (see Footnotes 6 and 29). Thank you to an anonymous referee for prompting me to consider the relevance of veridicality.


Alston, William (1988). An Internalist Externalism. Synthese, 74(3), 265–83.

Arpaly, Nomy (2000). On Acting Rationally against One’s Best Judgment. Ethics, 110(3), 488–513.

Arpaly, Nomy (2018) Comments on Emotions, Values, and Agency by Christine Tappolet. Philosophy and Phenomenological Research, 97(2), 520–24.

Bell, Machalaster (2009). Anger, Virtue, and Oppression. In Tessman, L. (Ed.), Feminist Ethics and Social and Political Philosophy: Theorizing the Non-Ideal (165–83). Springer.

Bergmann, Michael (1997). Internalism, Externalism and the No-defeater Condition 110(3), 399–417.

Brogaard, Berit and Chudnoff, Elijah (2016). Against Emotional Dogmatism. Philosophical Issues, 26(1), 59–77.

Brady, Michael (2009). The Irrationality of Recalcitrant Emotions. Philosophical Studies, 145(3), 413–30.

Brady, Michael (2013). Emotional Insight: The Epistemic Role of Emotional Experience. Oxford University Press.

Brun, Georg, Doğuoğlu, Ulvi, and Kuenzle, Dominique (Eds.) (2008). Epistemology and Emotions. Ashgate.

Campbell, Sue (1994). Being Dismissed: The Politics of Emotional Expression. Hypatia, 9(3), 46–65.

Carter, Adam (2019). Epistemic Perceptualism, Skill and the Regress Problem. Philosophical Studies.

Cowan, Robert (2016). Epistemic Perceptualism and Neo-Sentimentalist Objections. Canadian Journal of Philosophy, 46(1), 59–81.

D’Arms, Justin and Jacobson, Daniel (2000). The Moralistic Fallacy: On the “Appropriateness” of Emotions. Philosophy and Phenomenological Research, 61(1), 65–90.

David, E. J. R. and Derthick, Annie (2014). What Is Internalized Oppression, and So What? In David, E. J. R. (Ed.), Internalized Oppression: The Psychology of Marginalized gGroups (1–30). Springer.

de Sousa, Ronald (1987). The Rationality of Emotion. MIT Press.

de Sousa, Ronald (2007). Truth, Authenticity, and Rationality. Dialectica, 61(3), 323–45.

Deonna, Julien and Teroni, Fabrice (2012a). From Justified Emotions to Justified Evaluative Judgements. Dialogue, 51(1), 55–77.

Deonna, Julien and Teroni, Fabrice (2012b). The Emotions: A Philosophical Introduction. Routledge.

Döring, Sabine (2007). Seeing What to Do: Affective Perception and Rational Motivation. Dialectica, 61(3), 363–394.

Döring, Sabine (2015). What’s Wrong with Recalcitrant Emotions? From Irrationality to Challenge of Agential Identity. Dialectica, 69(3), 381–402.

Dotson, Kristie (2012). A Cautionary Tale: On Limiting Epistemic Oppression. Frontiers: A Journal of Women Studies, 33(1), 24–47.

Faucher, Luc and Tappolet, Christine (2006). Introduction: Modularity and the Nature of Emotions. Canadian Journal of Philosophy, 36(suppl. 1), vii-xxxi.

Fricker, Miranda (1991). Reason and Emotion. Radical Philosophy, 57(Spring), 14–19.

Fricker, Miranda (1999). Epistemic Oppression and Epistemic Privilege. Canadian Journal of Philosophy, 29(suppl.), 191–210.

Fricker, Miranda (2007). Epistemic Injustice: Power and the Ethics of Knowing. Oxford University Press.

Friedman, Marilyn (1986). Autonomy and the Split-level Self. The Southern Journal of Philosophy, 24(1), 19–35.

Frye, Marilyn (1983). The Politics of Reality. Crossing Press.

Greco, John (2010) Achieving Knowledge: A Virtue-Theoretic Account of Epistemic Normativity. Cambridge Core. Cambridge University Press.

Harding, Sandra (2004). A Socially Relevant Philosophy of Science? Resources from Standpoint Theory’s Controversiality. Hypatia, 19(1), 25–47.

Hartsock, Nancy (1983). The Feminist Standpoint: Developing the Ground for a Specifically Feminist Historical Materialism. In Harding, S. and Hintikka, M. B. (Eds.), Discovering Reality: Feminist Perspectives on Epistemology, Metaphysics, Methodology, and Philosophy of Science (283–310). Springer Netherlands.

Haslanger, Sally (2011). Ideology, Generics, and Common Ground. In Witt, C. (Ed.), Feminist Metaphysics (179–207). Springer Netherlands. 10.1007/978-90-481-3783-1_1110.1007/978-90-481-3783-1_11

Jaggar, Alison (1989). Love and Knowledge: Emotion in Feminist Epistemology. Inquiry, 32(2), 151–76.

Johnston, Mark (2001). The Authority of Affect*. Philosophy and Phenomenological Research, 63(1), 181–214.

Jones, Karen (2003). XI. Emotion, Weakness of Will, and the Normative Conception of Agency. Royal Institute of Philosophy Supplement, 52, 181–200.

Kenny, Anthony (1963). Action, Emotion and Will. NY: Humanities Press.

Kitcher, Philip (1994). Contrasting Conceptions of Social Epistemology. In Shmitt, Frederick F. (Ed.), Socializing Epistemology: The Social Dimensions of Knowledge. Rowman & Littlefield.

Langton, Rae (2009). Sexual Solipsism: Philosophical Essays on Pornography and Objectification. Oxford University Press.

Lorde, Audre (1981) The Uses of Anger: Women Responding to Racism. In Sister Outsider (124–33). Crossing Press.

MacIntyre, Alison (1990). Is Akratic Action Always Irrational? In Flanagan, O. and Rorty, A. (Eds.) Identity, Character, Morality (379–400). MIT Press.

MacKinnon, Catherine (1987). Feminism Unmodified Discourses on Life and Law. Harvard University Press.

Majeed, Raamy (2019). What Can Information Encapsulation Tell Us about Emotional Rationality? In Candiotto, L. (Ed.), The Value of Emotions for Knowledge (51–69). Springer International.

Michaelian, Kourken (2008). Privileged Standpoints/Reliable Processes. Hypatia, 23(1), 65–98.

Mitchell, Jonathan (2017). The Epistemology of Emotional Experience. Dialectica, 71(1), 57–84.

Narayan, Uma (1988). Working Together across Difference: Some Considerations on Emotions and Political Practice. Hypatia, 3(2), 31–48.

Pelser, Adam (2014). Emotion, Evaluative Perception, and Epistemic Justification. Oxford University Press.

Pollock, John (1986). Contemporary Theories of Knowledge. Rowman and Littlefield.

Raz, Joseph (2011). From Normativity to Responsibility. Oxford University Press.

Rossi, Mauro and Tappolet, Christine (2019). What Kind of Evaluative States are Emotions? The Attitudinal Theory vs. the Perceptual Theory of Emotions. Canadian Journal of Philosophy, 49(4), 54–563.

Scanlon, Thomas (2014). Being Realistic about Reasons. Oxford University Press.

Skorupski, John (2010). The Domain of Reasons. Oxford University Press.

Solomon, Robert (2003). Not Passion’s Slave: Emotions and Choice. Oxford University Press.

Sreenivasan, Gopal (2018). Emotions, Reasons, and Epistemology. Philosophy and Phenomenological Research, 97(2), 500–506.

Srinivasan, Amia (2020). Radical Externalism. The Philosophical Review, 129(3), 395–431.

Tappolet, Christine (2014). Emotions, Reasons, and Autonomy. In Veltman, A. and Piper, M. C. (Eds.), Autonomy, Oppression and Gender (163–80). Oxford University Press.

Tappolet, Christine (2016). Emotions, Values, and Agency. Oxford University Press.

Tappolet, Christine (2018). Replies. Philosophy and Phenomenological Research, 97(2), 525–37.

Teroni, Fabrice (2007). Emotions and Formal Objects. Dialectica, 61(3), 395–415.

Thompson, Simon (2006). Anger and the Struggle for Justice. In Clarke, S., Hoggett, P., and Thompson, S. (Eds.), Emotion, Politics and Society (123–144). Palgrave Macmillan.

Toole, Briana (2019). From Standpoint Epistemology to Epistemic Oppression. Hypatia, 34(4), 598–618.