The Accessibility of Moral Virtue in the Context of Depressive Episodes

Despite efforts to make virtue-acquisition more accessible, neo-Aristotelian accounts of virtue currently exclude those who occasionally experience depressive episodes from potentially possessing moral virtue. This problem of accessibility is especially relevant given the increased prevalence of depression due to, e.g., the COVID19 pandemic. Through an interdisciplinary analysis, I argue that one’s ability to adequately recognise and respond to virtuous possibilities for action is impoverished during a depressive episode. This is illustrated through the depressed agent’s field of affordances: the collection of possible actions an agent perceives. As moral virtues are conceptualised as stable dispositions, even a temporary yet significant decrease in moral competence excludes the agent from possessing the relevant virtue. To solve this problem of accessibility, I argue that, similar to Aristotle’s discussion of sleep, depressive episodes can be conceptualised as temporary states that cause one’s virtues to be inactive, rather than absent.

After all, there are plausible examples of agents who are, e.g., reliably courageous without being consistently honest (e.g., a political activist who on occasion cheats on exams). If we would not accept that such an agent possesses the virtue of courage even though she lacks honesty, virtue possession would be rare and potentially an idealisation. Moreover, while Aristotle's views on disability, specifically those found in his Politics, seem to exclude those with disabilities from possessing true virtue, contemporary accounts adjust his essentialist view of the norms of species to account for virtue possession in people with disabilities, as well as their ability to flourish (e.g., Garret 2010).
In this paper I argue that such non-idealised and increasingly inclusive accounts of virtue remain exclusionary towards those suffering from depressive episodes, contradicting the contemporary goal of the accessibility of virtue. This is especially relevant given the prevalence of such depressive episodes, its causes ranging from Seasonal Affective Disorder (SAD) and post-natal depression to depression induced due to external circumstance -an increasingly common condition since the global COVID-19 pandemic (Bueno-Notivol et al. 2021;Renaud-Charest et al. 2021;Shevlin et al. 2020).
In Sect. 1 I discuss existent attempts to include depressed agents as potentially virtuous (Foot 2002;Hursthouse 2001). I argue that these attempts are not sufficient for including those suffering from depression as potentially virtuous agents, as depressive states negatively affect moral competence, rather than merely the motivational quality discussed by Foot (2002) and Hursthouse (2001). The depressive state does not merely make it more difficult or effortful for the agent to respond to a possibility for virtuous action; it inherently affects which possibilities for virtuous action are perceived and how they are interpreted. Section 2 conceptualises these possibilities for virtuous action as virtue affordances -the agent-relative possibilities for virtuous action an environment offers -and discusses how they are perceived.
In Sect. 3 I argue for the detrimental effect of depressive moods on the perception of virtue affordances, illustrated through the agent's field of affordances -the collection of affordances the agent is engaged with. Through an interdisciplinary analysis of the effect of depressive moods on affordance perception, I argue that depressive episodes affect the field on all its axes, i.e., the height (how inviting the affordances appear), scope (the number of affordances with which the agent is engaged), and depth (how far into the future some affordances are perceived). From this I conclude that depressive states have a significant negative effect on one's ability to perceive and adequately respond to virtuous possibilities for action.
Given the requirement that the virtues are stable over time and across situations, agents who occasionally experience depressive states are currently excluded from being potentially virtuous -a result that is in opposition to contemporary attempts to increase the accessibility of moral virtue. To solve the problem of accessibility presented in this paper, I compare depressive moods with two other states which compromise one's ability to respond to virtue affordances: being inebriated and being asleep. The latter of which is accepted by Aristotle as a state where one's virtues are inactive in a way that does not negatively reflect on the agent's virtuous character, in contrast to inebriation where the agent is to blame for being in this state. I argue that depressive moods are more akin to being asleep than being inebriated, as in both cases the agent is not blameworthy for being in this state. If we accept this, then experiencing episodes of depression should not exclude the agent from possessing the virtues.

Depression and Moral Virtue
While Aristotle provides the blueprint for contemporary neo-Aristotelian virtue ethics, philosophers working in this field reject aspects of his framework that do not fit with our contemporary values. This includes claims pertaining to slaves and women, as well as the inclusion or definitions of particular virtues. For example, magnificence, which requires the agent to possess and redistribute substantial wealth (NE, 1122a30-35), is rejected to make the virtues inclusive towards those who have little economic security, let alone wealth to redistribute. Moreover, neo-Aristotelian virtue ethicists explore the virtues in the context of updated accounts of psychology (Anscombe, 1958), biology (Foot 2001), and ethology (Hursthouse 1999;, as Aristotle's views on these topics are dated. However, the potential impact of mood disorders on one's ability to be morally virtuous is currently underexplored. The closest the current literature gets to an analysis of depression and its effect on moral virtue entails the discussion of sorrow, provided in response to Kant's example of a philanthropist whose mind "is clouded with sorrow of his own which extinguished all sympathy with the fate of others" (Kant 1996, 4:398) yet who nevertheless performs morally valuable actions through his philanthropy. According to Kant, the actions of this philanthropist are a greater indication of a moral character than moral behaviour motivated by potentially contingent affective states. This seems to fit with our intuitions: someone who, despite her current lack of motivation due to a depressive mood, acts with kindness as she values this trait appears to be equally kind or kinder than the person who acts kind because she feels like it.
Aristotle, however, states that lacking virtuous desires, or experiencing emotions that contradict with these desires, makes the agent continent (enkratic) rather than fully virtuous (NE, 1145b). For example, an agent may have the opportunity to donate 20% of her paycheck to a worthy cause, yet is aware that this money could also be used to purchase a new camera. While she desires the new camera, making her reluctant to donate the money, she reminds herself that she values helping others and should donate the money instead of buying herself an expensive treat. Another agent may come across the same possibility to donate 20% of her paycheck and immediately feels motivated to donate the money. While she may also have treats in mind that her donation would prevent her from buying, she nevertheless donates without reluctance and feels good about doing so. The latter agent's act of generosity is, following Aristotle, more virtuous than the former enkratic act of generosity. Similarly, then, the depressed and thereby unmotivated philanthropist would be less virtuous than the carefree benefactor who takes pleasure in helping others.
Yet prominent neo-Aristotelian virtue ethicists such as Hursthouse (2001) and Foot (2002) argue that Kant's philanthropist may be truly virtuous, despite not experiencing the related virtuous desire. After all, they argue, this lack of desire was due to the depressive state, rather than the agent's character. Moreover, the philanthropist's affective state makes acting on his generosity more difficult, as he no longer experiences an affective push to act on his generosity. As this increased difficulty is not due to his character, the philanthropist rather overcomes a hurdle, which indicates that he is more virtuous, not less, than an agent who does experience an affective push.
The effect of depression on virtue is thereby conceptualised as increased difficulty or experienced effort, where the latter is similar to, for example, helping a friend move house after only 2 h of sleep instead of 8. To illustrate the former, Foot (2002) notes that a courageous act is more difficult when one has to overcome a severe phobia to perform it. Neither the agent's fatigue nor her fear is due to a defect in her character. 1 As such, similar to the depressed agent, these agents show more, rather than less, virtue than those who are not fatigued or do not possess the phobia. Through this interpretation of virtuous motivation, virtue ethics seems more inclusive towards people experiencing depression and other mood disorders.
However, I argue that this is not sufficient for virtue ethics to be inclusive to those who experience depressive moods, as it only accounts for one aspect of depression, i.e., the decreased or absent experienced motivation when confronted with a possibility for virtuous action. To illustrate, in an example presented by Foot (2002) and discussed by Hursthouse (2001), the agent is in a position to show more courage when he comes across an object he is irrationally afraid of. This agent is still aware of what the right action entails and overcomes his phobia to perform it. 2 Yet the decreased moral competence that, as I argue, accompanies depression may make one unable to assess, or even notice, the right action given the situation. Moreover, one may even be unable to recognise a situation as normatively laden, i.e., that her virtuous disposition would be relevant. Depression is thereby not a hurdle that can be overcome through "much virtue" (Hursthouse 2001, 95), as it may rather make an otherwise virtuous agent unable to act in line with virtue. When this effect is applied to the condition that moral virtues are stable character traits, even such a temporary inability to act in line with virtue negates one's possession of the virtue full stop.
The stability requirement commonly refers to situational and temporal stability (Alfano 2013). Situational stability entails the agent's enaction of a virtue over a wide array of situations. For example, if an agent is truly honest, she will not lie to her mother or commit plagiarism. In other words, agents cannot be honest in some aspects of their life and dishonest in others, yet still possess the relevant virtue. Temporal stability denotes a virtue's persistence over time. The temporal stability requirement is discussed in strict terms by Zagzebski (1996), who argues for the implausibility of losing a virtue once it is acquired. She does this through two lines of argument: first, a fully virtuous agent remains virtuous over time as she values the virtues intrinsically, i.e., her positive evaluation of the virtues is not contingent on circumstance. Second, 1 Note that Aristotle would disagree that one can be courageous despite a phobia, as courage specifically entails being afraid of the right objects. If the agent is not blameworthy for this phobia, then she would not be vicious due to cowardice, but she still cannot be courageous (NE 1115b). 2 After all, they merely discuss examples that relate to enkratic or akratic (incontinent) action, i.e., situations where the agent does or does not act virtuously when experiencing a virtuous and non-virtuous desire, rather than propeteia (NE 1150b), where an agent automatically acts in line with passions that oppose virtue, not experiencing a conflicting virtuous desire, yet where her positive evaluation of the virtues makes her regret this action in retrospect.
once an agent acquires a virtue, she does not forget how to act virtuously (Zagzebski 1996: 108-110). So, as a virtuous agent continues to value virtuous behaviour and does not forget how to behave virtuously, the virtues are stable over time. This seems right: we would not ascribe the virtue of, e.g., justice to someone who is only just three weeks a month. 3 So, if a virtue is acquired, ceteris paribus, the agent would be just as virtuous two days, months, or years from now as she is today. Moreover, in relation to both stability requirements, one's virtue must shape one's actions even in challenging situations, such as during a fight or other distressing conditions. 4 A virtuous agent, then, acts virtuously no matter the time or situation; if an agent regularly fails to act virtuously, she would not be considered virtuous. If an episode of depression negatively affects an agent's ability to act on her virtuous values, then, she would not be considered to possess the relevant virtues. This extends to the time where she does act virtuously, i.e., when she is not experiencing a depressive mood, as she does not adhere to the stability requirement. The following sections argue that depression indeed negatively affects moral competence due to its effect on one's ability to perceive and respond to virtuous possibilities for action, or virtue affordances.

Perceiving Affordances
In recent years, psychology, neuroscience and philosophy of mind have become increasingly open to the possibility that cognitive processes, including perception and action, are inherently intertwined with affect (e.g., Duncan and Barrett 2007;Gallagher 2017;Thelen et al. 2001;Woodward 2016). Moreover, the effects of emotional states on perception are well-documented (e.g., Barrett and Bar 2009;Brady 2016). It is therefore not surprising that depressive states would affect virtuous processes as well, as virtuous processes depend on adequately interpreting a normatively laden situation, perceiving a fitting possible action and, in prototypical cases, experiencing the motivation to act on this possibility. This can be illustrated through the interaction of the agent and the world in terms of affordances, focussing on the role affect has in affordance-perception.
Affordances are the possibilities for action the environment offers; the good or bad animal-relative properties it provides (Chemero 2009;Gibson 2015). They are the possibilities for action that an organism could perceive if it was present, based on its physical capabilities, its learned skills, and its needs or concerns. Organisms, including humans, do not perceive all affordances they stand in relation to. After all, a water bottle that affords drinking from also affords throwing, crushing, hiding, balancing on one's forehead, and countless other options that may or may not be perceived. In other words, every situation is laden with countless affordances, from which we only perceive a fraction.
To explain how our attention is directed to some affordances and not others, our affordance-perception can be compared to that of less complex organisms. As every organism has certain needs or goals, it aims to satiate these needs or achieve these goals through responding to the relevant affordances the environment offers. For example, an organism experiencing hunger will attend to affordances which would help it to reduce this negatively valenced state, i.e., the affordance to consume certain (eatable) objects. Likewise, when a cat comes across a bad smell, the resulting negatively valenced affective state directs the cat's attention to the affordance of moving away their nose, which they promptly respond to. Or, when the cat feels threatened, they may perceive the affordance to run away or scratch the threatening object.
A similar process directs our attention to relevant affordances, yet often in a more complex manner than that of an organism needing to eat. The cognitive abilities of humans allow for additional epistemic or mental affordances such as remembering, believing, reasoning and conveying. For example, the fridge may afford memorising what it contains; an affordance that is more likely to be perceived when the agent is hungry, yet may also be perceived in the process of deliberating what to make for dinner. Likewise, our ability to communicate allows us to attend to affordances that are irrelevant to our own experienced physical needs. 5 For example, when someone asks for a cup of tea, one may perceive the affordance that one's kettle offers which would have gone unnoticed without this request. Additionally, our environment is shaped by cultural and social forces and institutions, which affect affordance perception (Van Dijk and Rietveld 2017;Rietveld and Kiverstein 2014). For instance, the bus stop affords waiting for the bus, a field affords playing football, and a ballot affords voting. Moreover, affordances for certain kinds of conduct are perceived or experienced as more attractive than others when they are seen as fitting given the location and situation (e.g., a shop or a dinner party). So, because of socio-cultural practises our environment is socio-material, which shapes the affordances available and attractive to the agent.
Note the intuitive pull of this explanation: when hungry, you will be quick to notice the smell or sight of freshly baked bread or ripe strawberries and when tired, the affordance one's bed offers is quickly perceived and experienced as attractive. Likewise, when scared, you may perceive the affordance to run away and when angry the affordance to shout or to punch a pillow. Similarly, experiencing anger upon witnessing a co-worker being marginalised may direct attention to the possibility to call out this unjust behaviour, while experiencing fear in relation to potentially losing one's job may direct attention to the affordance to look away.
In short, individuals perceive the environment in relation to their needs, goals, and concerns, being drawn to affordances that are affectively experienced as relevant. So, as those who possess the virtues have moral behaviour as a concern, they tend to perceive affordances relevant to these values. For example, a kind person may perceive the affordance of complimenting her friend on her attire and would not perceive the affordance to throw red wine on this outfit. I call these morally laden affordances virtue affordances: possibilities for virtuous action. As with normal affordances, a virtue affordance is a relation between the agent's abilities and the relevant aspects of the environment. For instance, if a child is drowning in a lake, a virtuous agent who can swim would perceive the affordance to swim towards the child and save them. A virtuous agent who cannot swim, however, does not have access to this possibility for action. 6 Rather, she might perceive the affordance of calling attention to the drowning child.
The virtuous agent, then, perceives what would be, e.g., honest, just, or brave in her given situation, informed by her abilities, previous experiences, and other needs or concerns. Note that there is not necessarily only one virtue affordance that ought to be perceived by the virtuous agent. For instance, when encountering a homeless person, a kind agent could perceive the possibility of giving her money, a sandwich, or to start a conversation. These are all kind acts, and a kind person might perceive some, all, or only one of these affordances (Hursthouse 2001). Where one kind agent might be stressed due to a problem at work, therefore quickly handing some coins to the person in need before returning to thoughts of possible solutions, another might be hungry, planning on buying a sandwich, thereby perceiving the affordance to also buy one for the homeless person. Likewise, when relaxed, bored, or lonely, an agent may perceive the affordance to chat with the homeless person, possibly enquiring about what she could do to help. So, which virtue affordances an agent perceives is partly based on which affordances are available given her abilities and the situation, but this perception is also informed by the agent's general affective state.
This can be illustrated through a study by Riener et al. (2011), who explore whether mood alters one's perception of steepness. In their experiment, two groups of subjects estimated the angle of a hill, with one group brought in a sad mood and the other in a positive mood. The results of this experiment suggest that mood influences slant perception: there was a significant increase in the overestimation of the hill's angle by subjects who experienced sadness, while subjects in a positive mood reported a slight decrease in their overestimations. 7 Note that this effect did not occur when subjects were shown depictions of a side view of this hill; increased overestimation only occurred when subjects stood at the bottom of the hill, as if they were about to climb it (while aware that they would not be asked do so). This suggests that moods affect perception of objective characteristics of the environment in relation to one's abilities or potential actions, i.e., that they alter affordance perception.
With this rough overview of affordances and how they are perceived in mind, the following section analyses how the perception of virtue affordances is influenced by episodes of depression. If this analysis is accepted, it should provide sufficient support for the thesis that depressive moods negatively affect an agent's perception of possibilities for virtuous action to such an extent that their virtuous capacities can no longer be considered stable.

Affecting the Moral Field of Affordances
The previous section discussed the role of our skills, concerns, and affective states in directing attention, shaping which affordances we are engaged with. In other words, these elements shape our field of affordances. This field of affordances is constituted by all the affordances an agent is perceptually engaged with at a specific time (Bruineberg and Rietveld 2014;De Haan et al. 2013;Rietveld and Kiverstein 2014). 8 Following De Haan et al. (2013), the field of affordances can be further specified in terms of three axes: the scope (horizontal axis), height (vertical axis), and depth of the field.
The scope of the field represents the affordances an agent perceives or is engaged with. So, when sitting at a desk, an agent may perceive that her computer affords typing and checking social media, a cup of tea affords drinking from, etc. These perceived affordances affect the agent to a certain extent. This is expressed in the height of the field, signifying how relevant or inviting the affordances in the field appear to the agent. For example, when working at her desk, a sudden loud bang affords checking what caused the noise, with this affordance being experienced as very pressing and therefore high in the field. 9 Once a harmless source is detected, the drinkability of the tea presents itself as high in the field as a means to calm down. Lastly, the depth of the field represents the temporal axis upon which affordances present themselves. For instance, after being startled by the loud noise, the agent can respond to the affordance her tea offers in order to respond to the affordance to work that her computer offers. So, the affordance to work is deeper in the field than the affordance to drink tea. Similarly, when the computer affords checking her email, but the agent only plans to perform this task after finishing her work, the possibility to check her email is even deeper in her field of affordances.
In short, the field of affordances illustrates the ways in which an agent perceives the possibilities for action she has access to. In what follows I discuss the impact of depression on affordance perception, focussing on virtue affordances. Each section is dedicated to a different axis of the field: height, scope, and depth respectively. I argue that during episodes of depression, the agent will be unable to reliably 10 perceive and respond to virtue affordances. If this is the case, then people struggling with depression cannot be called virtuous, even if they do reliably act virtuously at other times.

Height
When discussing the effect of depressive episodes on affordance perception, the first intuitive effect relates to the height of field: possibilities for action are experienced as less inviting, and one's emotional responses to the environment seem stifled. Com-8 This is in contrast to all the affordances available to the agent, which is called the landscape of affordances (Rietveld and Kiverstein 2014). 9 Note that affordances need not be consciously perceived to be experienced -from the agent's point of view, she automatically looks towards the direction of the sound. 10 In this context, reliability does not refer to being reliably successful in one's virtuous actions once the relevant affordance is perceived, as discussed by, e.g., Annas (2003) and Stichter (2018). It rather entails whether one is sufficiently stable in adequately perceiving and responding to possibilities for virtuous action.
pare this to an agent who sits at her desk with a bottle of water next to her computer. Whether the agent is thirsty or not, this bottle affords drinking. However, while the non-thirsty agent may perceive this affordance, it will not be experienced as inviting, i.e., as high in her field. Ratcliffe (2014) uses first-person accounts to hypothesise that the affective changes the depressed agent undergoes cause a change in every aspect of experience. Depressed agents sense the world differently, feeling alienated from others and from possibilities to change their situation. Due to this sense of alienation, small acts of kindness performed in social situations are experienced as less inviting for depressed agents. Asking someone how their ill child is feeling or celebrating a friend's accomplishment might still be perceived as possible actions to undertake, yet without the affective push that usually accompanies this perception when the agent is not depressed. Similarly, depressive episodes can make virtue affordances seem more difficult to respond to, as the required behaviour seems exhausting (e.g., Buyukdura et al. 2011).
This increase in perceived difficulty and decrease in how inviting a virtue affordance appears mirror the effects of depression discussed in Sect. 1: Foot (2002) and Hursthouse (2001) take these elements into account to argue that a depressive state tests the agent's virtuous dispositions, akin to external hurdles. The lack of experienced motivation is thereby not reflective of enkrasia, and the depressed agent passing the 'test' by responding to the virtue affordance is rather indicative of a higher level of virtue possession. However, in the remainder of this section I argue that these effects on virtue affordance perception do not capture how a depressed agent interprets these perceived affordances, where the potential misconstrual of these affordances negatively affect the agent's ability to respond to virtue affordances.
Virtuous agents perceive virtuous possibilities for action in response to the normatively laden situation and their abilities (NE 1142a). Yet if a depressive episode makes an agent less capable in adequately estimating, recalling, or employing her abilities, she can lack the insight of how she should, or whether she can, respond to normatively laden situations. That this effect regularly accompanies depressive episodes is supported by the low self-confidence or under-confidence expressed by depressed agents in relation to their abilities (e.g., Stone et al. 2001;Reck et al. 2012;Szu-Ting Fu et al. 2012). Moreover, the comorbidity of depression and anxiety reinforces this claim (Cummings et al. 2014;Kircanski et al. 2017;Lamers et al. 2011). Anxiety tends to entail the belief or feeling that one is unable to cope with (potential) negative events or situations, which results from an interaction between overestimating the danger or negative effects of certain situations or events, and underestimating one's abilities (e.g., Mennuti et al. 2012;Reilly et al. 1999). Moreover, Graydon et al. (2002) demonstrate a relation between experiencing anxiety and underestimating even one's frequently employed physical capacities which, they argue, negatively affects the accurate perception of relevant affordances.
When a virtue affordance is perceived, the agent can dismiss responding to it due to this effect. For instance, when a kind agent sees a crying child, she would usually attempt to help this child by, e.g., cheering them up. When depressed, however, she might still perceive the virtue affordance the crying child offers, yet may deem herself unable to perform this action, thereby not responding to the virtue affordance.
Compare this to an agent who experiences a humiliating event, who then desperately wants to travel back in time to prevent this event from happening. This person might feel affectively invited to this line of action but disregards it, as she knows it is not an available possibility for action. Similarly, the depressed agent who comes across the crying child might experience the very real affordance to help as high in her field, but may write it off as equally impossible as time travel. As such, the affordance is not perceived as more difficult, but as impossible. 11 Similarly, when a depressed agent underestimates her abilities, she may thereby misconstrue related virtue affordances as non-virtuous or even vicious. This can range from an affordance to overcome their fear of public speaking by answering a question in class (enacting courage) to the affordance of responding to a job vacancy for humanitarian work (enacting kindness). They may perceive the affordance of answering the question posited in class, yet as they deem themselves incapable to respond adequately, they may conceptualise this act as a likely cause of embarrassment which would only increase their fear of public speaking. The affordance is thereby perceived as reckless rather than courageous. Likewise, responding to the job vacancy may be quickly dismissed if an agent does not see herself as capable enough to perform this job. Responding to this affordance, then, may be conceptualised as unkind by the agent as she construes sending in an application as wasting the time of the charitable body. Moreover, if she does apply for the job and is successful, she may construe the affordance to accept the job offer as unjust as she believes that there are others who are more capable and therefore more deserving of this job offer. In these cases, even though the agent is aware that the situation they could respond to is normatively laden, they do not perceive the relevant affordances as virtuous due to the decrease in their estimated competence.
Lastly, depressive states may make one more susceptible to respond to affordances that conflict with virtuous values. As Milgram (1974) argues, feelings of insecurity or incompetence strengthen conditions of conformity, an effect that Steger and Kashdan (2009) explain through the heightened sensitivity to social rejection that depressed agents experience, as well as through the discussed negative estimation of their abilities. If we accept this, the depressed agent is more susceptible to the bystander effect, i.e., not helping someone in need due to the inaction of other bystanders, as well as other morally disadvantageous forms of conformity (Hortensius and de Gelder 2018). 12 In short, while virtue affordances may be experienced as less inviting when depressed, thereby affecting the height of the field, this does not merely indicate that the relevant virtue affordances are perceived as more difficult or necessitating more effort to respond to. Rather, they may be dismissed as they are deemed impossible, non-virtuous, or even vicious.

Scope
This section argues that the scope of the field, meaning the number of affordances the agent perceives, is also affected during a depressive episode. In the previous section I introduced the example of a bottle of water that affords drinking, where this drinkability is not experienced as inviting due to the agent's lack of thirst. However, this agent's lack of thirst may inhibit this affordance from being perceived at all, i.e., it may not be included in the agent's field of affordances. Even when this agent aims to drink more water as she believes it to be beneficial for her health, she may simply forget about the bottle of water, or she could be focussed on her work, stopping her from perceiving its drinkability. So, even affordances related to an agent's goals and values may not be perceived due to a lack of affectively experienced motivation. In the case of the bottle of water, its drinkability is not experienced as uninviting or difficult to respond to; it simply is not experienced at all.
To apply this to virtue, recall the reported alienation from others that often accompanies depression. This causes an increased difficulty in forming and maintaining the connections required for friendship, as possibilities for social contact are either not experienced as attractive, or are not perceived at all. For example, a depressed agent might not perceive the affordance of talking to a friend that her phone offers, in the same way that they do not perceive the affordance to stand on it or to put it in the fridge. Especially when the agent is not visually reminded of the possibility of interaction by, e.g., seeing a missed call notification, contact needs to be proactively established by perceiving an affordance that her depression may blind her to. So, even when the depressed agent theoretically knows that calling her friend would be kind (as, e.g., her friend is going through a difficult time, has an important event coming up, or enjoys speaking to her friends), the agent's depressed state can inhibit her from perceiving this affordance.
Moreover, a depressive mood biases a narrow perceptual scope and localised attention (Basso et al. 1996;Brady 2016;Fredrickson and Branigan 2005;Gasper and Clore 2002;Derryberry and Tucker 1994;Schmitz et al. 2009). This can cause the agent to miss affordances she would otherwise perceive, including those related to epistemic or moral virtues (Brady 2016). For instance, a depressed agent may miss a homeless person lifting their hand or a co-worker crying at their desk. As such, unless the agent actively looks around with the goal of virtuous action in mind, virtue affordances she would otherwise respond to might simply not be perceived. 13 Additionally, even active attention guidance to potential virtue affordances may not be sufficient. Recall that depressive episodes often induce under-confidence in one's abilities. In the previous section this effect was applied to perceived virtue affordances, where the depressive mood causes the agent to dismiss the relevant affordance. Yet as affordances are relations between the agent's abilities and relevant characteristics of the situation, it seems plausible that one is unlikely to perceive affordances that are based on abilities the agent does not believe to (sufficiently) possess. 13 Yet recall that perceived virtue affordances may nevertheless be misconstrued.
Recall the student who wants to be courageous by answering a question in class. Rather than perceiving the related affordance as reckless, she may not perceive the affordance at all as she does not believe herself to possess the ability to contribute in class. She may rather sit silently in the back of the classroom, not perceiving the questions asked by the teacher as affording her to respond. This can be applied to other examples as well: if an agent does not believe herself to (sufficiently) possess the relevant skill or ability, she may not perceive the virtue affordances of, e.g., swimming to save a child, calming down someone undergoing a panic attack, or physically protecting a friend, which she would perceive if she did not underestimate her abilities, i.e., when she is not suffering a depressive episode.
Moreover, when the agent experiences psychomotor retardation -a common symptom of depression (e.g., Buyukdura et al. 2011;Mendels and Cochrane, 1968) -both her mental and physical functions slow down. When experiencing this symptom, it is plausible that she is simply unable to consider or perceive how to respond to normatively laden situations that require a quick response -a result applicable to the examples of virtue affordances presented in the previous paragraph. 14 This temporary lack of insight with regards to her abilities and situation may thereby make the otherwise virtuous agent unable to perceive the virtue affordances available to her.
In short, depressive states may stop one from perceiving virtue affordances by making one unaware of the normatively laden qualities of one's situation, or due to a decreased estimation of, or accessibility to, one's abilities. Moreover, the general dampening of one's experience of affective changes makes perceiving virtue affordances less likely, similar to the agent not perceiving the drinkability afforded by her water bottle when she is not thirsty.

Depth
Depressive states also affect the depth of the field of affordances, i.e., the temporal axis. While the introduction of this section illustrated the temporal axis with consecutive actions an agent performs at her desk, the depth of the field is not limited to a temporally extended action sequence. Rather, it can include the affordance a coffee shop offers to meet a friend in a month, or the agent's engagement with the affordance to pay for dinner if she were to work on weekends.
This temporal dimension is shaped by depressive states to such an extent that it affects hospital policies. Studies on the medical decision-making capacity (DMC) of depressed patients regarding their possible treatments suggest that these patients occasionally lack the competence to accurately interpret future possibilities for action. The standard criteria to determine patients' DMC (e.g., Grisso and Appelbaum, 1997) include that patients need to be able to "appreciate the significance of […] one's own situation, especially concerning one's illness and the probable consequences of one's treatment options" (Hindmarch et al. 2013: 1).
Studies analysing the DMC of depressed patients indicate that this ability is impaired in a significant number of patients (e.g., Grisso and Appelbaum 1997;Hindmarch et al. 2013). While these patients seem to understand the procedure and its consequences, they are unable to relate these consequences back to themselves or their own future. Leeman (1999) reaches a similar conclusion, stating that depressed patients, despite having excellent reasoning skills, do not grasp the situation-specific consequences of the treatment, nor the personal implications of these results. These studies thereby indicate that depressed agents have an abstract understanding of their treatment options, yet do not apply the relevant effects to themselves. In other words, while depressed agents may understand the theoretical possibilities for action available to people who undergo the relevant treatment, they do not perceive them as affordances within their field. As such, they may not experience the treatment possibility that would allow them to act on these affordances as inviting. 15 This analysis indicates that a depressed agent who normally acts on virtue might not experience certain affordances as inviting on the (otherwise motivating) basis that they could lead to future virtue affordances, as these future virtue affordances might not be perceived. For instance, a generous agent may value donating to worthwhile charities, which she would be able to do if she would, e.g., switch energy providers or opt for a cheaper broadband deal. However, as her depression makes her less able to apply potential scenarios of generosity to herself, on top of the experienced hopelessness telling her that her situation is unlikely to change (e.g., Abramson et al. 1989;Thimm et al. 2013), she is less likely to make the small changes required to acquire access to the relevant virtue affordance. In other words, due to the decreased depth of her field, potential future virtue affordances related to generosity are not experienced by the agent who otherwise values this virtue.
Additionally, the depressed agent has restricted access to imagining a future state of happiness, experiencing a disconnect from both joyous experiences in the past and possible positive experiences in the future (Beck 1963;Fuchs 2001Fuchs , 2013Habermas et al. 2008;Jacobs 2013;Ratcliffe 2014). Therefore, affordances that the agent does not experience as pressing or attractive in the present are difficult or even impossible to imagine as attractive in the future. So, if virtue affordances are affected with regards to height and scope, as I argue they are, their depth is also affected. Depressive states, then, negatively affect one's perception and interpretation of virtue affordances on all three axes of the field of affordances. This leads agents to be less stable in their otherwise virtuous dispositions.

Virtue's Stability
As noted in Sect. 1, virtues are conceptualised as stable dispositions to act. Therefore, if we accept the analysis from the previous section, it seems that agents who occasionally experience depressive episodes cannot be truly virtuous, even when they possess virtuous dispositions when not depressed. This has some unintuitive consequences. For example, when someone suffers from Seasonal Affective Disorder (SAD), their virtuous disposition is negatively affected due to the reduced exposure to sunlight in winter. Yet should this mean that they are not virtuous during summer? Similarly, when an agent has multiple children, experiencing post-natal depression after every birth, should they be excluded from possessing the virtues despite being reliably virtuous before and after experiencing post-natal depression? Lastly, should someone who reliably responds to virtue affordances and works hard to reduce the likelihood of depressive episodes, but nevertheless occasionally experiences a prolonged bout of depression which diminishes their moral competence, not be considered virtuous? 16 Giving a positive answer to these questions seems unintuitive, especially given the contemporary aim of presenting an accessible account of moral virtue. In what follows I argue that we need not give this positive answer.
As discussed in the first section of this paper, Foot (2002) and Hursthouse (2001) argue that affectively experienced virtuous desires, normally a requirement of virtue possession, need not apply to depressed agents as their lack of such desires is due to their depression, not their character. Similarly, I argue that one's inability to act in line with one's virtuous values during a depressive episode is due to this depressive episode, rather than one's character. To do so, I argue that a temporary inability to act in line with virtue need not mean that the virtues are absent. They can rather be inactive, as Aristotle argues they are during sleep. Second, I note that there are some states, such as intoxication, which also negatively affect virtue affordance perception temporarily, yet where this affected ability is nevertheless due to the agent's character. 17 In other words, if depression is similar to intoxication in this manner, then one's affected behaviour is nevertheless attributable to the agent's character, rather than depression alone. To respond to this concern, I argue that being depressed is more akin to being asleep than to being intoxicated in terms of how it reflects on the agent's character, as in neither sleep nor depression, the agent is blameworthy for being in said state.
First, note the difference between a virtue being inactive and a virtue being absent. When asleep, one is temporarily unable to adequately respond to possibilities for virtuous action and yet one, according to Aristotle (NE 1095b), possesses the virtues one possesses when awake -these virtues are merely inactive. In the latter case, on the other hand, the agent simply does not possess the relevant virtue. In both cases, an agent need not be blameworthy for their inability to adequately respond to virtue affordances. For example, when an agent has not enjoyed a virtuous upbringing and never came across possibilities to acquire the virtues later in life, she is not blameworthy for lacking the virtues and thereby, some may argue, not blameworthy for not responding to virtue affordances. Likewise, as I will argue, one is not to blame for being asleep and therefore for their inaction. The difference between these two cases, is that the former agent's inability to reliably respond to virtue affordances, while not blameworthy, is nevertheless due to her character; while her upbringing is to blame for her lack of, e.g., generosity, this agent not performing generous acts is due to this lack of generosity, i.e., her character.
When we analyse depression along the same lines, the depressed agent's virtue can either be absent, where her depression or other circumstances have prevented her from acquiring a virtue, or the agent has acquired the virtue and adequately responds to virtue affordances because of this virtue, except while depressed. In the latter case, the agent is in a temporary state that prevents her from adequately responding to affordances connected to this virtue. This is thereby akin to an agent who possesses a virtue, responding to the relevant virtue affordances while awake, yet while asleep being unable to do so. When an otherwise virtuous agent experiences an episode of depression, then, her virtue is inactive rather than absent.
However, one may argue that there are other states which impair the agent's ability to respond to virtue affordances, despite reliably acting in line with virtue when not within this state, where this temporary state and thereby the resulting behaviour can nevertheless be attributed to the agent's character. For example, Aristotle notes that being drunk may entail being in a state where one is unable to act in line with virtue, yet that in this case, 'the origin of the offence was in the man himself, as he might have avoided getting drunk' (NE 1113b.20). This seems to be in line with our intuitions. If someone is asleep, they are not responsible for their inaction at that time. Yet if they are drunk or otherwise under the influence of substances that negatively affect their ability to perceive or act on virtue affordances, the agent is, ceteris paribus responsible for their inebriated state and therefore their temporary inability to reliably act on the virtues they reliably act on while sober. 18 In what follows I therefore argue that depression does not share this element with inebriation, i.e., that the state and thereby the resulting behaviour is due to the agent's character. To do so, I argue that this differentiation depends on whether an agent is blameworthy for being in a state which negatively affects their otherwise possessed ability to respond to virtue affordances.
Comparing sleep and inebriation, they mainly seem to differ in terms of the agent's ability to prevent being in said state. After all, if an agent knows that she is unable to adequately act in line with virtue after a couple of drinks, it is prudent to only drink within her limit, or not at all, to prevent possibly risky, unkind, or excessively generous behaviour. In contrast, even when an agent is aware that she may miss 18 While considering the effects of addiction or alcohol-dependency in the context of one's ability to be virtuous is an interesting and pertinent avenue to explore, it falls outside of the scope of this paper. Moreover, where the consumption of the substance was not by choice, such as through a spiked drink, or necessary, such as administered morphine for post-operative pain, the inebriated state is not due to the agent's character. opportunities for virtuous action when she goes to sleep, she also requires sleep to adequately function. While she should consider whether virtue requires her to be awake for longer (e.g., staying with a friend in need after her usual bedtime), her need for sleep requires her to regularly place herself in a state where she cannot respond to virtue affordances.
When we compare this to the depressive states an agent may experience, it seems more akin to sleep than intoxication: by not drinking an agent may avoid being inebriated, while one seems to have less control with regards to falling asleep or experiencing a depressive episode. After all, when an agent suffers from a mood-disorder or is placed in an unexpected distressing situation, as was the case for many during the COVID-19 pandemic, she may not be able to avoid experiencing a depressive episode.
Even when someone may theoretically be able to avoid a depressive episode, but fails to do so, this does not seem to suffice as a reason to deem her blameworthy for her temporary inability to act in line with virtue. For example, if someone is aware of her disposition to experience depressive moods, she could face the choice between going to therapy, which potentially helps her to avoid a depressive episode, or being sufficiently financially stable to know that she can afford to feed herself. Similar to the agent who needs sleep, the agent who decides to feed herself does not seem blameworthy for doing so, despite thereby missing potential virtue affordances.
Likewise, someone who has experienced post-natal depression, therefore being aware of the risks, may become pregnant. If she decides to have another child, whether this is by planning her pregnancy or by deciding against an abortion, is she responsible for her post-natal depression and therefore the effect this has on her moral competence? Blaming depressed agents for their depression in this manner, noting the methods they could have employed to minimise the risk of experiencing this state, seems like a slippery slope.
After all, even when someone is aware that every minute she sleeps puts her at risk of missing some possibilities for virtuous action, we would not expect her to sleep only six hours a day instead of her usual eight. In other words, we do not consider the agent sleeping eight hours a day to be blameworthy for not responding to the virtue affordances she missed while asleep. Failing in avoiding depression, then, would be ill-fitting as an exclusionary condition of virtue possession in a similar way as denying virtue possession to those who require more sleep to feel fully rested.
One could argue that depression is nevertheless more akin to being inebriated as in both states the agent is still able to perceive affordances, which she is not able to do while asleep. Yet this does not seem entirely accurate: we often move around in our sleep to find a more comfortable position, or grab the blanket as comfort during a nightmare or to bring it closer when we are cold. In this scenario, the agent responds to the affordances her blanket and mattress offer in a way that takes their concerns into account and is based on prior beliefs (e.g., the location of the blanket). 19 Additionally, when asleep one can still respond to the affordance a ringing phone offers. Yet instead of waking up to answer, which would entail responding to the potential virtue affordance the phone offers, the agent can rather respond to the affordance of stuffing the phone under her pillow to muffle the sound. In this instance, being asleep or in a sleep-like state stopped her from considering the potential importance of answering the call. Moreover, during REM-sleep, she may even consider the call to be from the nemesis she is fighting within her dream, thereby ignoring the call as she falsely construes it as imprudent to answer. While a kind agent would not even perceive these affordances when awake, as a friend might be calling her in need of support, while asleep or in a sleep-like state these affordances are responded to without the agent being thereby blameworthy for their inaction. In other words, while asleep, or in a sleep-like state, the agent can respond to certain affordances yet, akin to the depressed agent, may misconstrue certain affordances or not perceive them at all. 20 Note, however, that depression does not have to be similar to sleep to determine whether one's inaction during that state negatively reflects on the virtuous status of the agent. This, as I have argued, is rather based on whether the agent is blameworthy for bringing herself into a state that prohibits her from responding to virtue affordances she would otherwise respond to.
Yet there is another element of this comparison between being depressed and being inebriated that must be addressed. Namely, that in both cases, the agent's ability to reflect on the affordances she perceives, choosing to respond to them despite inclinations to do otherwise, may lead them to navigate their normatively laden environment more skilfully, despite their inebriation or depression. In what follows I argue that this potential ability should nevertheless not be employed to argue that the inaction in the face of virtue affordances by the depressed agent is due to their character. First, by arguing that one is not blameworthy for lacking the abilities that allow one to skilfully assess and respond to possibilities for action while depressed. Second, by arguing that lacking these abilities should not entail that one thereby lacks the relevant virtue. Third, by noting that these abilities are unlikely to lead to the reliable perception of virtue affordances in the first place.
The skills needed to reject beliefs or percepts brought about by the depression, and to consider beliefs or percepts that depression tends to suppress instead, are hard to come by. They require access to information, time, effort, and plausibly external help to acquire. As noted, not everyone is able to afford therapy or other services that provide an understanding of how depression affects one's thoughts and perception. The tools that allow an agent to reject depression-induced beliefs that affect her interpretation of virtue affordances (e.g., 'you are unable to φ' or 'none of your friends like you') are thereby simply not available to everyone. Moreover, when an agent is, e.g., unaware of how depression presents itself, that she suffers from a mood disorder that induces depression, that she is currently in a depressive state, or how depression affects her reasoning and perception specifically, she is unlikely to reliably reassess and adjust her interpretation of her abilities and environment. As such, even the virtue affordances that are perceived, yet are perceived as non-virtuous or even vicious, may not be reconsidered as potentially affected by one's depression.
However, while this means that the depressed agent should not be blamed for their temporary inability to adequately engage with perceived virtue affordances, one could argue that the mentioned knowledge and skills to subvert depression-induced beliefs are constitutive of the relevant virtue. 21 As such, the objection goes, the depression and the agent's lack of access to abilities that allow her to regulate this depression rather prohibit the agent's acquisition or possession of the relevant virtue, instead of this virtue being merely inactive. There are however two arguments that reject this conclusion.
First, that such a deep understanding of how depression presents in oneself would have to be included for all agents who would otherwise possess a particular virtue. Imagine an old and kind man who, after the passing of his husband, falls into a deep depression despite never before experiencing such a state. While this man may have theoretical knowledge on how depression tends to affect thought and action, without the relevant experience he never found out what effect this state would have on his perception and beliefs, nor does he have the experience necessary to develop the skills required for subverting depression-induced beliefs. As such, he falls victim to the thought patterns discussed in Sect. 3.2, without possessing the ability to recognise these thoughts as induced by depression, nor the ability to subvert these thoughts. As such, during this depressive state, he no longer responds to the affordances for kindness he would have responded to before his husband's death. If we therefore conclude that this man never possessed the virtue of kindness, given the stability requirement of virtue, then it seems like we ought to include the relevant abilities to possess kindness full stop, or at least for everyone who might theoretically, at some point, become depressed. Given his age, after all, it would have been likely that he passed away before his husband, making it mere chance that he required these abilities to adhere to the temporal stability requirement. However, this would lead to a very odd conclusion. Namely, that only those who have been depressed, were able to go to therapy, and acquired the difficult to develop skills needed to recognise and subvert depression-induced beliefs, can acquire virtue. Including these abilities as necessary for virtue possession thereby contradicts the inclusive and non-idealised characteristics aimed for by contemporary virtue ethicists.
Second, this objection is merely applicable to the affordances that are perceived yet rejected due to false beliefs, i.e., the scope of the field of virtue affordances. Recall that due to the decreased scope and depth of the field of virtue affordances, the depressed agent may simply not perceive certain virtue affordances. 22 As such, while some depressed agents may be able to reason that the virtue affordances they perceive are, in fact, possibilities for virtuous action, rather than non-virtuous or vicious, they 21 After all, it is plausible that this knowledge and these skills are constitutive of practical wisdom. 22 Note that methods to decrease depression's effect on these elements mainly entail decreasing depression itself (e.g., reducing psychomotor retardation or an automatic misevaluation of one's skills). are still less likely to perceive the virtue affordances they would perceive when not experiencing a depressive episode.
A such, virtue ethicists can either accept that depression, similar to sleep, is a state where one's decreased ability to act virtuously is due to a temporary inactivity of the virtues, rather than the absence of them, or they can bite the bullet, arguing that agents whose virtues are not stable over time due to the experience of depressive episodes simply do not possess these virtues. In this section I have argued that the former option seems preferable. After all, Foot's (2002) and Hursthouse's (2001) reason behind including depressed agents as virtuous despite their lack of virtuous desires is applicable to the resulting effect on virtuous action as well: the inability to reliably perceive virtue affordances or to adequately construe them as virtuous possibilities for action is not due to the depressed agent's character, but rather due to the depressive episode.

Conclusion
Contemporary virtue ethicists tend to aim for an inclusive and accessible theory of virtue, moving away from the elitist and impracticable components of the Aristotelian account. This paper argued that, despite these efforts, neo-Aristotelian accounts of moral virtue currently exclude those who suffer occasional depressive episodes from potentially possessing the virtues. This problem of accessibility is especially relevant given the increased prevalence of depression due to, e.g., the COVID19 pandemic, on top of generally pervasive experiences of mood disorders such as SAD and postnatal depression.
Through an interdisciplinary analysis of the effects of depression I argued that one's ability to adequately recognise and respond to virtuous possibilities for action, or virtue affordances, is impoverished during a depressive episode. This was illustrated through depression's effects on all three axes of one's field of affordances. As moral virtues are conceptualised as temporally and situationally stable dispositions, even a temporary decrease in moral competence due to a depressive episode would exclude an agent from possessing the relevant virtue. To solve this problem of accessibility, I argued that, similar to Aristotle's discussion of sleep, depressive episodes can be conceptualised as temporary states that cause one's virtues to be inactive, rather than absent, as the relevant state is not reflective of the agent's character.

Declarations
Conflicts of interest I hereby confirm that this is an original article which has not been previously published, nor is it under consideration to be published elsewhere. Moreover, there are no conflicts of interest to disclose.
to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/ licenses/by/4.0/.