Collins, Stephanie (2018). The Government Should Be Ashamed: On the Possibility of Organisations' Emotional Duties. Political Studies 4 (66):813-829.
1. Introduction
Emotional duties are regularly attributed to organisations. Here are some examples.
In 2011, a media investigation exposed the abuse of patients at Winterbourne View hospital in England. Five years later, a parent of one of the patients spoke out: ‘The government have had five years to do something. They haven’t done anything. I’m going to be honest – I think they should be ashamed of themselves’ (BBC, 2016). In response, a National Health Service (NHS) spokesperson claimed an emotional response: ‘We sympathise with the frustrations expressed’ (Bingham, 2016).
In February 2016, Labour leader Jeremy Corbyn addressed the United Kingdom’s University and College Union: ‘We have the highest tuition fees in the industrialised world – it’s not something we should be proud of, it’s something we should be utterly ashamed of as a country’ (Morgan, 2016).
In July 2016, after the United Kingdom voted to leave the European Union, Scottish National Party leader Nicola Sturgeon claimed that the Conservative party had failed to plan for the referendum’s outcome, adding, ‘The Conservatives should feel deeply ashamed of themselves right now’ (Nutt, 2016).
Governments, countries and political parties are not the only addressees. Corporations also face such claims. For example, there exists a Facebook Community, with 91 members, called ‘BP should be ashamed of the mess it made in the gulf’ (Anonymous, 2016).
Perhaps these particular statements are metaphorical, rhetorical, hyperbolic, or a cynical attempt to get the public on-side. They might be ‘bullshit’: something spoken without regard for truth or falsehood, but to further some other aim (Frankfurt, 2005). Nonetheless, if BP were to be ashamed (assuming such a thing were possible), that would strike us as morally appropriate. Moreover, some of these statements – such as that of the Winterbourne View parent – are not bullshit. They express a sincere moral demand.
This article asks whether such demands can ever be legitimate. Can a government have a duty to be ashamed? More generally, can organisations have duties to have emotions? A positive answer requires two claims: first, organisations can have morally valuable emotions; second, they can have duties over those emotions.
These claims might seem bizarre. The first claim might seem bizarre because emotions have a ‘felt’ aspect: they are sensed, subjectively experienced, pleasurable or displeasur able. But organisations lack sensations (subjective experience, pleasure or displeasure). This is possibly what the first Baron Thurlow meant when he famously said, ‘Corporations have neither bodies to be punished, nor souls to be condemned’ (quoted in Coffee, 1981: 386).
The second claim might seem bizarre because duties imply voluntary control. If all entities (human or organisational) lack voluntary control over emotions, then emotional duties are a non-starter (organisations’ emotional duties are rejected, not always with explanation, by Isaacs, 2011: 84–85; Kutz, 2000: 196; O’Neill, 2004: 248; Pettit, 2004: 188; Velasquez, 1983: 124).
This article will argue that organisations can have emotional duties. Specifically, I focus on negative, object-directed, episodic emotions – like shame, remorse and guilt. The argument marries two distinct philosophical literatures – on organisational agency and on emotions – and brings that marriage to bear on the political problem of how to interpret and react to discursive attributions of organisations’ emotional duties.
The first two sections prepare the ground, explaining what organisations are and how they have duties over actions. The subsequent discussion distinguishes the ‘functional’ aspects of emotions (judgements, goals and intentions) from the ‘felt’ aspect (sensations) and argues that both have moral value in organisational contexts. The argument then tackles the two sources of scepticism mentioned above. Here, I argue that organisations, indeed, lack sensations. This is a shortcoming of organisations’ emotional duties. But I argue that they (and their members) have the voluntary control necessary for duties over organisations’ functional aspects.
The argument stakes out a new position among the few theorists who have already married groups to emotions. Some of these deny the importance of emotions’ felt aspect in collective contexts (Björnsson and Hess, 2016; Gilbert, 2002: 119–122; Huebner, 2011; May, 1987: 104–106). Others argue that individuals’ felt aspects can be attributed to groups under certain conditions (Gilbert, 2002: 141–142; Mercer, 2014; Sasley, 2011; Tollefsen, 2006: esp. 234–235, 2008: 11–12).
By arguing for the moral importance of the felt aspect in organisational contexts, yet against attributing feelings to organisations, my argument exposes the moral limits, as well as the moral potential (in the functional aspects), of organisations’ emotions.
2. Organisations
Definitionally, organisations have
‘(a) criteria to establish their boundaries and to distinguish their members from non-members,
(b) principles of sovereignty concerning who is in charge and
(c) chains of command delineating responsibilities within the organization’ (Hodgson, 2007).
Each organisation possesses, and rationally operates, its own decision making procedure, which uses judgement-like and goal-like inputs to produce intention like outputs (List and Pettit, 2011: 21). Crucially for their status as irreducible and distinct entities, the judgement-, goal- and intention-equivalents in organisations are distinct from the judgements, goals and intentions of members. For example, an organisation might hold the profit goal above all else, even though no member does likewise in their personal life. Such divergence of an organisation’s features from those of its members arises because the organisation’s method for arriving at these features is different from that any member uses for arriving at their analogous features. For instance, an organisation might establish its judgements, goals and intentions via a majoritarian or committee-based procedure, while no member uses such a procedure to arrive at his or her own judgements, goals and intentions [see. Goodin, compromise]. Because they arrive at distinct judgements, goals and intentions via distinct procedures, I will assume organisations are distinct agents (List and Pettit, 2011; Wendt, 2004: 296–305).
Organisations can be moral agents. That is, responding to moral reasons can be a goal in an organisation’s procedure. After all, if an organisation’s members are human moral agents, and so can recognise moral reasons, they can (under favourable conditions) design an organisational procedure (1) to which they can present the moral reasons the organisation should respond to and (2) that processes the moral reasons as morality demands of the organisation (Pettit, 2007). So (some) organisations are moral agents and can, therefore, bear duties. I focus on organisations that are moral agents (from now on, simply ‘organisations’).
Entities like ‘the international community’, ‘men’ or ‘the wealthy’ are not organisations, so my argument will not apply to them. But organisations are diverse: they can be more or less democratic or authoritarian, large or small, diverse or homogenous, long term or short term and so on. Ontologically, organisations are singular entities composed of individuals structured in a certain way. This is like how a table is a singular object composed of individual atoms structured in a certain way: if atoms (individuals) cease to be structured table-wise (organisation-wise), then there is no longer a table (organisation). While enough individuals remain thus structured, the organisation can be identified with that instantiated structure (Richie, 2013).
Literature on the politics of emotions has tended to eschew this picture of organisations as unitary agents. For instance, Nussbaum (2011) argues that individual sentiments are necessary for sustaining political values; Sasley (2011) and Mercer (2014) each examine groups’ emotions via the convergent or interacting emotions that individuals experience when consciously identifying as or feeling like the group; Pettigrove and Parsons (2012) theorise the emotions of informal networks. My starting point is not (1) individuals, nor (2) individuals identifying as/like groups, nor (3) networks, but rather (4) organisations as distinct agents. All these perspectives are enlightening: my goal is to supplement, rather than supplant, the other perspectives (I consider individuals’ emotions in the section on ‘Individuals’ Duties’, but only insofar as these are experienced qua formal organisational role-bearer, and without building upon members’ identification as the organisation).
3. Organisations’ Actions and Duties
There are (at least) three arguments for organisations’ actions being non-identical to combinations of members’ actions. This tripartite distinction deserves recognition, because viewing the three arguments alongside one another reveals a compelling case.
The first argument concerns multiple realisability: there are multiple ways distinct organisation-level actions can be realised by individual actions. Consider the United Kingdom’s action of maintaining the NHS. The individual actions that constitute this organisational action could go numerous ways: individuals could work more or less reluc tantly, callously, carefully, obediently and so on. Each of these realisations is ‘The UK maintaining the NHS’, just so long as (1) in each realisation, the individual actions are performed by members within and because of their role in the United Kingdom and (2) the NHS’s maintenance results from each realisation. The NHS’s maintenance exists in each realisation, so we cannot identify it with any one of them. It must be something distinct from, and that unites, them all (Jackson and Pettit, 1992; List and Menzies, 2009; List and Spiekermann, 2013).
The second argument concerns explanation: to adequately explain or describe the members’ actions that constitute the organisation’s action – to capture members’ motiva tions, constraints, goals and so on – we need to refer to the organisation’s distinct features (judgements, goals, intentions and procedures). If one doctor offered medical care by herself, for her own reasons, while off-duty, without the NHS’s resources, then we would not describe this as the NHS’s action. The NHS’s features did not motivate, constrain or guide the doctor’s actions. But if an adequate explanation or description of her actions does require referring to the NHS’s features – she did it because it supports the NHS’ strategic plan or because its processes told her to – then we face pressure to view the doc tor’s action as not merely her own (though it is that too) but also the action of the NHS (List and Spiekermann, 2013: 637–640). The more dependent the action was on the organisation’s features, the more such pressure we face. This is why the Winterbourne View abuse was a problem with the hospital, not just with the individuals within it.
The third argument is perhaps most important. It concerns emergent possibility: organ isations can perform actions their members, taken in aggregate but independently of the organisation’s structure, cannot perform. The NHS could not be maintained by one person alone, nor even many people together who were not structured by an organisation-level decision-making procedure. Of course, the NHS might be maintained by sheer fluke via the actions of many unstructured individuals, but then we would not describe the mainte nance as an action. It would be too unreliable and random. By contrast, if the individuals maintained the NHS via an NHS-targeted organisation-level decision-making procedure, then the maintenance is reliably and intentionally controlled; therefore, an action (Strand, 2012). Yet no individual maintains it on their own: the action must be attributed to the organisation itself.
These three arguments together provide sufficient reason to believe organisations can act in ways non-identical to collections of individual actions. If duties range over actions, then organisations can have their own, organisation-level, duties – for the same three reasons.
First, organisations’ duties are multiply realisable in the duties of members. The precise members that have duties – and what their duties are – vary across possible scenarios in which the organisation’s duty remains constant. Second, the duties members acquire from the organisation’s duty make ineliminable reference to the organisation’s features: plausibly, when an organisation has a duty, the members have duties to act in certain ways within their roles in the organisation (Collins, 2017). Additionally, members’ duties can sometimes be justified only via the organisation’s duty (Wringe, 2016). Third, organisations can have duties that members taken in aggregate do not have. This follows from the fact that organisations can do things their members taken in aggregate cannot do. Individual doctors, managers and so on – considered as a mere aggregate and not under an organisation – could not have a duty to maintain the NHS. This is because they could not do that without the procedure that renders them an organisation. ‘Ought’ implies ‘can’, so since they could not, they ought not. Whereas the United Kingdom can, so, assuming maintaining the NHS is valuable, the United Kingdom ought.
Much more could be said on all these issues. This summary is intended to support, rather than prove, the existence of organisation-level actions and duties over actions. The question is how this extends to emotions.
4. Emotions’ Two Aspects
1 Functional Aspects
Most philosophers of emotion agree that emotions involve judgements, goals and intentions, alongside feelings (De Sousa, 2014).2 I will assume this is correct. To illustrate, take an individual emotion: my remorse overlying to you. This remorse is constituted [???] partly by my judgement that my lie was wrong. It also contains [???] a goal-directed aspect; I have the goal of making amends, or, at least, avoiding similar lies in future. And it includes [???] intentions, for example, the intention to apologise or avoid similar lies in future.
These three aspects admit of a functionalist reading: they can be identified with what causes them and what they cause, within the system that is me operating in the world. Indeed, in the discussion of organisational agency in the ‘Organisations’ section, I assumed such a functionalist account is acceptable. I will continue to assume this (if one believes that judgements (goals and intentions) have a non-functional aspect, then one can include that aspect under the upcoming discussion of emotions’ felt aspects).
Considering only the functional aspects, organisations can have their own emotions – for the three familiar reasons.
First, organisations’ judgements, goals and intentions are multiply realisable in those of members. If the UK Cabinet judges that the Winterbourne View abuse was unjust, it could be that all Cabinet members judge this, or that most do, or that the leader does, or even that no members judge this, but the judgement is entailed by earlier judgements of the Cabinet (List and Pettit, 2011, Part I). These are multiple ways the Cabinet’s judgement might be realised.
Second, Cabinet-level judgements, goals and intentions will sometimes be necessary to adequately explain or describe what is going on with members. This is clearest when members are operating within their ministerial roles. When the ministerial ‘hat’ is on, most of what they intend is intended (at least in part) for reasons that make ineliminable reference to what the Cabinet believes, wants or does.
Third – and most compellingly for irreducibility – organisations can have judgements (or goal or intentions) even when no member has that judgement (or goal or intention). An obvious reason for this is compromise. Imagine the Cabinet is split into two groups: half believe the abuse was unethical but not unjust; half believe it was unjust but not unethical. All agree that ‘wrong’ is halfway between ‘unethical’ and ‘unjust’. So, as a compromise, the organisation takes the position that the abuse was wrong – despite no member hold ing, or coming to hold, a private view on this third possibility.
Thus, the functional aspects of organisations’ emotions are as distinctive as their actions and duties are. For an organisation to have (the functional aspects of), say, remorse, it is not necessary that any member has those aspects. This is just as organisations can have actions and duties no members have. Indeed, members might be entirely indifferent to the objects of the organisation’s remorse-constituting judgements, goals and intentions. This disrupts the natural thought that ‘emotions realized by human beings … are the only way to make sense of corporate affect’ (Tollefsen, 2008: 9).
Much of emotions’ moral value comes from these functional aspects. In humans, the functional aspects direct our attention to morally relevant facts, help us consider responses to those facts, encourage valuable goals and so on (Evans, 2002). This is also true in organisations, where morally worthy judgements, goals and intentions can combine with existing ones to generate new ones. This can snowball into virtuous cycles of belief-, goal- and intention-formation (Björnsson and Hess, 2016). Ultimately, such changes lead to morally valuable organisational behaviour, such as compensation, apology and restruc turing (Wringe, 2012: 441). Often, the judgements, goals and intentions that (partially) constitute emotion – and the actions that follow from them – are all that matter to those who call upon organisations to have emotions. Thus, some collective agency theorists rest content with a functionalist characterisation of groups’ emotions (Björnsson and Hess, 2016; Gilbert, 2002; Huebner, 2011).
2 Felt Aspect
However, philosophers of emotion broadly agree that emotions also comprise feelings (De Sousa, 2014). Whether or not feelings are definitional of emotions, it is true that emotions tend to come with feelings in humans. We should consider feelings’ moral value. If feelings have moral value and if organisations cannot have them, then this will have implications for the number and content of organisations’ duties. These implications are my ultimate interest. For exegetical simplicity, I assume feelings are, indeed, part of emotions.
When I refer to ‘feelings’ or the ‘felt aspect’, I refer to what it is like to be in the emotional state, the state’s subjective character, or the raw sensation of the emotion. In humans, feelings have physical correlates (e.g. changed heart rate, hormone production, neurons firing). Theorists of psychology disagree over whether these physical correlates are identical to feelings (Smart, 2007). If they are identical, then everything I shall say about feelings will also apply to the physical correlates.
While feelings are part of emotions, not all feelings are emotion-feelings. Feelings that are not emotion-feelings are found outside emotions (as an analogy, flour is part of cake, but not all flour is cake flour and non-cake flour is found outside cake). In this way, feelings are a broader category than emotions. For example, feelings arise with moods. When someone is in a mood – happiness or sadness, say – they are not happy or sad about any thing. In contrast, to have the emotion of happiness is to feel happy about something (this ‘something’ is the object of the functional aspects). In this way, ‘feelings’ is a genus, of which ‘emotion feelings’ and ‘mood feelings’ are two species.
To identify feelings, it is best to introspect slowly on examples. Consider the lightness of joy, the blurriness of rage, the pang of remorse, the butterflies of excitement, the bitter ness of envy, the longing of love or the tickling of hilarity. Ryle (1949: 83–84) gave an even longer list: ‘thrills, twinges, pangs, throbs, wrenches, itches, prickings, chills, glows, loads, qualms, hankerings, curdlings, sinkings, tensions, gnawings, and shocks’. Additionally, emotion-feelings (as opposed to mood-feelings) include the ‘comfort or discomfort’ that ‘holds the content [object, e.g., an action you regret] in mind in the rel evant sense’ (Greenspan, 1995: 166).
Of course, examples are not definitions. Indeed, these examples of feelings will aid understanding only if the reader is already familiar with feelings. The impossibility of describing feelings non-circularly arises because feelings are a form of ‘phenomenal con sciousness’. Other examples of phenomenal consciousness (beyond feelings) include the taste of garlic or the appearance of red. The impossibility of describing phenomenal con sciousness – other than by listing examples – is notorious among philosophers of mind. As Block (1993: 241) says, quoting Louis Armstrong’s remark about jazz, ‘If you’ve got to ask [what phenomenal consciousness is], you ain’t never gonna get to know’.
Assuming the above examples enable the reader to identify feelings, we can ask: do emotion-feelings have moral value? I will argue ‘yes’. This argument will focus on individuals’ emotion-feelings because – as I will soon argue – organisations cannot have their own feelings. If the upcoming argument concerned organisations’ emotion-feelings, then it would concern the impossible. But my argument shall focus on individuals’ emotion-feelings within and because of organisational membership. By demonstrating that these particular feelings are valuable, the present argument justifies their inclusion in my later taxonomy.
Consider, then, the 2003 invasion of Iraq. Imagine Blair’s Cabinet had, ex-post, devel oped the functional aspects of remorse over the invasion. The Cabinet publicly stated that (1) the invasion was an error (reflecting a judgement), (2) it would not perform such actions again (a goal), and (3) it would revise fact-checking procedures to ensure this (an intention). Suppose these reports on the Cabinet’s functional features were accurate, lead ing to associated actions, following the virtuous cycles mentioned earlier.
Now consider two possibilities. In one, the Cabinet’s functional states were accompa nied by pangs of remorse among members, when conceptualising themselves as mem bers. In another, there were no such feelings: confronted with those whose lives were destroyed by the invasion, Cabinet members – when acting and thinking within their roles in the Cabinet that authorised this – just did not feel bothered. We are comparing (1) an organisational scenario with functional and felt aspects (the former held by the organisa tion; latter held by individuals qua organisation members), against (2) an organisational scenario with functional aspects only. Which scenario is morally better?
Scenario (1). Why? On a view sceptical of feelings’ intrinsic moral value, perhaps only because feelings are evidence of the judgements, goals and intentions that the individuals, or the Cabinet itself, are likely to have in the future. This is undoubtedly part of the story – but not, I think, the whole story. To see this, suppose our imagined scenarios occur just before the Blair Cabinet disbanded in 2007. Adopt the perspective of someone outside the United Kingdom. Consider those Cabinet members who were not in the subsequent Cabinet. Why should you care about the future judgements, goals and intentions of these soon-to-be-internationally-inconsequential individuals? It would be reasonable for you not to care about these functional upshots of Cabinet members’ feelings, alongside caring whether these individuals have appropriate backward-looking felt reactions about harms their organisation caused.
First, then, a remorse-feeling ‘serves essentially to register the standpoint of the [wronged person]’ (Blum, 1980: 16; Greenspan, 1995: 119). Of course, perhaps judging them wronged or intending to apologise (functional aspects) could achieve similar regis tering – but not quite the same. The difference is like that between judging an artwork to be beautiful and experiencing it as beautiful: the latter registers the beauty with a depth the former lacks (Greenspan, 1995: 113). My claim is that remorse felt by Cabinet members would have similar depth. Indeed, such feeling is perhaps what the Winterbourne View parent was demanding.
Second, remorse-feelings are perhaps the only way of responding appropriately to moral dilemmas (Greenspan, 1995: 118). A non-organisational example will clarify this. In the novel Sophie’s Choice, Sophie ought to give one of her children to the Nazi guard. She ought to have judgements, goals and intentions accordingly. But this right action is also intrinsically wrong: it is a dilemma. It is therefore right that she feels bad about doing so. The feeling attests to the action’s intrinsic wrongness – even alongside conflicting functional aspects and regardless of whether anyone knows about the feeling. Likewise, a Cabinet member may believe that both invasion and non-invasion would have been disastrous; that the invasion decision was a dilemma. A felt pang of remorse over the invasion acknowledges the tragedy of the choice made in the dilemma, even if analogous functional aspects would have been inappropriate.
Third, remorse-feelings are painful. In non-dilemma cases, they, thus, inflict deserved punishment upon oneself (Greenspan, 1995: 133). If a remotely retributive view of pun ishment is correct, then deserved punishment is valuable – regardless of accompanying functional aspects. And insofar as an organisation’s deserved punishments are justifiably distributable to (some) members (Collins, 2016), members’ remorse-feelings are likewise valuable. For example, if BP’s leaders felt remorse for the mess in the Gulf, this would amount to deserved (distributed) punishment.
Fourth, feelings imbue their experiencer with morally significant properties. Again, a non-organisational example will help the initial explanation. Consider a person who feels amused by a sexist joke. Suppose they judge the joke unfunny and successfully aim at not laughing: they lack the functional aspects of amusement (De Sousa, 1987: 290). Such a person is still (somewhat) sexist. They have a morally negative property because of the feeling alone. Likewise, failing to feel remorse, when thinking as a member of a wrongful organisation, makes one (somewhat) morally insensitive, whatever functional features one has
Fifth, feelings are appropriate to the extent they match the facts, so they can be more or less fitting (De Sousa, 1987: 315; D’Arms and Jacobson, 2000; Roberts, 2013: 36). Likewise, feelings are morally fitting to extent they ‘match’ the moral facts, that is, have the same positive/negative valence as those facts. Non-cynically interpreted, Corbyn is imploring UK members to feel as fits the moral facts, when he demands shame over tui tion fees.
Finally, feelings sometimes provide quasi-perceptual acquaintance with the objects of our judgements, goals and intentions, such that our justification for those functional aspects is unmediated. To explain, compare a situation where you learn something (1) by seeing it for yourself versus (2) via testimony from a trustworthy friend. The former gives you unmediated perceptual acquaintance and, thereby, a distinctive kind of justification for your belief. Likewise, when you feel remorseful, you have a distinctive immediate justification for holding the goal of making amends (Johnston, 2001; Roberts, 2013: 49). In organisational contexts, this distinctive feeling-derived justification can be input into the organisation’s procedure to provide distinctive justification for resultant organisa tional goals.
If at least some of these reasons hold up, then emotional duties are not merely duties to perform ‘emotional labour’, understood as ‘the management of feeling to create a pub licly observable facial and bodily display’ (Hochschild, 2012: 7). Instead, they are duties to produce a more profound, unobserved source of value: genuine feelings.
5. A Shortcoming and Two Objections
When arguing for feelings’ moral value, I focused on individuals’ feelings in organisational contexts. This was for a simple reason: there is near-consensus among philosophers of mind that organisations’ feelings are a chimera (Giere, 2006: 315–317; List, in press). Roughly, this is because organisations’ feelings would require ‘a supra-individual experiential unity or some sort of a communal stream of consciousness’ (Szanto, 2013: 109). IR theorists tend to agree. As Hutchison and Bleiker (2014: 49) put it, ‘States, for instance, have no biological mechanisms and thus cannot experience emotions directly’ (similarly Wendt, 2004: 311–314).
Some argue that organisations can, nonetheless, have feelings. For example, perhaps there are feelings that are shared by multiple individuals: one emotion, many individual bearers (Schmid, 2014). However, this is not the same as ‘one emotion, one (group) bearer’, which would be a truly organisation-level feeling.
Or perhaps, if an individual has a feeling because of her membership, then the feeling ‘“belongs” to the group itself’, with the member being the ‘conduit’, even though ‘the group itself does not “feel”’ (Tollefsen, 2006: 235, 236, 2008: 12; Gilbert, 2002: 141–142).
But if it is a member’s feeling that belongs to the group, then we have one feeling, housed in an individual, with two bearers (herself and her group) – not a feeling that belongs to the group alone, able to depart from and explain the feelings of members.
Others argue that groups feel emotions insofar as members cultivate, validate and police emotions among themselves (Mercer, 2014) or self-identify as members of the group while having convergent emotions via this identification (Sasley, 2011). These are closer to genuinely group-level feelings. Here, individuals’ feelings are felt because of membership via identifying with or imaginatively becoming the group.
Such feelings are group level in a significant sense: the group level explains and maintains them. But the group ‘has’ the feeling via the individuals, who each must feel it themselves and in broadly similar ways. By contrast, organisations can judge, want and intend things even while all members are, and remain, wholly indifferent to those things, or when members are (and remain) sharply divided over them. Organisations’ decision-making procedures can produce functional aspects that do not require individuals’ ongoing personal endorse ment. Such group-level conflict with, and departure from, the individual-level is not pos sible with feelings.
Thus, feelings are not irreducibly group-level in quite the same way that the functional aspects are. This is a shortcoming in organisations’ emotional duties.
However, there are two other alleged shortcomings, which should not be accepted. These shortcomings purport to apply to all emotional duties, held by organisations or individuals, over the functional or felt aspects. The first is the ‘control’ objection: duties imply voluntary control over the duty; neither individuals nor organisations voluntarily control their emotions (functional or felt); therefore, emotional duties are impossible (Schroeter, 2006: 346; Sidgwick, 1962: 239). The second is the ‘spontaneity’ objection: even if agents control emotions, only uncontrolled emotions have moral value; there fore, any duty to have an emotion would be self-defeating (if an emotion arises in a controlled, duty-bound way, then the value that gave rise to the duty would not be real ised; Stocker, 1976: 462).
However, there are two other alleged shortcomings, which should not be accepted. These shortcomings purport to apply to all emotional duties, held by organisations or individuals, over the functional or felt aspects. The first is the ‘control’ objection: duties imply voluntary control over the duty; neither individuals nor organisations voluntarily control their emotions (functional or felt); therefore, emotional duties are impossible (Schroeter, 2006: 346; Sidgwick, 1962: 239). The second is the ‘spontaneity’ objection: even if agents control emotions, only uncontrolled emotions have moral value; therefore, any duty to have an emotion would be self-defeating (if an emotion arises in a controlled, duty-bound way, then the value that gave rise to the duty would not be real ised; Stocker, 1976: 462).
First, I will reply to the control objection. In many circumstances, both organisations and individuals can bring themselves to have emotions synchronically, that is, in the circum stance where the emotion is demanded. They can consider the reasons they have to feel some way, downplay the reasons they have against feeling that way (Carver and Scheier, 1998: chs 8–9), or act as if they feel that way (with the aim of producing the feeling; Diefendorff and Gosserand, 2003). This all regards the felt aspect, but it extends to the functional aspect: we have all had the experience of inducing certain judgements, goals and intentions in ourselves, for example, by attending to certain information or to long-term life goals. Individuals and organisations can, thus, control for the functional and felt aspects.
Even if such synchronic control is impossible, agents can produce emotions diachroni cally, that is, can make themselves such that in future circumstances, they are likely to have morally required emotions. Individuals can exercise long-term cultivation of dispo sitions and capacities (Snow, 2015). Organisations can do likewise. For example, they can instigate intra-organisation norms and expectations (Schein, 2010). In addition to reply ing to the control objection, the possibility of diachronic control generates a reply to the spontaneity objection: if the disposition is properly cultivated and maintained, then the emotions arise without value-undermining synchronic calculation.
Thus, the statement (say) ‘you ought to be ashamed’ means either (synchronically) ‘you ought to attend to what you’ve done, or act as if it shames you, or consider your reasons for feeling ashamed …’ or (diachronically) ‘you ought regularly to consider whether you’ve acted shamefully, downplay other demands on your attention, place yourself in environments where your shameful actions can be revealed to you …’.
Individuals and organisations control these actions, which produce both functional and felt aspects (the latter only in agents capable of feelings). What is more, the diachronic actions produce the functions and feelings in a synchronically spontaneous way (Greenspan, 1995: 136–142).
As the above citations suggest, these points have already been made about individual agents. The possibility of applying them to organisations opens space for a more thor oughgoing ethics for organisations, including the possibility of organisation-level cul tivation of virtuous dispositions (the possibility of which is denied by, for example, Cordell, 2017).
6. A Taxonomy
The upshot is that there are more organisations’ emotional duties than one might have thought, although perhaps not as many as one might have hoped. Table 1 depicts a taxonomy of these duties. The duties are pro tanto: they must be balanced against other demands morality places on their bearers and are subject to general limits on moral demandingness.
1 Organisations’ Emotional Duties
Organisations can have two types of emotional duties: duties over their own emotions and duties over the ‘membership emotions’ of their members (explained shortly).
Organisations’ duties over their own emotions are duties only over the functional aspect, since the felt aspect (unlike the functional aspects) requires ongoing housing in individuals. These duties demand actions that direct organisational attention (for synchronic emotion production) or cultivate an organisational disposition (for diachronic emotion production). Broadly, such actions are targeted at changing the organisations’ (synchronic or diachronic, specific or general) judgements, goals and intentions.
We saw an example in the NHS spokesperson’s response to the Winterbourne View parent: ‘we sympathise with the frustrations expressed’. The NHS’s sympathising might involve (inter alia) it committing to the judgement that the parents’ complaints were legitimate, and to the goal and intention of not causing such frustrations in the future. By saying to non-members ‘these are our judgments, goals, and intentions’, the organisation makes it more likely that those judgements, goals and intentions will cause morally appropriate subsidiary judgements, goals and intentions (because the organisation is put ting itself ‘on the hook’ for having them). Of course, the statement was made by a spokesperson – a human individual. But the statement was an organisation-level action nonetheless, because it was multiply realisable, explanatory and emergent: the statement could have been made by a different NHS staff member; NHS-level facts explain the statement; the individual could not meaningfully make the statement if not for the organisation. What is more, for all we know, the spokesperson themselves did not sympathise. The truth of the NHS’s sympathising (in the functional aspects) does not depend upon any individual member sympathising. It depends only on the judgements, goals and intentions that were produced by the organisation’s decision procedure. The statement can, thus, be interpreted as the NHS itself partly discharging its duty over (the functional aspect of) its own remorse emotion.
What about when emotions have moral value via their felt aspect? Here, members can have ‘membership emotions’, with functional and felt aspects. These are emotions that individuals experience within and because of their role in an organisation. These emo tions refer to the organisation and its features (judgements, goals, intentions and proce dures), and are caused by the member’s role in the organisation. These emotions need not build from – though do not preclude – members’ identifying with, or as, the organisation (Mercer, 2014; Sasley, 2011).
Unlike functionings, the group cannot feel how members do not feel. This makes feelings an essentially member-level phenomenon. Thus, membership emotions are housed within – felt by – individual members, not the organisation. The NHS’ statement can be interpreted as purporting the existence of some (functional and felt) membership emotions, alongside the organisation-level functional aspects described above. While membership emotions are not organisations’ emotions per se, they will often (perhaps not always) realise moral value, as the Blair Cabinet example suggested.
In sum, there are morally valuable membership emotions, which organisations can not have duties to experience but that they can have duties to influence. Regarding these, when we say ‘it ought to feel ashamed’ of an organisation, we are saying some thing strictly speaking false. What we really mean is: ‘it ought to use its procedure to make it likely that members will feel ashamed within and because of their role in the organisation’. This is one way to interpret the demand of the Winterbourne View parent. One way for an organisation to, diachronically, ensure it will have such membership emotions is for it to recruit members who are prone to feel membership emotions. Another way is to encourage the development of an emotion-supporting organisational culture (Schein, 2010).
Organisations’ duties over membership emotions may seem objectionable: how can organisations be not just permitted, but obligated, to force or manipulate their long- suffering members to feel emotions? The possible negative implications – exploitation, burn-out, misogyny – loom large (Hochschild, 2012: ix–xiv, 11, passim). But notice two things. First, if an individual voluntarily joins an organisation, she voluntarily commits to having its obligations distribute to her.3 Second, organisations’ duties over membership emotions are not duties to force anyone to experience anything. Instead, the duties might mean drawing members’ attention to certain things, getting them to talk to one another or to third parties, or assigning members the role of engaging with some person or situation. The intra-organisation ethics of such inducements are no more problematic than the intra organisation ethics of organising physical or mental labour to produce a material outcome (indeed, Hochschild (2012: ch. 1) implies this with her analogy between flight attendants and factory workers).
A full account of this intra-organisation ethics would explain, more generally, how organisations’ duties should be realised by members. This includes explaining which members should take on various burdens: realising organisations’ functional emotions, inducing other members’ functional or felt membership emotions, and having one’s own functional or felt membership emotions induced. The simple fact ‘emotions are valua ble’ does not solve this distributional puzzle. I favour a ‘source-tracking’ solution, on which the distribution of the burdens should be sensitive to the ground(s) of the value of the organisation’s emotion (for a defence of this in non-emotional duties, see Collins, 2016). For example, if the demanded emotion is remorse, then the emotion’s value is partly grounded in there having been an organisational wrongdoing. Then, costs should be distributed among members in proportion to their contribution to that wrongdoing. In contrast, if the required emotion is compassion, then the emotion’s value might be grounded simply in the organisation’s ability to respond compassionately. Then, costs should be distributed among members in proportion to their ability.
Social convention partly determines which attributes (e.g. wrongdoing and ability) ground duties over which emotions (e.g. remorse and compassion) in which contexts and exactly which members (e.g. leaders, representatives, rank-and-file members) have those attributes. Thus, which members should face the highest burdens will vary from case to case. Consider Sturgeon’s assertion that the Conservatives should be ashamed because (she claims) they lack a Brexit strategy. Suppose Sturgeon is right. She is evoking the following shame convention: shame is valuable when one has reneged on earlier obligation. The source-tracking model takes this convention and adds that the party’s shame duty therefore distributes primarily to those Conservatives who had (and reneged on) membership duties that were distributed from the party’s earlier obligation to form a Brexit strategy. Who were these people? To answer, we must turn to another social convention: that the duties of the party in government distribute primarily to party members who are in parliament. The source-tracking account takes this convention and gives the verdict that the party’s shame duty distributes primarily to party members in parliament. Importantly, the two mentioned social conventions could have been different, with different results for the moral distribu tion of burdens within the organisation (e.g. we can imagine a shame convention that does not require reneging on obligations, or a convention of parties as strongly member-led, either of which might distribute burdens primarily to members at large.)
This does not imply that the intra-organisational distribution of emotional duties should blindly follow social conventions; still less that all discursive attributions of emotional duties are accurate. There is a difference between how an organisation’s duty should be discharged (which is determined by convention in combination with the substantive moral apparatus of the source-tracking model) and whether it exists in the first place (a matter of morality). Obviously, some emotional duty attributions amount to the exploitation of pre vailing emotion expression conventions for the attributor’s own ends. Corbyn’s and Sturgeon’s statements are arguably like this. So, to assess the existence of an emotional duty, we must assess the proposed emotion for the values described in Section ‘Emotions’ Two Aspects’. If the duty exists, then (under the source-tracking model) conventions partly inform how its costs should be distributed to members. I lack space to defend the source tracking model of distribution here. Alternative models might see burdens justified by members’ voluntarism (French, 1984: 188–190), or distributed relative to members’ authorship (May, 1987: 91–106), or distributed equally (Pasternak, 2011), or relative to members’ capacities (Lawford-Smith, 2012), or relative to a combination of solidarity, prior notice and members’ opportunities for control (Feinberg, 1968).
2 Individuals’ Organisationally Situated Emotional Duties
Organisations’ actions will, at least sometimes, not be enough. Thus, we have indi viduals’ organisationally situated emotional duties. These include duties to have func tional and felt membership emotions. It is morally plausible that, in most circumstances where an organisation has a duty to promote membership emotions, members will have duties to have membership emotions (after all, it is unclear what else would explain the organisation’s duty to promote them). These duties boil down to duties to direct one’s attention, cultivate one’s disposition and so on. If no members of the Blair Cabinet experienced membership remorse – that is, remorse as a membership emotion – then that might be because the organisation has reneged on its duties over felt mem bership emotions, or because members have reneged on their duties over their own felt membership emotions, or both. The latter duties can be interpreted as the distributed components or realisations of the former duties. But the latter have a distinctive char acter: they are directed inwardly at a member’s own individual feelings and function ings, rather than outwardly at the organisation’s functionings or at other members’ feelings and functionings.
Finally, we have individuals’ organisationally situated duties to influence the organisa tion’s emotions. These range over only the functional aspects of the organisation’s emo tion. They will often be duties to present certain inputs into the decision-making procedure, for example, the moral judgement that the Iraq War was wrong. Or they might be a duty to re-design the procedure, such that the organisation from now on acknowledges its past mistakes (as the NHS spokesperson was attempting).
Obviously, there will be various positive and negative feedback loops between the duties I have taxonomised here. The categories are not sharply isolated. The dynamic interactions between duty-types will often render some types superfluous. For example, if members attend to their own membership emotions, then the organisation’s duties over membership emotions become superfluous. Thus, although the proposal might seem overly demanding, it is unlikely to be in practice. In any case, all duties arise only if the relevant emotion aspect would be of sufficient value. The taxonomy here is simply that a conceptual categorisation of the possibilities. It is not an assertion that all organisations and all members have all these duties, all the time.
7. Conclusion: Where Next?
Although a government cannot itself feel morally valuable shame, it can – as a distinct agent with distinct actions, duties, judgements, goals and intentions – have duties over the organisation-level functional aspects of shame and duties to influence and encourage morally valuable shame that its members feel within and because of their membership in the government. Members of a government, in their turn, can have duties to induce mem bership emotions in themselves, where those have functional and felt aspects. Finally, the members can have duties to act on the organisation to induce organisation-level func tional aspects of emotions. There are at least two places where this apparatus could illu minate empirical political research.
The first is analysing the causal-explanatory role of (1) organisations’ (lack of) per ceived morally required emotions and (2) attributions of emotional duties to organisa tions, for example, in electoral or international politics. For instance, take voter mistrust of political elites. One underexplored hypothesis is that such mistrust sometimes arises due to the failure of political organisations (governments, parties, etc.) to fulfil the emo tional duties that voters believe the organisations have. My framework posits that there are two sub-hypotheses to be investigated here: that organisations have failed at organisa tion-level functional aspects, and that they or their members have failed at member-level functional and felt aspects. To test the former, we must assess the organisation’s judge ments, goals and intentions at the bar of public expectations. To test the latter, we must investigate (e.g. through surveys or interviews) the judgements, goals, intentions and feelings that members have within and because of their organisational roles. These are all – at least to some extent – measurable. Both levels matter because feelings are housed only at member level. Thus, my conceptualisation of organisations’ (and members’) emo tions and emotional duties opens the door to systematic measurement of those emotions, which opens the door to analysing the causal-explanatory role of those emotions (or lack thereof) – and analysing the effects that attributions of emotional duties have on those emotions. Such measurement is one place for future research.