Petersson, Björn (2020). Collective Guilt Feelings. In Saba Bazargan-Forward & Deborah Perron Tollefsen, Routledge Handbook of Collective Responsibility. Routledge.
16.1 Introduction
In Complicity (2001), Christopher Kutz admits that there are important differences between assigning responsibility to a group in a holistic manner and summing up judgments about individual responsibility. He nevertheless denies that a group in itself can literally be culpable. One of Kutz’s reasons for refraining from saying that collectives can be guilty is that
a collective cannot respond affectively to these expressions, only its constituent members can. The lack of affective counter- response is troubling, because the efficacy of responses of accountability partially depends upon affect. The responses of shame, guilt, and regret help to register the significance of the harm. (Kutz 2001: 196)
The view that evoking feelings of shame, guilt, and related social affective attitudes is essential to our practices of blaming and holding groups morally responsible is common, at least within a broadly Humean or Strawsonian tradition (Hume 1740; Strawson 1962). And the claim that groups can feel guilt, or indeed feel anything at all, appears to “stretch phenomenological credibility” (Smith 2008: 241). So, Kutz’s worry seems legitimate.
There is a relatively rich literature on the social psychology of collective guilt and on how different kinds and degrees of experiences of collective guilt affect and are affected by the degree of group identification, and by other ingroup/ outgroup attitudes and behavior.
Social psychologists regard the feeling of collective guilt as an existing phenomenon; one which is available for empirical studies. This may seem to undermine Kutz’s objection. However, even in the context of empirical research it is still not clear exactly what assumptions research subjects commit themselves to when they express feelings of collective guilt.
The studied phenomenon is social and hence collective in the broadest sense of the word – it is a feeling that requires an us/ them categorization and it explains attitudes that are directed towards one’s own group or other groups. However, it is unclear whether it is collective in the sense required to mitigate Kutz’s worry. Empirical studies in this area typically focus on individual experiences and perceptions, and as Ferguson and Branscombe note in a research review, “it is not clear whether such measures actually assess collective guilt” (2014: 259). In other words, it is not clear whether empirical research supports the idea that groups as such can react with the kind of affective response that Kutz finds essential for assigning meaningful responsibility to them.
Section 16.2 is a brief elaboration of Kutz’s assumption that a subject’s capacity for guilt feelings is a reasonable condition for holding it morally responsible in a meaningful way. Section 16.3 maps current defenses of the possibility of collective guilt into two cat egories: positions that assign guilt feelings to groups as such but play down the phenom enological or experiential component in guilt feelings, and positions that do justice to our intuitions about the phenomenology of guilt feelings but understand collective guilt feelings in terms of individual experiences. Section 16.4 focuses on two examples of the f irst type of approach, by examining the analogy between collective and individual guilt from two different collectivistic viewpoints – Gilbert’s plural subject theory of group guilt, and the suggestion from Gunnar Björnsson and Kendy M. Hess that standard functionalist arguments for basic corporate agency extend to reactive attitudes like guilt feelings. The fourth section presents an approach of the second kind: an individualistic but “perspec tival” understanding of collective guilt, related to the “we-mode” approach to collective intentionality.
16.2 Blame and Guilt Feelings
Kutz’s objection to collective responsibility rests on the assumption that evoking feelings of shame, guilt, and regret is an essential function of moral blame in general. It is beyond the scope of this chapter to defend this assumption, but like David Hume, John Stuart Mill, Richard Brandt, Alan Gibbard, and others, I will assume that it is basically correct.
Hume emphasizes that moral feelings, i.e. “pleasure and pain of that peculiar kind, which makes us praise or condemn” are social in character, and differ radically from feelings we have towards inanimate objects. They include love and hatred, pride, and humility (1739, book III, section 1:2). These are feelings of the sort that are bound up with “the very great importance that we attach to the attitudes and intentions towards us of other human beings, and the great extent to which our personal feelings and reactions depend upon, or involve, our beliefs about these attitudes and intentions” (Strawson 1962: section 3). Like Strawson, Hume understands moral praise and blame in terms of interpersonal feelings that typically involve a mutual affective concern – a desire or expectation that the other cares about my attitudes towards her, and that she has a similar desire for me to care about her attitudes towards me.
When John Stuart Mill characterizes morality as a system of social sanctions, the type of sanctions he treats as most important are self- reproach and feelings of guilt.
For the truth is, that the idea of penal sanctions, which is the essence of law, enters not only into the conception of injustice, but in that of any kind of wrong. We do not call anything wrong, unless we mean to imply that a person ought to be punished in some way or other for doing it; if not by law, by the opinion of his fellow creatures; if not by opinion, by the reproaches of his own conscience. (Mill 1863: 33)
According to Richard Brandt, the essential element in the moral responsibility of a person X is that it is fitting or justified in itself for X to have some blaming attitudes, including … remorse towards himself, and for many other persons Y to have some blaming attitudes including retributive indignation towards X, and to express them in their behavior. (Brandt 1958: 16–17)
Alan Gibbard claims that “[t] o hold a person to blame for an action will then be to accept norms that tell him to feel guilty for having done it, and tell others to be angry with him for having done it” (Gibbard 1990: 150)
In line with this, I will assume that to hold you responsible for some bad event is to have an attitude with a conative element towards you: The wish that you feel guilt for your involve ment in the event. When you feel guilt for what you did, you feel bad about your actions and wish them undone, due to your demand for regard from others. (As Mill notes, this demand may be completely internalized and it does not necessarily require actual social interaction in each case.) Blame typically appeals to your demand for goodwill from others and your sense of having been part of the explanation of why something bad happened.
On the face of it, we blame the deceased, we blame psychopaths who allegedly lack the capacity to feel guilt,4 and we even appear to blame inanimate objects like malfunctioning computers. On the view assumed here, it appears that we must dismiss such behaviors as non sensical, since the blamer in such cases aims at a reaction she knows cannot be realized. A less dismissive strategy is to regard them as distinct forms of blaming- like practices, whose meaning in various ways are parasitic upon the basic sense of blame, sharing some but not all of its essen tial characteristics. The latter strategy is compatible with holding on to the idea that blame in that basic sense still occupies the most central “region in our moral thought” (Gibbard 1990: 52), and hence that the question of whether a certain type of being is capable of feeling guilt is of great moral relevance.
As Mill notes, guilt feelings may be regarded as negative moral sanctions, closely related to other forms of punishment. Although guilt feelings may have valuable functions, such as helping ameliorate damaged relations, prompting personal improvement etc., like other forms of pun ishment they are in themselves generally unwanted and unpleasant. This becomes obvious when we consider cases where guilt feelings occur but lack the usual positive effects, such as when people cannot help but feel strongly guilty about some innocent choice that by coincidence turned out to have disastrous consequences.5 Such feelings are generally considered as bad for the person, and clearly constitute a form of suffering. The unpleasant quality of guilt feelings appears essential to their function as negative moral sanctions.
16.3 Guilt Feelings and Phenomenology
The view that groups can have their own beliefs, desires, and intentions has been defended in different ways by Philip Pettit, Margaret Gilbert, and others (Gilbert 2000; Pettit 2003). This does not mean that they would be prepared to assign phenomenal consciousness to groups. Christian von Scheve’s and Mikko Salmela’s anthology on various perspectives on collective emotions (2014) illustrates the fairly broad philosophical and scientific consensus on phenom enology being absent in groups.
Like Burleigh Wilkins, most would probably find the idea of guilt feelings with a “total absence of any phenomenological accompaniments… extremely puzzling” (Wilkins 2002: 152). While a more technical term like “emotion” has been given non- phenomenological interpret ations, at least in the philosophical literature – as referring to an essentially conative state or a type of evaluative judgment, for instance – the word “feeling” typically refers to something felt by a subject, i.e. an occurrent subjective experience of a particular kind, or at least a disposition for such experiences. Non-phenomenological accounts of emotions are usually explicitly contrasted with feeling accounts, thereby implying that feelings, unlike emotions, by definition come with phenomenology. So, on the common use of the word “feeling,” “feelings without phenomenology” is a contradiction in terms
1. one natural move among philosophers defending collective guilt feelings is to question the essentiality of phenomenology in this context.
① Gilbert explicitly doubts that some phenomenological condition must be met for someone to have guilt feelings (Gilbert 2002: 141).
② Thomas Szanto argues that corporations can have negative reactive emotions, such as feelings of humiliation, which “are not individuated by their phenomenology” and Deborah Tollefsen makes a similar point (Szanto 2016: 271– 2; Tollefsen 2003: 232).
③ In their functionalist defense of corporate reactive attitudes, among them guilt feelings, Gunnar Björnsson and Kendy M. Hess “fail to see why purely qualitative aspects of a phenomenal point of view would matter” (2017: 282).
2. The natural alternative to stripping the notion of “feeling” (in “collective guilt feelings”) of phenomenological connotations will be to reinterpret the notion of collectivity in a way that makes it possible to assign the feeling in question to individual group members rather than to the group as such.
① This appears to be in line with how the term is mostly used in empirical investigations or case studies of collective guilt feelings (see e.g. Pettigrove and Parsons 2012).
② Versions of this strategy have been defended, for example, by Stephanie Collins, who interprets normative ordinary language claims about organizational emotions in terms of the organization’s duties to promote said emotions in their members (Collins 2018)
③ and by Anita Konzelmann Ziv, who accounts for “collective guilt feelings in terms of individual members’ we- feeling of guilt” (Konzelmann Ziv 2007).
In other words, given that no reasonable approach assigns phenomenal consciousness to collectives as such, we need to stretch the meaning of “collective guilt feelings” beyond the most natural reading, 1. either by allowing for “feelings” without experiential components, or 2. by assigning a “collective” attitude to individuals.
If we allow for the first strategy, it seems reasonable to require at least that the functional analogy between individual and collective guilt is very strong, phenomenal differences aside. If we allow for the second, the challenge is to give an account of what it means for an individual attitude to be collective in a substantive and interesting sense.
16.4 Joint Commitment to Feel Guilt
According to Margaret Gilbert, there may be a radical disjunction between the intentions and beliefs of a group, and those of their members.
This is the core of her account: “A population P has a collective intention to do A if and only if the members of P are jointly committed to intending as a body to do A” (2002: 125, also Gilbert 2000, 2006 and 2009).
A population fulfilling this condition constitutes a plural subject of the intention to do A. A joint commitment is not a set of individual commitments, it is not created by individual decisions, and it “does not have parts” although it has implications for the parties.
On the plural subject account, groups can have intentions of their own, and Gilbert argues that they can have moral beliefs. Therefore, groups as such can be guilty of performing wrongful acts. Gilbert questions Kutz’s claim that collectives cannot respond affectively to moral blame in the appropriate way. “In particular, I argue that there is an important sense in which a collective can feel guilt” (2002: 117).
On the plural subject account of collective action, it is clear that groups as such can be guilty of wrongdoing, and that the group’s guilt may have no implications for the guilt of its members (Gilbert 2000, section 8.10).
Gilbert argues plausibly that the individual member’s feeling of guilt in such a case may be different from ordinary feelings of guilt over the member’s own individual actions: the object of the member’s guilt feeling is the collective action of the group in which she is a member, and such feelings may not be proportionate to the member’s assessment of her personal contribution to the collective wrongdoing. As Gilbert notes, a “problem with rejecting such feelings is that people seem to have them” (2002: 134). We might ask whether they can be rationally justified – Gilbert quotes Karl Jaspers’ description of the predicament of being unable to rid himself of guilt feelings for what other Germans did during the Second World War, while as a philosopher he found such emotions to be rationally refutable – but for the present purposes it is sufficient to accept that we can have such feelings, feelings that we have in virtue of our membership in a collective. Gilbert calls them “membership guilt feelings” and we might imagine other sorts of membership feelings, like membership sadness over a collective loss, or membership pride over a collective achievement (Gilbert 2002; Jaspers 1947).
Still, the occurrence of membership feelings of guilt does not imply any feelings of the group as such, or vice versa.
For genuinely collective guilt feelings to occur, it is neither necessary nor sufficient for members of the group to feel membership guilt over an act of the collective… neither necessary nor sufficient for members to feel personal guilt over their participation or other act relating to the collective act… [W] hat is needed, to put it abstractly, is expressions of readiness on everyone’s part to be jointly committed to feel guilt as a body. (Gilbert 2002: 140)
Gilbert’s general plural subject account is formulated in terms of a joint commitment to intend as a body to do something. From such a commitment follows individual entitlements to actions from other members, and personal obligations to act so as to constitute with others an entity that acts “as a body.”
The language of commitments seems more appropriate for capturing collective agency than collective emotions, for the simple reason that while we commit ourselves to act in various ways, we do not normally seem to make commitments to feel anything.
In later work, Gilbert articulates the core idea by saying that “roughly, the parties are jointly committed as far as possible to emulate, by virtue of the actions of each, a single body that intends to do the thing in question” (2009: 180). The reference to commitments to act so as to constitute a body that does something is still essential in this formulation.
We do not make decisions about what to feel, and more generally, we do not seem to have the capacity for voluntary shifts of attitudinal modes. I do not deliberately switch from fearing that p to hoping that p, grieving that p, or feeling guilty that p. Admittedly, there are pre- commitment devices, more or less efficient self- help methods, and other ways of attempting to affect one’s emotions and motivation but those sorts of activities are not what the parties to the joint commitment to feel guilt as a body are committed to in Gilbert’s examples of such cases.
In her illustrations of collective guilt feelings, the act of expressing guilt feelings as a body is really what the parties to the joint commitment are committed to.
When Joe and Lisa have failed to look after Phyllis’ daughter Mary properly and as promised during the weekend, thereby making Mary depressed, a joint commitment between Joe and Lisa is prompted by Lisa’s expressing their guilt feelings to Phyllis, in the presence of John. This results in a commitment “to feel guilt as a body” but the function of that commitment is e.g. to make Joe “feel constrained to do and say things that echo or conform to Lisa’s claim that she and Joe feel guilty about the way they treated Mary” (2002: 140). Certainly, this type of situation occurs, but note that these conditions may be completely fulfilled – Joe and Lisa may both be constrained by a joint commitment to do and say things that are consistent with feeling guilty as a body – without any of them actually feeling guilty about anything.
Gilbert explicitly stresses the latter possibility: “No one of these feelings [i.e. collective, membership, and personal guilt feelings] seems to carry another with it as a matter of logic” (2002: 142). On Gilbert’s notion of collective guilt feelings, their collective behavior would suffice for assigning such feelings to this group, even if we know for certain that neither Joe nor Lisa cares about Mary or Phyllis, and that their initial expressions of readiness to enter the joint commitment to do the things that are significant of collective guilt is fully explained by, say, their desire to conform with socially accepted behavior.
Gilbert explicitly questions the idea that some phenomenological condition must be met for someone to have guilt feelings. Let us grant that the apparent absence of an independent phenomenology of the plural subject may, as Gilbert claims, be “no issue” for an account of collective guilt feelings (2002: 141).
My remaining worry is that nothing in the account gives us any reason to believe that the plural subject as such would care about being in this state. Joe and Lisa may each wish that they were not bound by a joint commitment to express guilt feelings as a body, but the account assigns no such wishes to the plural subject, which is distinct from Joe and Lisa. To assign such wishes to the plural subject would seem to require that yet another joint commitment comes into play – a commitment to express resentment over the first commitment, perhaps. This seems farfetched, and would clearly be an ad hoc move. So, I am inclined to think that the plural subject’s Gilbertian guilt feelings lack the element of being unpleasant and unwanted that is essential to guilt feelings and their functions in relation to blame.
My contention is that while Gilbert sketches an interesting account of how members in a group may feel obliged to express certain emotions on behalf of the group, and of how in such situations collectively tainted emotions may occur in the minds of individual members, the properties that she assigns to the plural subject – the group as such – do not constitute guilt feelings in the sense that is tied to moral blame. Nevertheless, I think that elements in her reflections on the phenomenology of collective guilt feelings point to an interesting sense in which individuals may have distinctively collective feelings of guilt. I will return to that issue in the last section.
16.5 Corporate Agency and Reactive Attitudes
In “Corporate Crocodile Tears? On the Reactive Attitudes of Corporate Agents” (2017), Gunnar Björnsson and Kendy M. Hess challenge anyone who finds it reasonable to regard corporations as agents in a basic sense but refuses to assign feelings of guilt, sadness, or indig nation to them. The challenge is to “explain why the sort of arguments that support basic corporate agency do not extend to reactive attitudes” (2017: 293). Elsewhere in the text, they state their conditional argument more cautiously and merely claim that functionalist arguments for regarding corporations as agents will also show that corporations can have “the moral equivalent” of reactive attitudes or something “sufficiently similar” to reactive attitudes (2017: 274).
The strategy of Björnsson and Hess, as I understand it, is to establish, first, that if we accept functionalist arguments to the effect that corporations have basic agency, then we should also accept that they can be in states that are functionally equivalent to, for example, individual reactive attitudes like guilt feelings.
Second, they assume that if the corporation’s and the individual’s states are functionally equivalent, then they are morally equivalent. The step from functional to moral equivalence could be problematized. Hess’ and Björnsson’s dismissal of the phenomenal point of view in this context indicates that they have a narrow sense of “functional equivalence” in mind, where references to subjective experiences are excluded. (Functionalism in a broader sense might admit that a specific attitude’s functional role could include its relation to subjective experiences, or dispositions for subjective experiences.)
On that kind of functio alism, we would have to say that e.g. a robot programmed to imitate a mammal’s physical pain perfectly in terms of inputs and behavioral outputs is in a state that is functionally equivalent to the mammal’s physical pain. It seems reasonable to think that a machine could be programmed in this way without being extremely complex or displaying AI in any substantial sense. Would this kind of functional equivalence entail moral equivalence? But I will disregard this issue in the following, and focus on the first step in the argument.
The standard argument for regarding corporations as independent agents is that corporations can make decisions, express beliefs and follow policies that do not reflect the views of their individual members. This is taken to prove that “corporate commitments are distinct from member commitments, and this remains true regardless of whether they conflict or cohere” (Björnsson and Hess 2017: 276). I agree. It is obvious that such discrepancies can occur when the corporation is undemocratic, but it is also a fact that no collective decision procedure that fulfils reasonable rationality constraints can guarantee an outcome that reflects the beliefs and desires of its members. This is Philip Pettit’s main reason for thinking that groups have “minds of their own” with their own beliefs, desires, and intentions (Pettit 2003).
It could also be the case in periods that no collective or individual decisions at all are made about the corporation’s overarching goals and policies – corporate policies and commitments have simply been passed on from previous generations of members, and new members carry on executing those policies without contestation, “because that is how things are done” in this company, as Pettit says. As Björnsson and Hess note, members may continue acting on the corporation’s commitments and make various decisions based on them in response to external events, and they may even adjust the relative weight of those commitments when they come into conflict, without necessarily embracing these commitments as individuals.
I accept that this kind of distinction between member commitments and corporate commitments is sufficient for admitting that there is a basic sense in which a corporation can be regarded as an autonomous agent. Assume also with Björnsson and Hess that desires, beliefs, and intentions need not have any essentially phenomenal qualities, and that a corporation’s autonomy is no more undermined by the underlying control by its members than an individual’s autonomy is undermined by her decisions being caused by underlying mechanisms.
① My first worry concerns the next step in the argument: from ascribing basic agency to corporations on account of the distinction between corporate commitments and member commitments, to equating corporate beliefs and desires with individual beliefs and desires in the standard sense. What we have established is rather agency in a very basic sense. As Pettit admits,
If we are to recognize the integrated collectivity as an intentional subject, then we must admit of course that it is a subject of an unusual kind. It does not have its own faculties of perception or memory… it is incapable of forming degrees of belief and desire in the ordinary fashion of animal subjects; its beliefs are recorded as on- off judgments, its desires as on- off intentions. (2003: 182)
What we may think of as a corporation’s beliefs, desires, and intentions are “its commitments about how the world is, what goals to pursue, and how to act” (Björnsson and Hess 2017: 277). As far as I can see, there are essential functional differences between corporate commitments and the beliefs or desires of an individual. Take ACME’s commitment to be environmentally responsible. Such a commitment typically consists in a directive expressed in a policy document or a declaration that may be internal or public. The directive, in turn, is the end product of an individual or collective decision process. The corporate directive is the consequence of a decision taken at a specific point in time. The function of the directive is to promote or restrict future behavior of the corporation in this or that direction. The proper functional analogy to this kind of organizational commitment in the individual case would be the adoption of a pre- commitment device (like making a public New Year’s promise that I will be environmentally responsible in order to raise the cost for certain future choices, e.g. in terms of lost respect from those who heard my promise) rather than a continuous desire.
A full- fledged desire would be a continuous internal state of ACME, which explains why they took the specific decision to adopt this policy document, or would enable us to predict future commitments. But nothing in the description of what happens makes it necessary to postulate a state of ACME with that functional role. What explains how the directive came about is a specific individual or collective decision procedure (whose outcome admittedly need not reflect the views of ACME’s members).
If I would like to be environmentally responsible and know that I fail to be so, I have a continuously frustrated desire, i.e. I am in a state, presumably unpleasant, making me disposed to act and think in various ways.
If ACME has declared its commitment to environmental responsibility and fails to live up to it, there may be external sanctions of various kinds, and individual members may feel frustrated if they identify with ACME or sympathise with its stated policies. However, there is no reason to think that ACME as such thereby has a continuous frustrated desire. It has a policy document that was decided upon at a specific point in time. If ACME declares regret over its failure, this new declaration is the result of another individual or collective decision procedure.
② My other worries are more empirical and concern corporate guilt feeling behavior in real life. My evidence is anecdotal, but to begin with I find it rather unusual for corporations or their official representatives to display behavior indicative of such feelings on the part of the corporation as such. On the contrary, it seems to me that when the representatives of a business company or a public authority want to express concern for the victims of some harmful act that the corporation has committed or contributed to, they typically do this in a personal manner, marking that they and possibly their colleagues personally feel for the victims and their families etc. rather than that the impersonal corporation does so.
It is true, as Björnsson and Hess state, that ACME as a corporation “might issue apologies and compensate victims because ACME’s position is that this is what one does when one is responsible for some harm” (2017: 18). This could be a policy and a standard procedure in the company. Apologies and compensation are weak indicators of guilt feelings though. I might apologise and compensate you after having broken your vase even if both of us know that it was an accident and completely unintended. Compensation may be fully justified even in cases where it is clear that there is no moral guilt to begin with. In the context of tort law, the notion of strict liability often applies precisely to cases where people have been harmed by commercial products. In such cases there may be a common assumption that the company causing harm is also responsible for compensating the victims, regardless of any wrongdoing.
Moreover, we should not underestimate the role of pure business strategy as the explanation of why apologies and compensations are offered. Think of cases where the risk of costs for future lawsuits, fines, apologies, and compensations have been calculated and weighed against expected profit before the harmful decision is made.
Even if my empirical speculations about corporate behavior are true, they would not by themselves show that it would be impossible for a corporation to display all the subtle behavioral signs that are significant of guilt feelings. If that happened occasionally, only an oversimplified form of behaviorism or role functionalism, identifying guilt feelings with its external displays, would force us to conclude that the corporation in such a single case feels guilt. Unless this happens regularly and typically, we have no reason to believe that the company is in a continuous state warranting predictions about future patterns of behavior in relation to guilt in new decision contexts and circumstances.
In line with what I said in the first section of this chapter, I think that guilt feelings like other reactive attitudes have an essentially social and involving character, reflecting our sensitivity to how we are regarded by others and our caring about how they believe that we regard them. Moreover, like Hume I do not find it improper to regard feelings of guilt as a specific form of pain – it is unpleasant to be haunted by guilt feelings. Both properties are, I think, central to the function of blame and guilt feelings in morality as a system of social sanctions.
Björnsson and Hess agree that the unpleasant nature of guilt feelings might be important for its role in practices of holding responsibility (even though they do not think that this is abso lutely necessary for fully fledged moral agency, p. 288). They claim, though, that something like the motivational role of unpleasantness is an element in a corporation’s moral equivalent of guilt feelings, “since an organization instantiating the moral equivalent of guilt… will be thrown into disruptive internal conflict, conflict of a sort that it is motivated to avoid” (2017: 288). That rests on their assumption that the organization as such desires to avoid changing its commitments and values, and that taking on the moral equivalent of guilt feelings requires such changes. It is not clear to me why an organization as such necessarily should be motivated to avoid chan ging its commitments (nor that companies displaying elements of guilt behavior typically make fundamental policy changes). That seems to presuppose that the organization is attached to its commitments in a stronger sense than the assignment of basic agency, consisting of on- off intentions, and on- off judgments, can justify.
16.6 Blaming Collectives as a Way of Evoking Collectively Tainted Guilt in Individuals
As Gilbert points out, we may distinguish between the feeling of guilt that a person may have over her personal contribution to a collectively produced effect, and guilt feelings that are directed at the collective behavior of one’s group.
It is evident that groups to which we belong in various ways can figure in the objects of our frustration or displeasure. I may not only feel guilty about what my group has done, but also be dissatisfied with its performance, worried about its future, or depressed over its declining reputation. Such attitudes can differ from my views or feelings about my own individual performance, future, or reputation.
However, the mere fact that my personal preferences about my group can be frustrated, or that I can have unpleasant feelings over what my group has done or things that happened to it, does not invoke any need for a special category of desires or emotions. These may be ordinary first- person indi vidual attitudes, albeit with a conception of the collective in their content.
Gilbert tentatively suggests that an individual may not only have membership feelings (or, as she says, “feeling- sensations”) of this unproblematic kind – individual feelings directed at some aspect of one’s group – but also that distinctively collective feelings may occur in the individual’s mind.
When Gilbert discusses how the pangs of guilt feelings that a person may have due to some act that her group has performed should best be described – as pangs of personal guilt, mem bership guilt, or collective guilt – she says that although these phenomena are distinct, “from a phenomenological point of view, there may be no way of deciding this issue: a pang is a pang is a pang” (2002: 141). That caveat may be unnecessarily cautious. Among authors defending the possibility of genuine collective intentionality (without postulating group minds or plural subjects), it is not uncommon to associate certain intentional states with a collectively tainted phenomenal feature, a “sense of we- ness” (Pacherie 2012: 343) or “feeling of togetherness” (Zahavi 2015: 91). I think that there is some intuitive plausibility to this.
Consider again the case of Karl Jaspers. One way of understanding his predicament would be to say that although he was neither causally involved in the atrocities, nor felt any kind of sym pathy for these deeds or the ideology they were driven by, he felt guilt for them from the per spective of the German people – a collective with which he identifies. The idea that individuals may identify with groups has been commonly expressed in social psychology at least since the 1970s (for an overview, see e.g. Tajfel and Turner 1986, who discuss group identification in rela tion to social identity theory) and I believe that it is quite common in various popular debates on collective phenomena – people are said to identify to various degrees with their sports team, ethnic group, political party, etc. In the present context, we need to give a more exact meaning to such expressions. My suggestion is that there is a sense in which you can experience the situ ation from the group’s perspective.
To make that claim plausible it is not enough to appeal to phenomenological intuitions or introspective evidence. We should be able to give a functional characterization of this perspectival feature of our attitudes. Elsewhere I have suggested a way of developing such an account of the collective perspective, and I will not go into all details here but just give some hints about how I think one should proceed.9 My approach relies on the common assumption that we can characterize intentional states functionally in terms of what would make them successful. Some types of attitudes, like perceptions and action intentions, are only veridical or successful if their object stands in a certain relation to the subject of inten tion – my intention to raise my arm is not successful unless I raise my arm by way of carrying out this very intention, to use a standard example from John Searle.
Like François Recanati, I think that Searle is wrong in thinking that this implies that the subject of the intention (I) must figure in the content of the intentional state in question. When you perceive this text, a complete description of what makes your perception veridical must include a description of how the text is causally related to you and your experience, but the content of your percep tual experience is mainly the text. The content of the experience need not, and normally does not, include any conception of how the text is related to you. Moreover, we must distinguish between the subject of intention, which is a perspectival feature of some types of intentional states, and the ontological subject, i.e. the individual in whose head the intentional state resides. All intentional states have bearers, but some intentional states, like beliefs, need not have a subject of intention – it may not be necessary to refer to the believer when exposing the truth conditions for a belief.
As Recanati has argued, some types of intentional states admit of perspectival variation when it comes to time and place. This means that the context in which the content of the intentional state should be evaluated (according to its success conditions) need not be the context in which the ontological subject is situated when possessing the state. An episodic memory is not ver idical if its object occurs in the present context, but is if it occurred in the context of an earlier perceptual experience, for instance. So, the point in time and space from which the content of an intentional state is conceived need not be “here” and “now.” In a similar manner, I suggest that it is possible that some types of intentional states admit of perspectival variation when it comes to the subject of intention, i.e. that the subject of intention may be either “I” or “we.”
So, my claim is that there is a coherent and analyzable sense in which “the subject is immanent in the attitude” (to borrow a phrase from Hans Bernhard Schmid10) without necessarily being part of its content, and that this allows for genuinely collective attitudes without requiring the existence of group minds or independent plural subjects.
This section proceeded from Gilbert’s claim that individuals can experience collective feelings, and a phenomenological intuition about the plausibility of the claim that a sense of we- ness may accompany certain attitudes. The suggested functional characterization of how attitudes can be held from a group perspective does not rely on such phenomenological intuitions, but fits well with them. Perspectival features of a mental state may be part of the phenomenology of that state. As Recanati says, “there is absolutely no reason to consider that phenomenology supervenes on content in the narrow sense” (2009: 51).
According to Deborah Tollefsen, “one could understand collective emotions as those emotions that are expressed through the group members qua group members” (2003: 232). A collective guilt feeling is “the guilt one feels in response to the actions of one’s own group” (Tollefsen 2006: 237). I agree that these kinds of feelings occur and that they are distinct from the guilt an individual may feel over her individual contributions to the collective act. Such feelings are what Gilbert (and I) call “membership guilt feelings,” i.e. individual feelings directed at the collective action of the group in which the individual is a member. What makes my membership guilt feeling “collective” is its object.
But like Gilbert, I believe that individuals can have collective guilt feelings in a sense distinct from mere membership guilt feelings. I cannot only feel guilt for my group or what it has done. I can feel guilt from my group’s perspective. What makes this kind of feeling collective is not its object, but its intentional subject.
This approach will not provide any grounds for taking literally the claim that groups as such can feel guilt, but it may capture some of the intuitions behind such claims. It gives an explication of how feelings and preferences in the individual mind can be based on the individual’s identification with the group, and thereby have a genuinely collective character. Such feelings can be unpleasant even if there are no corresponding feelings of discomfort held from the individual perspective.
Let me return to Kutz’s challenge against the idea of collective guilt. Kutz thinks that it would be pointless to hold collectives accountable, since they are unable to respond affectively. Blame must therefore be directed towards the individual accomplices (Kutz 2001: 196). I am sympathetic to Kutz’s restrictive attitude to collective blame and sanctions, partly because of considerations in the previous sections of this chapter. If those considerations are correct then there may be no meaningful way of punishing the collective as such – collective blame, if effective, will merely make individual members feel guilty.
That said, I think that Kutz overlooks the possibility that by holding a collective responsible, we address its members in a way that is substantially different from what we do when we assign individual guilt. And although the collective as such is unable to respond affectively, its members may do so from the collective perspective. As Tollefsen says, the norms governing the relevant affective reactions might differ, and so might the motivational role of the distinct types of feelings evoked by assignments of individual and collective responsibility (2003: 232).
I would add that there is a motivational difference not only between the guilt I may feel over what I have done and the membership guilt I may feel over what we have done. There is also a difference in terms of motivation between feeling guilty as a member over what we have done and feeling guilty from our perspective. The latter but not the former kind of feeling fits in with the kind of “agency transformation” that makes a distinct motivational difference in certain prob lematic social choice situations, as Bacharach’s work on “team reasoning” indicates (Bacharach 2006). In other words, the functions and consequences of affective responses that are collective in this sense may be substantially different from what we are after when we blame each member individually. This way of addressing them may be appropriate when the collective character of the action for which they are blamed needs to be stressed.
16.7 Conclusion
Reasonable conceptual analyses of “collective guilt feelings” face the choice between assigning such feelings to groups as such while playing down the phenomenal element in the notion of “feelings,” or assigning them to individual group members while weakening the collectivistic element in the notion of “collective.” By discussing two examples, I have tried to show that the f irst strategy is less viable than the second, due to functional dis-analogies between the group case and the individual case.
My concluding claim is that an essential function of holding a collective morally responsible is to make its members feel guilt, albeit from the group’s perspective. Guilt feelings are reactive attitudes of an essentially social character, intimately connected with our view of others, and with our thoughts and wishes about their view of us. Such feelings are appropriate responses to moral blame. Guilt feelings are essential elements in a social sanction system that may not con stitute the whole sphere of morality but at least, as Hume, Mill, Gibbard, and others noted, occu pies the most central “region in our moral thought” (Gibbard 1990: 52). Like other negative sanctions, guilt feelings are in themselves unwanted and unpleasant even when they are justified and fulfil their function. Fairness therefore requires that blame for collective actions is directed towards properly delimited collectives, consisting of people whose membership features make them complicit in the act the group is blamed for. We need criteria of complicity that explain why Jaspers is unfair to himself when he feels guilty just in virtue of sharing nationality with the perpetrators of atrocities.
Very generally speaking, the two basic pillars of moral as well as legal accountability are causal involvement and intent. However, in accounts of co- responsibility for collective or corporate action, both of these conditions have been questioned. Due to causal or epistemic complications such as over- determination or undetectable causal links, philosophers like Christopher Kutz (2001) and Brian Lawson (2013) have argued that people sometimes should be held to account as accomplices to collective wrong- doing merely in virtue of participatory intentions, regardless of causal involvement. Because of possible discrepancies between a group’s collective decisions and the beliefs and desires of its members, other philosophers, like Torbjörn Tännsjö (2007), have abandoned or at least played down the requirement of intent.
I am more optimistic about upholding these two basic requirements for co- responsibility. I also think that giving them up would either have unwanted implications for moral and legal security, or make assignments of co- responsibility toothless. But I have discussed both kinds of worries in some detail elsewhere (2008, 2013) and they do not affect the main points of the present chapter.