Konzelmann Ziv, Anita (2007). Collective Guilt Feeling Revisited. Dialectica, 61 (3): 467-493.
Introduction
While collective guilt understood as collective responsibility is a common subject in political theory, theory of law and social psychology, the problem of collective guilt feeling has not found wide attention until recently.
As the term ‘guilt’ is ambiguous, denoting both factual guilt (blame, responsibility, liability, culpability) and the emotional state of feeling guilt for a fact, it is not clear in many cases whether notions of collective guilt refer to objective states of being guilty for p, or, in addition, to intentional states of feeling guilt for p.
It is, however, obvious that most authors who include intentional feelings of guilt in their analyses of collective guilt, would attribute such feelings to individual members of the collective and not to the collective itself. It is widely held that ‘the question of morals is always and fundamentally cast in individual terms’ (Narveson 2002, 180).
Nevertheless, some philosophers defend a view of genuine collective intentionality, according to which there is not only irreducible collective cognition and intention, but also irreducible collective emotionality exemplified by collective guilt.
The aim of the present paper is to evaluate arguments advanced in favour of the irreducibility of collective guilt feeling, and to argue for an account of collective guilt feeling in terms of feeling membership guilt. This alternative is ‘individualist’ with respect to the emotional subject of the feeling, while it is ‘collectivist’ with regard to the irreducibility of a we-feeling.
It presupposes the idea of a fundamental distinction between intentional states in the I-mode and intentional states in the we-mode, the latter not being reducible to the former. In recent years, this idea has been articulated and developed by several philosophers, among them John Searle and Raimo Tuomela. Without going into details of their approaches, I take we-feelings to be feelings in Tuomela’s strong ‘we-mode’ sense, according to which individuals do have we-states in their function ‘as group members because of a group reason’.
This interpretation allows accounting for the personal and societal impact of moral we-feelings without giving up the naturalist view of individual subjects of intentionality.
As for the question of an appropriate emotional response to collective wrongdoing, I propose to substitute the minimal requirement of feeling collective regret for the requirement of feeling collective guilt. This proposal is due to some characteristics of guilt and regret feelings, discussed in section 1 of the paper. The topic of section 2 will be Margaret Gilbert’s ‘plural subject’ account of collective guilt feeling, and section 3 is dedicated to a membership account of collective regret feeling.
Let me start with a few introductory remarks on affectivity and collective intentionality, domains that overlap in attempting to analyze the notion of collec tive guilt feeling or the notion of collective emotion in general.
Analytical research in collective intentionality started roughly 20 years ago with questions about collective intention and action, then extended to analyses of collective belief and has recently turned to the question of collective emotions.
Within affective science, philosophers’ interest concentrates on the question of emotions’ rationality, their relation to desire, belief and judgment, and criteria of their categorization.
Emotions like guilt, shame, remorse and regret constitute a family whose members are often qualified as ‘social’ and ‘self-reflecting’. While the term ‘social’ refers to interpersonal relations as being essential to the emotion, the term ‘self-reflecting’ usually qualifies emotions that presuppose or constitute self-consciousness or self-awareness on the side of their subjects. Since guilt-family emotions often arise in morally significant situations, they are also called ‘moral’ emotions. Guilt family emotions (e.g. guilt, remorse, regret) have an intentional object p whose satisfaction presupposes reciprocity within an interpersonal plurality. If x feels guilt for p, p implies an interpersonal relation between a wrongdoing agent x and a victim y suffering from some sort of wrongdoing.
In some accounts of collective intentionality, reciprocal and other interpersonal relations are identified with collective intentionality. According to such an account, personal states that neces sarily involve interpersonal relations– for instance promising (x,y,z) where the variables x and y represent persons– might be considered as manifesting collective intentionality even if displayed in the first-person-singular mode ‘I promise to y to do z’. In the following, however, I will confine the discussion to explicit collective intentionality, i.e. to cases where an intentional state is attributed to a collective grammatical subject, such as an ethnic group or a football team.
Analyses of explicit collective intentionality are usually given in either collectivist or individualist terms. The collectivist position holds that an ascription of collective intentionality refers to the ‘genuine’ intentional state of a collective, ‘shared’ somehow by the collective’s individual members without being reducible to their intentional states. The individualist position holds that collective intentional states are ‘aggregates’ of ‘joint’ individual intentional states.
The two positions may also be labeled in terms of ‘intentionality dualism’ versus ‘intentionality monism’. While collectivists believe that intentional states are experienced by at least two kinds of ontological entities, i.e. individuals and collectives, individualists believe that only natural individuals experience intentional states.
Experience of intentional states is displayed in the subjective mode of the first grammatical person, either ‘I’or ‘we’ taking the subject position. Investigating the question of collective guilt feeling, we investigate on subjective ascriptions of the form ‘We gp’, where g stands for an emotional predicate of the guilt family (guilt, remorse, regret), and index p for the emotion g’s intentional content. Since subjective emotion-ascriptions are part of the expression of the emotional experience ascribed, I will use the term ‘display of emotions’ to refer both to the experience and the subjective description of emotional states. Instances of ‘We feel guilt for p’ raise the question of what is experienced by whom, or, in other words, what a collective feeling of guilt is. Sections 1 and 3 of this paper are intended to shed some light on the question of what is exper ienced in collective guilt, while section 2 deals with the question of the subject of such emotional experience.
Reflections in section 1 will focus on
1) the normative feature of appropriateness,
2) the relation of guilt-family emotions to personhood and the self, and
3) the representative quality of guilt-family emotions.
The results of these considerations will serve as guidelines for a discussion of Margaret Gilbert’s collectivist interpretation of collective guilt feeling (section 2). According to Gilbert, we need to assume genuine guilt feeling of collective subjects in order to account for the moral function of guilt in intergroup relations. I cannot agree, though, with the claim that all individualist accounts fail to satisfy this function. Note that the term ‘individualist account’ does not imply a reductionist view as to we-intentionality. It does not imply that collective intentionality must be reduced to I-intentionality, but that collective intentionality has individual bearers.
My arguments for an individualist understanding of collective guilt feeling aim at showing that the emotional subject of such a feeling need not be identical with the agent of the wrongdoing involved in the feeling’s content. On the one hand, it is the notion of the self involved in all accounts of feeling guilt that posits a severe problem for a collectivist interpretation. On the other hand these accounts emphasize that emotional states of guilt can hardly be detached from certain phenomenal features such as feelings of discomfort and distress.
Defending an intentionality dualism, the collectivist position faces the difficulty of how to accommodate phenomenal features of emotion in collectives. The collectivist is more or less committed to adopt a judgmentalist view on emotion, according to which an emotion’s essential element is a judgment, while phenomenal feeling merely ‘accompanies’ it.
In contrast, monist accounts of intentionality allow for a non-reductive understanding of collective emotion without being bound to a specific view on emotion. The monist claim that only individuals experience emotions is not identical with the reductive claim that individuals only experience individual emotion, i.e. emotion displayed in the mode ‘I feel F’. It is conceivable that individuals equally experience collective emotion, i.e. emotion displayed in the mode ‘We feel F’.
It is important to distinguish the phrase ‘individual (or collective) emotion’ from the phrase ‘individuals’ (or collectives’) emotions’. While the former expression refers to a kind or mode of emotion, the latter refers to the subject of emotion. When I talk of ‘collectivist’ or ‘individualist’ accounts of collective intentional states, I refer to views on the subject of these states.
In section 3) of the paper, I will relate a membership account of collective emotion to the question of the appropriate emotional answer to collective wrongdoing. Considering the large distance that often exists between agents of collective wrongdoing and subjects of responding moral we-feelings, it seems that– in a general theory of moral response– regret is more adequate than guilt to fill the role of the morally appropriate we-emotion. Further, there is empirical evidence that expressions of we-feeling responding to collective wrongdoing rather occur in terms of regret and apology than in terms of guilt.
1.
‘Guilt’ is an ambiguous term in that it can refer
either
to states of factual responsibility for wrongdoing
or
to emotional states of feeling guilt for it.
We can distinguish the two senses by calling them ‘objective guilt’ and ‘subjective guilt’ respectively, or by referring to factual responsibility in terms of ‘culpability’, ‘blame(worthiness)’or ‘liability’.
In order to make things clear, I will use the term ‘feeling guilt’ or ‘guilt feeling’ for denoting emotional states of guilt.
As for the relation between feeling guilt and objective guilt, it seems obvious that x’s being objectively guilty does not entail x’s awareness of x’s liability, while x’s feeling guilt does entail such awareness [???].
X's being blameworthy (objective guilt)
X's belief that X is blameworthy (belief or judgement)
X's feeling guilt: X's awareness of X's belief or judgment accompanied by subjective emotional state (a painful emotional answer)
The blame for a certain outcome p can be attributed to x from an objective point of view, while x’s feeling guilt for p is a subjective state that requires not only x’s awareness (or belief) of objective guilt, but also a painful emotional answer to it.
It is usually agreed that feeling guilt for p entails believing that one is liable for certain wrong acts, but that the converse does not hold. Assuming responsibility for a certain act can go together with denying the act’s moral wrongness or its moral relevance for a blameworthy outcome p.
Another view on the relation between objective guilt and feeling moral guilt refers to the self-evaluative function of guilt feelings. Since the ‘self [. . .] evaluated by shame and guilt is not simply a product of the person’s own actions’, guilt feelings may arise without their being linked to actions of the self (Stocker 1990, 30). Guilt-family emotions that are experienced on behalf of someone else’s deeds are often called ‘vicarious feelings’. [대리감정]
Typical cases of feeling vicarious guilt occur in family relations, when, for example, a parent feels guilt for a wrong action of a child. Feeling vicarious guilt is due to an extension of personal responsibility for actions that are beyond personal control and, as such, is a function of specific interpersonal relations. It is, however, important to note that in order to feel vicarious guilt, acknowledgement of a morally wrong act in one’s realm of responsibility is also needed. We may thus assume that guilt feelings require acknowledgement of moral wrongdoing, stating that the standard case implies identity of the subject of the feeling with the agent of the believed wrongdoing, while the non-standard case of feeling vicarious guilt allows for nonidentity under certain conditions.
It is important to note that vicarious guilt feelings are usually not considered to be inappropriate, as, for instance, feelings of exaggerated self reproach or pathological guilt feelings are.
Inappropriate guilt feelings are based on mistaken beliefs about one’s realm of responsibility: a worker in a knife factory feeling guilt for the ferocities performed by a serial killer with a product of the factory would display an inappropriate emotion. [???]
Emotions’ qualification as ‘appropriate’or even ‘ought to be felt' is standard in contemporary discussion of affectivity and brings in a normative element.
It suggests that there are rules, or at least regularities, that determine whether displaying certain feelings is a sound emotional response given the specific circumstances. The notion of appropriateness emphasizes that higher emotions, such as guilt-family emotions, are considered to be ‘internal acts’ for which the subject is somehow responsible as it is for ‘external acts’. That such emotions are rule governed is not supposed to mean, though, that they are entirely subject to the will. The idea, rather, is that ‘appropriate’ emotional responses are the result of learning and other social processes shaping the individual personality. Many authors emphasize that guilt feelings, for example, are required in order to test the authenticity of wrongdoers’ moral ‘appreciation’ of their deeds:
If a moral lapse is sufficiently serious, it will not be enough to make up for the lapse with good deeds; rather, the agent must appreciate its seriousness, in a sense not unlike aesthetic appreciation to the extent that it rules out being left cold. [. . .] [A]t least some negative feeling seems to be needed to assure us that the agent’s negative evaluation of his act affects him personally (Greenspan 1995, 113).
A view of emotions as being internally related to values entails taking emotional appreciation of a moral failure to be itself a moral requirement.
According to this view, external acts of apology and reparation alone cannot account for the morality of a person, because they do not necessarily reflect an internalization of values. A moral person is supposed to internalize important values in a way that her self is severely touched by violations of these values. As ‘self-attributed responsibility for a wrong’, feeling guilt, for instance, is a sorrowful experience that ‘rests on identificatory self-anger’, constituting thus an ‘ “internal sanction” of the moral code involving some form of discomfort at the thought that one is responsible for a wrong’ (Greenspan 1995, 130f; 109).
As internal sanction of the moral code, feeling guilt posits a ‘threat to the self– to the estimation of character or merit’. The associated reparative desire amounts to a general ‘need to clear the self– to expiate a wrong, thereby erasing a “taint” [. . .] Guilt has [. . .] a potential future orientation mediated by its identificatory focus on the self’ (ibid. 135f).
It could be objected to this account that the highly controversial notion of the self is not ideal to rest a case on. It seems obvious, though, that in the context of guilt feeling analyses, the notion of the self represents the common sense idea of individual personality, of an individual’s ‘own’ as distinct from all others. Taking this idea as intuitively plausible, it seems to me that the supposed close connection between the self and guilt-family emotions threatens collectivist interpretations of collective guilt. The idea of internal sanction does not only underline a normative perspective on emotions– taking them ‘as able to serve as substitutes for action’12– but is also closely linked to the idea that a feeling of discomfort is essential to the emotional state of guilt. This point is important in view of the problem of the representative quality of emotions. In recent years, the cognitive aspect of emo tions has been more and more emphasized, resulting in the defense of so called ‘judgmentalist’accounts according to which the essential ingredient of an emotion is an evaluative judgment. Many judgmentalists claim that phenomenal features such as feeling ‘pangs’ or ‘twinges’ are to be considered as merely secondary to emotions, contingently accompanying the latter without constituting a defining value.
Other philosophers do not agree with that, although they also emphasize the representational function of emotions: provided that emotions are intentional states with a propositional content p, an emotion consists in a certain mental mode toward content p. In Patricia Greenspan’s words it is ‘emotional comfort or discomfort’that ‘holds the content in mind in the relevant sense’(Greenspan 1995, 166). We feel something– discomfort, sorrow, joy– on behalf of content p and it is this affective component that makes the state an emotion:
It is essential to this account that comfort and discomfort are taken as general intentional states– states of positive or negative affect directed toward evaluative propositions– rather than amounting merely to affective symptoms of emotional evaluation [. . .]. Thus, guilt amounts to discomfort at or about the thought of oneself as responsible for a wrong– not just to the thought plus an accompanying pang of discomfort (Greenspan 1995, 166).
It is important to note that this account of the essence of feelings for emotion does not claim that different emotions are individuated by specific phenomenal features. It rather claims that emotions differ from cognitive states such as beliefs in that they are different intentional modes or qualities towards an intentional content p. According to this view, an emotion is not constituted by a judgment or belief that p, nor is it just a certain kind of attitude towards a judgment that p. An emotion rather is an affective state towards p, whereas p does not need to be asserted in a judgment or in a belief. For p to become the content of an emotional state of subject x, it suffices to be held in mind by subject x. A corresponding assertion of p in form of an evaluative judgment or belief may then be considered as cognitive reaction to the emotion and the subject’s reflecting over it. Such assessment of an emotion’s content as well as different behavioral responses to it can be used in individuating different emotions.
The decision for a judgmentalist or a non-judgmentalist account of emotion plays an important role in the question of collective guilt feeling. A view that considers affective phenomena of emotional states as mere side-effects of the essentially involved judgments allows more easily for collectivist interpretations of ‘We gp’ than a view for which the phenomenal element is essential to the emotion.14 The ‘feeling’component of emotions is conceptually linked not only to an intentional mind but also to a body with physiological reactions. The self involved in guilt-family emotions is individualized by the space-time extension of a material body with its specific experiences. Therefore, neither the self nor its feelings can be literally shared with others.
I should like to close this brief overview on some features of guilt-family emotions with a few remarks on distinctive characteristics of guilt and regret feelings. According to Greenspan, it is precisely the presumed close connection to the self that functions as a distinctive mark between the emotion of guilt and its relatives regret and remorse. Although feeling regret or feeling remorse are con sidered to share guilt feelings actively moral character by supporting an ‘urge toward reparative action’ (Greenspan 1995, 132), neither is supposed to involve any kind of threat to the self. On the contrary, they ‘seem to use negative feeling simply to distance the self from its past acts without any intervening stage of self-punishment’ (Greenspan 1995, 135). Notwithstanding the acceptance of this distinction, it emphasizes the important notion of a ‘distance’ between the subject of regret feelings and the regretted acts, a distance that is larger than in the case of guilt feelings.
While Greenspan thinks that feeling regret differs from feeling guilt in not challenging the self to the same degree, Janet Landman sketches their difference in other lines. Discussing the question whether regret is rather a cognitive or an emotional state, Landman argues for the irreducibly emotional character of regret: ‘Regret is emotional because it always implicates the self to some degree– if not the self as agent or character, then the self as caring about the regretted matter’ (Landman 1993, 44). What Landman emphasizes is a different kind of a self’s involvement in feeling regret and in feeling guilt. While feeling guilt is usually taken to involve the self as an agent, feeling regret involves the self as caring about the object matter of the feeling. Asubject feeling regret is caring for a certain state of affairs, i.e. is taking it personally, independent of personal involvement in this state’s obtaining. Therefore, the concept of regret is a wider concept than that of guilt or remorse, since it includes in its extension feelings that are not only self-evaluative, but also concern the self in its moral evaluation of self-independent facts.15 While guilt and remorse feelings’ objects are morally wrong acts in the subject’s own past, regret applies to these circumstances as well as to others like ‘unexecuted intentions, thoughts, wishes, impulses; one’s own future, involuntary, and morally innocuous or virtuous acts; and circumstances and the acts of others that share the foregoing characteristics’ (Landman 1993, 53). In this perspective, regret can be said to have a stronger cognitive component than regret’s relatives remorse and guilt:
[R]elative to other emotions such as anger or fear, regret does seem more cognitively elaborated. And relative to related emotions such as remorse and guilt, regret may often [. . .] be experienced more as a matter of ‘cool’ cognitive assessment than of ‘warm’ emotional reactivity (Landman 1993, 37).
Since the satisfaction of feeling regret does not require the emotional subject to be involved as responsible agent in the emotion’s content, subjects experiencing regret usually feel a kind of distance to their regret’s content that subjects expe riencing guilt do not feel. Therefore, one may doubt whether feeling regret could meet the requirements for an appropriate emotional response to collective wrong doing. Since regret is not internally related to blameworthiness, one might think that feeling regret lacks the motivational strength towards acts of apology and reparation that characterizes shame or guilt feeling. It is not my objective here to provide a profound analysis of the different related emotions of the guilt family. Yet most available analyses of guilt-family emotions take action regret as having an actively moral character, i.e. supporting an ‘urge toward reparative action’ (Greenspan 1995, 132). We can, therefore, state that feeling regret fulfills at least the condition of necessity for a morally appropriate emotional response to collec tive wrongdoing. Whether it is a sufficient response in all cases of collective wrongdoing is a question I will leave open for the moment. I will come back to this question in section 3)
2.
Collective responsibility, i.e. collective objective guilt, is a subject commonly discussed in ethical theory, philosophy of law, and social psychology. It is widely accepted that there are good reasons to attribute objective responsibility not only to individuals but also to groups whose structure allows that common goals, decision processes and acting are ascribed to them. There is less unanimity, however, in the question of how such responsibility, as well as consequences resulting from it, can be distributed over individual group members. While one view holds that collective responsibility distributes equally to each group member, the opposite view holds that if the group is responsible, no member is. Between these two extremes, liable to arguments of unfairness in the first and of downplay ing individual responsibility in the second case, a less rigid position distributes group responsibility in different degrees of individual responsibility:
There can well be degrees of responsibility in the case of the shared responsibility of unorganized groups, as in the case of collective responsibility of organized ones (Held 2002, 166).
According to such views, blaming groups is not a way to downplay individual responsibility, but might be, for example, ‘a special way of coping with a moral situation characterized by uncertainty’. Statements of collective blame ‘are practical ways [. . .] of organizing a certain complex of limited information about individual blameworthiness in organizational contexts’ (Pfeiffer 1995, 149, 141)
The problem of the distribution of group responsibility to individual group members is mirrored in the problem of collective guilt feelings. Let us grant, for the present purpose, that groups do collectively intend and act, and, consequently, bear collective responsibility. Are we justified, then, to extend this assumption to the claim that groups do also feel collectively? According to Margaret Gilbert we are, at least with regard to moral feelings.16 Over the years, Gilbert developed a ‘plural subject’ account of collective intentions into a general theory of collective intentionality, according to which it is an obtaining fact that groups do have genuine intentions, beliefs and emotions. To have genuine intentional states means, in the case of groups, that these states are not reducible to intentional states of the individuals who constitute the plural or group subject. Gilbert’s approach is based on the key concept of ‘joint commitment’, which refers to the ‘objective ground’ and the ‘clear rationale’ of collective intentionality. As ruling principle of ‘plural subjecthood’, joint commitment implies 1) open expression of each mem ber’s readiness to be jointly committed; 2) all members’ understanding of their relations of rights and obligations among themselves; and 3) ‘transitivity’: joint commitment can be a function of a (preceding) authority-creating joint commit ment (Gilbert 2002, 126f; 136). On the basis of joint commitment, groups are the subjects of collective intentions, collective beliefs and collective feelings. Conse quently, Gilbert gives the following two options for an analysis of subjective ascriptions of feeling collective guilt:
For us collectively to feel guilt over our action A is for us to be jointly committed to feeling guilt as a body over our action A. For us collectively to feel guilt over our action A is for us to constitute the plural subject of a feeling of guilt over our action A (Gilbert 2002, 139).
Developing an explanation of this account, Gilbert relies roughly on four reasons: 1) common language; 2) the grammar of plural terms; 3) a judgmentalist view of emotion; 4) moral relevance.
As for the argument of common language, Gilbert mentions the ‘fact’ that ‘emotions are routinely ascribed to [groups]’ (Gilbert 2002, 118) and assumes that people frequently speaking in this way presumably ‘think that there is some thing [. . .] real, to which they refer: the feelings of a group’ (ibid.). Her project then is to give an account of these presumed real feelings of groups, arguing that if collective guilt feelings were possible, ‘they would be of great practical importance’ (ibid.).
I think there are mainly two objections against the argument of common language. First, it is doubtful whether ascriptions of guilt feelings to collectives really belong to our everyday linguistic routine. Second, it is doubtful whether people using such ascriptions think that, by doing so, they literally refer to something real. The first objection does not disavow that people indeed might be inclined to use normative statements such as ‘They really should feel guilt for p!’ It rather intends to show that descriptive ascriptions of guilt feelings to spe cific collectives like ‘The American government is feeling guilt for having neglected its duties concerning p’ are rather uncommon. Even less common are, as far as I could observe, subjective ascriptions of collective guilt feelings.
Japan’s former Prime Minister Koizumi’s apology for Japanese war crimes in terms of ‘We express our deep remorse and heartfelt apology’ is quite exceptional in this regard [???].
Speakers assuming collective guilt for p usually assume objective guilt by acknowledging responsibility for p, combined with expressions of regret and apology. It is also typical that officials speaking for a group express regret and apology not for ‘our’ wrong acting, but for the wrong acting of ‘some of our members’ [???]. For example, the 1998 Vatican document acknowl edging a lack of church members’ resistance against the Holocaust refers to ‘sons and daughters of the church’ that sinned, and its declaration of guilt with regard to Israel (2000) mentions the ‘sins of Christians’.
With regard to the second objection, such examples seem to show that– instead of clearly pointing to real feelings of a group– the phenomenon of displaying a collective moral feeling is naturally understood in terms of individual constituents, either of the subjects of these states, or of the agents involved in their content. Thus, even granted that ascribing guilt feelings to collectives were common practice, it would not imply that such ascriptions refer– or are understood to refer– in a non-metaphorical way to collectives as emotional subjects, or that emotional states of collectives obtain.
In her account of collective guilt, Margaret Gilbert claims that distributive analyses of collective feeling ascriptions ‘We gp’ in terms of ‘personal guilt’ or ‘membership guilt’ are ‘found wanting,’ and that it is ‘implausible’ to construe collective guilt feelings in terms of such individual feelings (Gilbert 2002, 115). While the ‘personal guilt’ account is rejected for inadequacy of the ‘object’ or content of the feeling, the ‘membership guilt’ account is rejected for inadequacy of its ‘subject’.
Given that personal guilt feeling is confined to one’s personal deeds, content p in ‘We gp’ stating a group’s wrongdoing does not correspond to the content of individual group members’ personal guilt feelings. Even if the whole distribution of group members would feel guilt on behalf of their personal contribution to the wrongdoing, this would not amount to feeling guilt for the group’s wrongdoing p. While such a distributive ‘personal guilt’ analysis of ‘We gp’ does not ‘seem to capture the object of the guilt in the analysandum’ (Gilbert 2002, 131), a distributive ‘membership guilt’ analysis will preserve this object, since individuals feeling membership guilt ‘feel guilt over what a group of which they are members has done’(Gilbert 2002, 134).
Nevertheless, Gilbert takes the mem bership account as inappropriate for the analysis of collective guilt feeling, its fault being that ‘the feeling does not have a collective subject’. Thus, a membership interpretation would not account for the ‘emotional state of the citizen body,’ but merely for a certain kind of emotional states of individual citizens. It seems to me that this argument against a distributive membership account lacks explanatory force, since it simply rules out the logical indeterminacy of the plural pronominal form. It assumes right from the start the claim that needs proving, namely that the collective interpretation is the right one, or, in other words, that a display of the form ‘We gp’ is principally unanalyzable.
In displaying an emotional state such as ‘We feel ashamed’ team members seem to be aware of themselves as team members feeling ashamed. The grammatically distributive character of we-statements insinuates that the default function of ‘we’ is rather to refer distributively to the members of the group than to refer collectively to the singular group entity. It does not, however, exclude collective analysis, especially in those cases where the predicate has a genuinely collective character such as in ‘We intend to perform the 5th symphony of Beethoven’. Generally speaking, the logical grammar of ‘we’ does not allow concluding that the meaning of subjective ascriptions combining ‘we’ with an intentional predicate is collective by default, nor that such ascriptions presuppose by default a joint commitment of individuals to undergo the intentional state ‘as a body’.
Talking about collective emotions confronts us with the problem of explaining them either as a kind of mental phenomena of natural individuals, or as being ‘mental states’ of collectives. While the first option in the eyes of many thinkers does not suffice to account for the strength of the notion of collective states, the second option challenges both common sense and philosophical analysis. Notions like ‘collective mind’, ‘collective consciousness’ or ‘collective self’, are cautiously avoided by contemporary philosophers, even by those defending a collectivist view on collective intentionality. Gilbert, for instance, explicitly claims that, in order to give a proper account of collective guilt, it should not be modeled on individual guilt feelings. But if the guilt feeling of a group is explained in terms of ‘feeling guilt as a body’, the model of a natural individual’s emotional states appears just applied to another type of individual.
From certain observations made with regard to common language Gilbert concludes that entities such as feelings of groups are in fact real entities. The group is individualized as ‘a plural subject’ that feels guilt ‘as a body’. Yet the group, as it lacks a natural body, can hardly count as the subject of phenomenal feeling such as ‘pangs’ or ‘twinges’ of guilt. In order to combine emotional intentionality with non-natural individuals we need a specific approach to emotion. Gilbert therefore assumes– as a third line in her argument– a judgmentalist account of emotion:
I can imagine that I felt guilty about something without meaning to imply that any particular phenomenological condition was satisfied. The central if not the sole thing at issue would be my judgment that I was wrong. [. . .] [W]hether or not guilt feelings necessarily involve feeling-sensations of one kind or another, they involve a kind of belief (Gilbert 2002, 120).
The judgmentalist account, according to which the principal constituent of an emotion is a judgment, is combined with the assumption that phenomenal condi tions for certain emotion types cannot be given. Therefore, we are allowed ‘to regard it as at least open whether guilt feelings necessarily involve either a qualitatively special “pang of guilt”, a generic type of “pang”, or any phenomeno logical condition at all’ (Gilbert 2002, 120). Although it is not made clear how, in such an account, feeling guilt is to be distinguished from judging that one is liable for a wrong, it is made clear that feeling sensations are not supposed to account for the difference. If the emotions attributed to non-natural individuals such as groups or collectives are void of feeling sensations, objections to collectivist accounts of the sort that a collective ‘cannot respond affectively,’ or that it lacks the possibility ‘of an affective counter-response’ (Kutz 2000, 196) loose their relevance.
According to the judgmentalist line of her argument, Gilbert assumes that a collective’s feeling guilt is constituted by the collective’s believing that it is liable for a wrong. As a number of philosophers hold, intentional behavior can be ascribed to collectives as a function of their ‘ability to formulate what they accept or decide in language’, without there being a need to state ‘chemical and physi ological stimulation of feelings’. Being ‘regulated by well-understood procedures and conventions’, collective speech-acts allow for rational justification of certain ascriptions (Cohen 1992, 53ff). Such a strategy is apparently followed by Gilbert in her account of ‘feeling guilt as a body’ according to which to be ‘jointly committed to feeling guilt as a body’ or to be ‘the plural subject of a feeling of guilt’ is the result of either a direct commitment of the collective’s members, or of an authority committing the collective to the feeling. For the first case, Gilbert gives the example of a couple neglecting their responsibility towards a little girl they are babysitting during the weekend. While the woman later expresses a we-feeling of guilt towards the mother of the baby, the man denies such a feeling. The woman’s confirming ‘Yes, we do feel that way’and the man’s ‘concurring’by not replying anymore ‘may be enough to establish a joint commitment [. . .] to feel guilt as a body’ between them (Gilbert 2002, 140). In the second case, a govern ment’s official, expressing the nation’s guilt feeling towards a third party, is taken to represent the government’s entitlement ‘to determine the emotional state of the citizen body’, by ‘jointly committing the citizens to feel guilt as a body’ (Gilbert 2002, 140f). These examples figure in the section ‘The genesis of collective guilt feelings’, suggesting that collectives’ feelings are actually generated in the circumstances evoked.
The examples suggest that the obtaining of a collective’s guilt feeling mainly involves a declarative act in conditions of joint commitment. If a sufficiently authorized representative of a plural subject declares the emotional state, the plural subject members are bound to it in a way that makes the state obtain.
The main problem of this account is how to integrate individual members’ feelings. Since the collectivist account rejects a distributive explanation of the collective feeling, the latter does not seem to be related to individual members’ feelings. The collectivist account in principle allows for a collective feeling to obtain with many or most or all individual members having contrary feelings. The husband in Gilbert’s first example, although not voicing his objections to his wife, may still continue not to experience any feeling of guilt.
The same most probably applies to a large part of a nation’s citizens, bound to feeling guilt by an official’s declarative act. If the collective feeling is generated by a declarative act in conditions of joint commitment, the fact of contrary individual feelings apparently does not have any influence on it. Yet Gilbert holds that the joint commitment to feel guilt as a body would be ‘violated’ by individuals ‘expressing contrary feelings’, even if they were ‘unaware of the official’s pronouncement’ (Gilbert 2002, 141).
This remark suggests that individuals ought to harmonize their personal feeling with the declared guilt feeling of the collective whose members they are. Given that collectivist accounts reject an analysis of collective intentional states in terms of individual intentional states, it is surprising to find such a requirement in Gilbert’s account. If a collective guilt feeling is not a distribution of individual (personal or membership) guilt feelings, it is not obvious that the latter should be related in a normative way to the former.
Apparently, the normative relation between individual and collective feeling is supposed to hold not in terms of ‘having a feeling’ but rather in terms of ‘expressing a feeling’. Expressions of contrary individual feelings would undermine the soundness of a representative’s expressing a collective guilt feeling and, therefore, justify doubts on behalf of the obtaining of the collective’s feeling.
Having contrary feelings, however, could not count as violating a ‘joint commitment to feel guilt as a body’, as long as such feelings were not publicly exhibited.
This reminds us Gilbert’s account of collective belief according to which a collective or group G can be said to believe that p if its members ‘jointly accept that p’ by openly expressing ‘a conditional commitment jointly to accept that p together with the other members of G’(Gilbert 1987, 195). Gilbert emphasizes that in her account of collective belief ‘it is not a necessary condition of a group’s belief that p that most members of the group believe that p. Indeed, [. . .] it seems that it is not necessary that any members of the group personally believe that p’ (Gilbert 1987, 191). What is important, though, is that the group members ‘at all public points’ in their lives speak and act in ways which conform to the jointly accepted belief. In so promoting their collective belief they might be led to adapt their contrary personal beliefs to it.23
Given a judgmentalist account of emotion, Gilbert’s analysis of collective guilt feeling appears to be rather an analysis of collective belief than an analysis of collective emotion. If the constitutive element of an emotion is understood as an evaluative judgment, and if judgments or beliefs can be ascribed to collectives, then a theory of collective belief seems sufficient to account for collective emotion. It might be objected that collective emotion differs from collective belief in that it carries a motivational force that the latter does not. Gilbert in fact refers to the motivational force of guilt feeling, using its moral relevance as a forth reason to justify a collectivist account of collective emotion:
Were collective emotions such as guilt feelings possible, then, they would be of great practical importance. They would help to ameliorate relations [. . .]. They would also be apt to improve the relevant collectives themselves (Gilbert 2002, 118).
In the argument of moral relevance, two aspects can be distinguished. On the one hand, we find the idea that the function of collective guilt feelings lies in their moral power to motivate actions of apology and reparation on the level of inter group relations: a collective’s emotional state of guilt generated through the right kind of declarative act ‘We feel guilt for p!’will constrain the collective’s members ‘to do and say things that echo or conform to the claim’ that ‘we’ feel guilt for p (Gilbert 2002, 140). On the other hand, it is suggested that this function cannot be fulfilled if collective guilt feeling is analyzed in terms of an individualist account such as ‘personal feeling’ or ‘membership feeling’.
As for the first aspect, we note that according to Gilbert’s and others’ accounts of collective belief it is the function of the latter to initiate action.24 If this is correct, there seems to be no need to stipulate collective emotion for that purpose: then a collective’s belief expressing acknowledgement of the collective’s objective guilt can motivate the collective to perform acts of apology and reparation. [but this is Gilbert's way of defining collective guilt feeling] The motivation is moral insofar as it springs from the belief that what has been done collectively was morally wrong. A collective’s emotion does not seem necessary in this scenario.
As for the second aspect, it is difficult to see why collective guilt in terms of membership feeling should lack a motivational force that it is supposed to have in terms of a collective’s ‘feeling guilt as a body’. The idea of membership guilt contains the idea of vicarious guilt as a function of membership, i.e. condi tions determining specific individual-group relations. Analyzing an ascription of collective guilt feeling in terms of membership guilt preserves the collective guilt’s content (‘object’). As a member of group c responsible for genocide G, individual x may feel guilt for G, although x may have contributed very little– perhaps nothing– to perform G. In this case, individual x’s displaying guilt feelings on behalf of G is a function of x’s awareness to be tied by specific relations to c as a constitutive member. Such an emotional identification with the group can plausi bly be a strong motivational force for acts of apology and reparation.
Consider as an example individual Serbian citizens experiencing membership guilt for the massacre of Srebrenica. They will presumably speak and act in ways to prevent ethnic hatred resulting in massacres of this kind. Consider now the collective Serbian subject committed by its government’s declaring: ‘We feel guilt for the massacre of Srebrenica’to feel guilt as a body over the massacre. Is it likely to respond in the same way? There are serious doubts about that:
The hard case is where the citizen(s) may know of the official’s pronouncement and react with apathy or even hostility. [. . .] In such circumstances, official pronounce ments of collective feelings of guilt instead of reflecting or increasing ‘solidarity’ amongmembersofagivencollective, may actually reveal a moral gulf separating the officials of a collective from the rank and file members (Wilkins 2002, 154).
It is difficult to consider a collective citizens’ body to speak and act in ways reflecting a collective feeling that is not supported by a majority of individual citizens. We may wonder how the force of such a feeling might be manifest, given that feeling guilt has any motivational strength at all. On the other hand, it seems reasonable to assume that individuals capable of having we-feelings of guilt for wrongdoing in which they did not personally participate are also able to be motivated by such feelings. Even the hypothetical case in which ‘each member believes that she or he alone feels membership guilt over the group’s act’(Gilbert 2002, 137f) does not rule out the possibility of a motivational potential of such feeling. Concerning its motivational force, the problem of membership feeling mayrather lie in its weakness of impact than in its absence.We might ask: how can individual members’ acts of apology and reparation for collective wrongdoing be relevant to the extent of the wrongdoing? But we cannot rule out the membership account of collective guilt feeling for lack of moral relevance or motivational force.
Defending a ‘plural subject’account of collective guilt feeling, Gilbert does not exclude individualist accounts in terms of ‘personal feeling’ or ‘membership feeling’ from being warranted in certain cases. But she claims that such feelings coming down to an ‘aggregate of guilt feelings attributable to individual group members’– are only ‘so-called collective guilt feelings’ (Gilbert 2002, 142). Contrariwise, the ‘plural subject’ account– appearing ‘to imply the existence of a collective subject’– is supposed to ‘capture the gist of everyday attributions of collective guilt feelings’ and ‘seems most plausibly to be considered a collective way of feeling guilt’ (Gilbert 2002, 142).
The problem of the ‘irreducibly collective subject’account of feeling concerns the relation between supposedly genuine collective guilt feelings and feelings of personal or membership guilt: while ‘important connections’ are assumed, it is also quite clearly stated that ‘no one of these feelings seems to carry another with it as a matter of logic’ (Gilbert 2002, 142). Being logically independent, a collec tive guilt feeling in the ‘plural subject’ sense can thus exist without requiring any corresponding membership or personal feelings. Mutually expressed, however, membership feelings might lead group members to ‘constitute themselves a plural subject of guilt feelings’ (Gilbert 2002, 142). This is one version given of an important but contingent connection between a collective’s c feeling of guilt for p and c-members individual membership feelings of guilt for p.
A problem in Gilbert’s account is that the terminology used in the paper seems to imply a necessary connection: we read that a government, ‘acting through its officials, is entitled to determine the emotional state of the citizen body’, by ‘jointly commit ting the citizens to feel guilt as a body’ (Gilbert 2002, 140f). On the one hand, the emotional state is attributed to the collective, ‘the citizen body’, while on the other hand, it is ‘the citizens’ that are committed ‘to feel guilt as a body’. These expressions seem to suggest that the emotional state of ‘the citizen body’– i.e. the emotional state of the plural subject– is displayed in ‘the [individual] citizens’, in their ‘feeling guilt as a body’. It seems as if the ‘citizen body’s’ feeling could not exist without citizens ‘feeling guilt as a body’.25 Thus, we are either back to a membership account, or we have to assume that the irreducible collective feeling is somehow split into a unique component borne by the plural subject incarnated as ‘the citizen body’and a component distributed onto individual citizens. In either case, an aggregate of individual feelings seems to be necessary to the collective guilt feeling, although this is explicitly denied in the collectivist account.26 Gilbert seems to adopt the latter view when she concedes that collective guilt might be accompanied by feeling sensations experienced in the minds of individuals.27 It appears bizarre to now reintroduce the afore disbanded phenomenal aspect of collective guilt, ascribing it to ‘the minds of individuals’ whose feeling guilt, even joined, cannot amount to a collective guilt feeling but remain a mere aggregate of individual feeling. How could individuals feel the ‘pangs’of irreducible collective guilt for an emotion they, as individuals, do not have?
The moral function of guilt feelings– constraining people to conform to the claim of guilt– is a good argument in favor of an account of collective guilt feeling, but it does not strengthen a need for a collectivist interpretation. The moral function of collective guilt feeling is entirely satisfied within an individualist account of membership guilt. Feeling collective guilt– if any such thing obtains is not a collective’s guilt feeling but its individual members’ feelings of collective guilt. Declaring we-feelings of guilt as individual feelings of collective guilt does not amount to claim priority of the individual over the collective. It does amount, though, to the claim that only individuals experience emotions. Individual self consciousness can be considered as essentially two-fold, being constituted of both we-consciousness and I-consciousness. There is no need to reduce intentional we-states to intentional I-states or vice versa. Being shaped by the relations in which they stand with others, individuals owe their singularity not only to distinct time-space-positions but also to their being individuals-in-relations.
3
Given a view of emotions as being internally related to values, we assume that emotional appreciation of a moral failure is morally required for rational beings. We further assume that ascriptions of guilt family emotions ‘We gp’ refer to obtaining states of collective emotion. The question now arises: what member of the guilt family is best suited to fill the role of the appropriate emotional response to collective wrongdoing in a general account of it? Asking this question is not suggesting that subjects reporting we-feelings of guilt on behalf of collective wrongdoing were possibly mistaken about their feelings. It rather suggests that an account of a general moral requirement should be given in terms of a minimal requirement applicable to all cases of individual involvement in collective wrong doing. The need for a minimal requirement is warranted by the fact that the term ‘collective wrongdoing’ extends to a wide array of heterogeneous cases, differing not only in size and structure of the collective agent and the wrong committed, but also in the distribution of the involvement of individual members. It seems odd, for example, to require that German Jewish citizens– as members of the German collective– feel guilt for the atrocities committed by the German collective on the German Jewish community. Likewise, it seems odd to say that people not directly involved in these atrocities should not have any reactive moral emotions on their behalf. A requirement for emotional responses coming in degrees proportional to degrees of personal involvement does neither seem a practical tool for a general account of the moral assessment of collective wrongdoing.
In a general account of appropriate emotional responses to collective wrong doing we need an emotion that is strong enough to assess the wrongness of an action and weak enough not to violate rational principles of attributing culpability. The function of such an emotion is to provide a minimal moral requirement for the appropriate reaction of a collective’s members to a wrongdoing of the collective. Stating such a minimal requirement is not stating that it is a sufficient requirement in all cases of collective wrongdoing and for all members of the collective. Depending on various parameters, the appropriate requirement has to be aug mented by means of stronger emotional responses.
As already mentioned in the first section of this paper, I take regret to be the emotion that fits the role of the minimal requirement for an appropriate response to collective wrongdoing. More precisely, I am talking about action regret, i.e. regret that is not directed to a mere state of affairs such as someone being sick, but to the fact that a certain action has been performed in a certain way. In the following, ‘regret’ is to be understood as ‘action regret’, and also, it is understood as moral regret. This means that we are talking about an emotion evaluating the moral character of an action, as distinct from an emotion evaluating simply the success of an action. Regret, like guilt (and unlike shame), is being directed to action, although there are important differences between these two members of the guilt-family emotions. In section 1, I sketched a few of these differences. Now I will concentrate on the aspect of the relation between the emotional subject and the responsible agent of the action constituting the emotion’s content. I assumed in section 1 that appropriate personal feelings of guilt imply the belief that one bears personal objective guilt, i.e. the belief that the feeling subject is involved as an agent in a wrongdoing. Extending objective responsibility to the agency of speci f ied others, e.g. one’s children, leads to a view of appropriate vicarious feelings of guilt.28 While guilt and also remorse are feelings primarily centered on one’s own acts, the case of regret is different. Contrary to guilt, moral regret incontestably includes within its contents wrongdoing that is out of one’s control. There is, for example, no question whether it is appropriate for me to feel regret for a decision taken by the Swiss government, while the appropriateness of my feeling guilt for it could be controversial
Unlike guilt, regret is conceptually independent of an identity between the emotional subject and the presumed agent of the emotion’s content. We do not have to stretch the concept of regret to include vicarious cases of agency into its extension. Whatever stance we take in the question of the appropriateness of vicarious guilt feeling, we can adopt regret as uncontroversial regarding the appropriateness of its direction to the deeds of others. On the other hand, regret is similar to guilt in being a reactive emotion whose function is to morally assess certain facts. The combination of these two features makes regret appear the ideal candidate for the role of an emotion required in response to collective wrongdoing
To state that feelings of regret fulfill best the moral requirement of an emo tional answer to collective wrongdoing is not to set up an error theory about reported feelings of collective guilt. I do not say that people feeling guilt for what their ancestors did are mistaken about the nature of their feeling. Given conditions of sincerity and privileged access, I take it that these people do actually feel guilt.
But the fact that vicarious guilt feeling for group action actually is experienced does not amount to stating that such feeling ought to be experienced from a moral point of view. The fact that such feelings occur may be the result of education, oversensitivity or other circumstances. If we agree that guilt feelings are intimately connected to the self, we may prefer to restrict conditions for vicarious guilt so tightly that only deeds of those being the closest to the self are admitted in the realm of guilt liability. For deeds not falling under these conditions, a person ought not to acknowledge objective guilt, nor display guilt feelings. [???]
As a general requirement of assessing collective wrongdoing, the demand for feeling (membership) guilt seems too strong. In view of the heterogeneity of cases of collective wrongdoing, it appears unfair, even wrong, to charge individuals indiscriminately with the requirement of feeling guilt for acts that– temporally, locally or in other respects relevant to influence– may be entirely out of their control. Even the Christian Myth, according to which one man could and did take humankind’s blame upon himself, does neither require nor state that this indi vidual ought to feel or felt guilt on behalf of humankind’s blame. Similarly, this individual did not repent of humankind’s blame or feel remorse for it.Yet, still in the scope of the myth, we can imagine that he regretted the wrong leading to humankind’s blame and that he felt pain on behalf of it. These feelings apparently had enough motivational force to induce him to engage in acts of reparation.
If we take regret feeling as the minimal requirement for an appropriate emotional response to collective wrongdoing, we say that each member of an objectively guilty group ought to feel regret for the wrongdoing committed by the group, regardless of the degree of personal involvement in the wrongdoing. This membership feeling of regret ought to be augmented with feelings of guilt, remorse and shame, corresponding to the members’ involvement in the wrong doing.
Wemight, for instance, say that, from a moral point of view, Serbian woman x living as a refugee in Germany ought to feel membership regret for the massacre of Srebrenica. But she is not committed to feel any kind of guilt for it. On the other hand, Serbian soldier y killing under general Mladic’s command in Srebrenica ought to feel more than regret for the massacre. He ought to feel guilt for having participated in it. Also General Mladic ought to feel more than regret for the massacre: he ought to feel guilt for the massacre carried out on his initiative and under his order.
It seems to me that an account of morally appropriate emotional responses to collective wrongdoing needs to respect the proportionality of personal involve ment. If feeling guilt is an ‘internal sanction of the moral code’, an act of ‘self punishment’, its strength depends on being applied to relevant cases of personal wrongdoing or involvement in collective wrongdoing. Sanctions lose their moral impact when their occurrence is inflationary. Thus, the claim that guilt ought to be felt not only for one’s own wrongdoings but also for those of groups one belongs to may lower the value of guilt as internal sanction. Inflation of requirements to feel guilt can lead to a deflation of taking wrongdoing seriously.
On the other hand, it seems important that the required emotional response be a ‘reactive attitude’ targeting the wrongdoing, and not an emotion primarily focused on character traits.29 It has been repeatedly proposed that shame rather than guilt actually is the emotion that fits contexts of collective wrongdoing. Larry May, for instance, made such a proposal, pointing out to Jaspers’s account of collective guilt feeling. May is seeking less accusatory forms of the required emotional responses, especially for cases of collective inaction: ‘Those who have not actually caused harm may often find that shame rather than blame or punish ment is the proper category to describe what should follow from their sharing responsibility for various harms’(May 1992, 51).
Also Peter Forrest sees the basis of collective guilt in shame when he claims that ‘the feeling component of collective guilt just is the sum of the feeling components of the individuals who feel shame over the events concerned’ (Forrest 2006, 152). Without joining this discussion in a detailed manner, I will only mention the fact that shame in principle is not an action- but a character-directed emotion. The function of shame is different from the function of feeling guilt and so is the typical behavior correlated to these emotions. In cases of collective wrongdoing, the need is for an emotion that assesses in the first place the wrong acting and the harm it did. Therefore we require an emotion such as guilt or regret with their associated behavior towards apology and reparation.
Membership matters and so does emotional capacity. Displaying regret on behalf of wrongdoings committed by one’s group, one manifests that one cares both for the victims of the wrongdoings and for the integrity of one’s group. The emotional ties to the group provide a motivational force for engaging in efforts to restore the damaged group integrity:
More defensible (due to the less stringent entailment of personal control in regret than guilt) is the notion of collective regret, whereby the members of a group experience misgivings, dissatisfaction, or distress of mind over the acts of other membersoftheir group. [. . .]The notion of collective regret makes sense also insofar as the self is implicated by virtue of the regretted matter’s being something you care about (Landman 1993, 54f).
The fact that moral regret is a more ‘distanced’ feeling than guilt does not imply that its motivational force is weak. The distance of the feeling does not reflect the subject’s moral indifference to the wrong, but reflects the subject’s lack of influ ence in the wrongdoing. The fact that regret feelings lack the ‘bite’of remorse and guilt feelings does not imply that they are less efficient emotional states. The one deeply regretting a situation can be equally afflicted, and, therefore, equally engaged in reparative action as the one feeling guilt for it.
It has been objected that regret– not being conceptually linked to responsibil ity– is too weak a requirement to assess collective wrongdoing. Not requiring a collective to feel guilt over its wrongdoing would amount to depreciate the suf ferings of the victims. It would amount to no one taking responsibility for the wrong and, therefore, no one being obliged to apologize or engage in reparative acts. If all members of a group liable for a wrong had only feelings of regret but none of guilt, the wrong would not be assessed as one for which the group (or its members) were responsible. The objection is apposite: escaping from responsibil ity is a serious matter. On the other hand, undue inculpating is an equally serious matter. I think that the concept of moral regret can meet the challenge of sufficient strength. If we feel moral regret on behalf of an action, our feeling does not only assess the performance or the outcome of the action but also its moral character.31 Regret is caring about the wrong that has been done and implies, therefore, the idea of responsibility. I do not think that the aforementioned advantage of regret, namely the conceptual independence of its subject from the agent involved in its content, should be turned into an argument against its moral relevance. Even if the responsibility implied in moral regret is not necessarily self-responsibility, this does not mean that in regret no responsibility is acknowledged. The pure fact that moral regret is felt indicates the awareness of some responsibility.
As already mentioned, the case for regret concerns only the claim for a necessary minimal requirement in a general account of morally assessing collec tive wrongdoing, and not the claim for a sufficient requirement in all such cases. The difficulty lies in the dilemma between escaping responsibility and inculpating the innocent. I presume that in view of the heterogeneity of collective wrongdoing a general account of the appropriate emotional response to collective wrongdoing cannot be given in necessary and sufficient terms. I therefore prefer to confine a general account to the requirement of a modest, but stable necessary basis of an emotional response instead of requiring a possibly sufficient but exaggerated emotional performance. I hold with Janet Landman that feeling guilt augments feeling regret:
In general, it seems impossible to imagine experiencing guilt without regret, but quite possible to imagine experiencing regret without guilt. [. . .] [R]egret is not limited to instances in which there is legal, moral, or psychological culpability [. . .]. Further more, regret [. . .] includes certain acts and omissions of others with whom one shares group membership (Landman 1993, 56). [what about guilt feeling about original sin?]
Feelings of moral regret constitute the basis for feeling moral guilt, since they imply the idea of responsibility. They fulfill the positive function to constrain those who experience them to act and speak in ways that conform to an improvement of the regretted state of affairs.
Simultaneously, regret feelings do not entail the emotional subject’s culpability and are therefore not a form of self-punishment. This is an advantage over guilt feelings, which, as internal sanctions, are unwelcome and often tried to be avoided by all kind of strategies. In order to suppress guilt feelings, people may deny responsibility for a wrong action, or deny that it was wrong. Another well known strategy used to soften the burden of guilt is to emphasize the blameworthiness of the victims which is declared worse than one’s own.
As a result of practicing such strategies, guilty people often appear in the end as the ‘true victims’. One can imagine that such defense mechanisms are even more efficient in cases of guilt feelings that lack a basis of actual culpability. Nobody likes to be punished, even less so for things one is not to be blamed for. Therefore, claiming that collective guilt feelings are a moral requirement may turn out to be counterproductive. In defense against such feelings, members of eco nomical, political or ethnic groups may practice strategies of denial over years, thus preventing efforts to improve the situation. If citizens ought to feel guilt as a body for deeds they neither intended to nor actually did commit, defense strategies against such a constraint could take over, even to the point of not acknowledging at all that wrong has been done.
I conclude that regret is the better candidate than guilt for the role of the required emotional response in a general account of assessing collective wrong doing. Regret, on the one hand, respects the distance between the feeling subject and the agent involved in the feeling’s content, and, on the other hand, manifests the subject’s personal care for the regretted state of affairs. Regret does not entail undoing, but is ‘phenomenally associated with undoing’ (Landman 1993, 56). People regretting p usually do and say things that echo or conform to their feeling. Expressions of collective regret ‘We regret that p’are important because they show that responsibility for wrongdoing is acknowledged and that there is a will for change and reparation. The best that can realistically be reached in situations of collective blame is a sufficiently high number of individuals displaying sincere we-feelings of regret, motivating them to join their forces in actions that improve the given situation.