Official apologies as reparations for dirty hands

Abstract

The problem of dirty hands is, roughly speaking, concerned with situations in which an agent is faced with a choice between two evils so that, no matter what they do, they will have to violate something of important moral value. Theorists have been primarily concerned with dirty hands choices arising in politics because they are thought to be particularly frequent and pressing in this sphere.¹ Much of the subsequent discussion in the literature has focused on the impact that such choices have on a well-functioning democracy and how, if at all, we can ensure that a politician dirtying their hands does not undermine core democratic values and processes. A particular concern has been whether dirty-handed politicians ought to publicly reveal their actions after the fact and whether they should be held accountable through some form of punishment. The first aim of this paper is to point out that the focus on what is required in the wake of a dirty-handed decision, based on the need to protect the democratic system ought to be balanced with a genuine concern for reparative justice and what is owed to the victims of dirty hands. In the second part of this paper, I then examine the suitability of one common way in which perpetrators partly repair the damage they have done to victims, namely apologizing, in the dirty hands context.

The problem of dirty hands is a particular kind of moral conflict. Moral conflicts arise whenever an agent is faced with a choice between two morally valuable options which can be performed separately but not jointly. As a result, the agent is forced to forego one of the moral values but does so for good moral reasons, that is, to pursue the other morally valuable option. Having to choose between these two valuable options results in a moral remainder because, even if they choose the all-thing-considered better one, the lesser option has a value that cannot completely be made up for by the chosen one. This remainder is associated with a corresponding negative emotional response such as regret or remorse.² The problem of dirty hands is concerned with a particular subset of such moral conflicts, namely those that result in a grave moral remainder.³

The theoretical example given most frequently to illustrate this problem is Michael Walzer's ticking bomb scenario (1973). Walzer asks us to imagine a morally good politician who genuinely wants to act in the best interest of her citizens. Unfortunately, she is confronted with a situation in which a suspect in custody is refusing to give up the location of a number of bombs hidden in the capital city. Were they to explode, hundreds of innocent citizens could die. The politician now has to consider whether she ought to authorize the torture of the subject in order to extract the information from them. She is torn between upholding the ban on torture on the one hand and protecting her citizens on the other. Whatever option she chooses, she will get her hands dirty,⁴ and her choice generates a moral remainder and negative emotional response termed by Stephen de Wijze (2004) as “tragic-remorse.”⁵ Even if she chose the lesser of two evils, she will feel anguished and polluted.

How do dirty hands theorists explain this remainder, and the fact that an agent may feel a form of remorse for having done what they considered to be the all-things-considered best option available to them? To understand this, it will be helpful to take a brief look at Michael Stocker's (1990) analysis of dirty hands situations. In the ticking bomb scenario, it is wrong to torture people and it is also wrong to let hundreds of innocent people die when one could prevent it. In this hypothetical scenario, however, these two requirements are mutually exclusive; what we are faced with, according to Stocker, are “impossible oughts” (1990, p. 13). There are two moral requirements that the politician ought to follow, but she can, in this instance, only do one or the other but not both. Let us assume that she now decides, after careful deliberation, that it would be the lesser evil to torture the suspect. The wrongness of letting innocent people die if she could prevent it then turns into the “action-guiding” (1990, p. 11) consideration, while the wrongness of torturing people in this case becomes the “non-action-guiding” (1990, p. 13) one. The former is the reason that pushes her to take a certain course of action, while the other one, even though accepted as a good moral reason, is not acted upon. It is at this point, Stocker argues, that most moral theories stop. One course of action has been identified as the lesser evil and is therefore the right thing to do; end of story. As long as the politician has chosen the lesser evil, she should wholeheartedly commit to it and embrace her action as the morally right one. This, however, overlooks what now happens with the non-action-guiding feature; because she sees the wrongness of torturing people as a genuine moral demand she “double-count[s]” it; “the dirty feature is taken into account once in determining the overall value of the act and again on its own” (1990, p. 12). While she has identified the lesser evil in the situation, she in this way gives due weight to something she takes to be a genuine moral requirement; she is taking the wrongness of torture seriously and it becomes a remainder that attaches to her evaluation of the course of action taken. This remainder is then the source of a strong negative emotional response. Such an approach emphasizes that she ought to react with due consideration to every feature of the situation, whether she acted upon it or not. This focus on the non-action-guiding side, for Stocker, better reflects the complexity of our moral experience than a theory that concentrates simply on “action-guidingness.” The problem of dirty hands therefore illustrates the tragic possibility of situations in which it can be wrong to do what is right and in which we can dirty our hands by doing what we ought to do.

For the purposes of this paper, I will not be able to take up a more sustained defense of the dirty hands analysis of such conflict scenarios; instead, I will simply assume that it presents us with an accurate reflection of our moral reality.⁶

Let us illustrate these ideas with a real-life example. While I acknowledge that real-life examples will be a lot messier than hypothetical ones, and that we are unlikely to have all of the relevant information available, they can help us focus our attention on the morally salient features of the problem of dirty hands that merit investigation. Some of these features may be easily missed when confining our attention to more commonly used theoretical examples, such as the ticking bomb scenario, in which we focus on abstract moral agents rather than real human beings whose lives are impacted by political decisions.

On the evening of 18th June 1994, several locals congregated in Heights Bar, a small establishment in Loughinisland, Northern Ireland to watch the opening match of the football world cup between Ireland and Italy, when two gunmen from the Ulster Volunteer Force (UVF) entered the bar with automatic assault rifles and opened fire. Six men between the ages of 34 and 87 were killed in the attack; all of them were Catholics. The next day, the then Secretary of State for Northern Ireland, Sir Patrick Mayhew gave an interview in which he condemned the attack and promised that the assailants were “going to be caught sooner or later. The RUC [Royal Ulster Constabulary] never give up” (Gibney, 2017). A few days later, local policemen attended the wake of one of the victims and promised the family “we will leave no stone unturned until we get the perpetrators of this” (Gibney, 2017). The message to the public and those directly affected by the attack was therefore clear: officials would do everything in their power to identify and prosecute those responsible for the attack.

Local police showed a good initial response to the incident: they sealed off the crime scene, put out vehicle checkpoints and used a helicopter to surveil the surrounding area. The morning after the attack, the fully intact getaway car was discovered by a local farmer in a nearby field. Because of a mechanical error the car had been pushed off the road and the two gunmen and the getaway driver escaped on foot across the field. This was a very lucky find because standard practice in the UVF would have been to burn the car in order to destroy evidence. Police did, however, fail to forensically examine the surrounding field for further evidence and, curiously, in early 1995 the car was disposed of by the police. Later that day the police also received important information from Special Branch, the unit tasked with investigating terrorist activities and handling police informants. They told senior investigators the names of five individuals who formed part of a particular UVF cell they believed were likely responsible for the attack. Three of them had been seen talking together earlier on the day of the shooting. Less than a day after the attack the police had the names of a handful of suspects and had found the getaway car, yet no attempts to bring in any of the suspects for questioning were made that day.

A few weeks later police discovered the rifles that had been used in the attack in a field close to where the car had been found. They also recovered a bag containing gloves, handguns, balaclavas and, most importantly, a hair follicle found inside one of the balaclavas. Throughout July and August, the RUC brought in a series of suspects for questioning. Unfortunately, we do not know anything more about these interviews because it is believed that all transcripts were destroyed in 1998 because of an asbestos contamination. What we do know, however, is that on the night before the August arrests the suspects were warned by a police officer of their imminent detention. In the end, none of the suspects were charged. Six months after the attack, an anonymous caller contacted the police stating that they knew the identities of the people involved in the Loughinisland Massacre. The names they provided matched up with those provided to police by Special Branch months earlier. The call was followed by an anonymous letter sent to a local councilor who forwarded it to the police as evidence. Local police knew the identity of the anonymous source because they recognized her voice on the phone; she was the wife of one of the suspects. She was brought in for questioning but was later on released without any charges despite her claims that she had been involved in planning the attack. After a while, the families of the victims stopped receiving regular updates from the police about what was being done to solve the crime, and the investigation petered out. Despite the initial assertions of both the Secretary of State and the local police force that everything would be done to apprehend the attackers, the case remains unsolved to this day.

As a result of the continuous pressure of the families, in 2016 the then Police Ombudsman for Northern Ireland, Michael Maguire, published a report on the massacre.⁷ Maguire states in it that “the failures to bring the killers to justice cannot be explained solely […] by a failure or otherwise of investigative actions” (Maguire, 2018, p. 139) and that he has “no hesitation in unambiguously determining that collusion is a significant feature of the Loughinisland murders” (2018, p. 7). According to him “a legitimate suspect, was not only an informant for the RUC at the time but continued in this role for a number of years after the Loughinisland attack” (2018, p. 118) and he could “only conclude that the desire to protect informants may have influenced policing activity and undermined the police investigation into those who ordered and carried out the attack” (2018, p. 138). One of the police officers who worked on the case concurred in this assessment, stating that “you'd have to say someone was being protected” (Gibney, 2017).

Maguire identifies a moral trade-off that had to be made in the messy context of a violent conflict. On the one hand, the use of informants can help prevent atrocities and save lives. On the other hand, in order to procure the information vital to preventing future attacks, informants have to be an integral part of their respective organizations, which likely entails a sustained engagement in the group's criminal activities. Public officials therefore have to weigh up the value of the future lives that could potentially be saved by protecting the informant with the value of bringing the perpetrators of crimes to justice. When public officials⁹ make such a trade-off, they will inevitably get their hands dirty.¹⁰

So, what should be done in the wake of a dirty-handed decision such as the one described above? The dirty hands literature has focused in particular on the question of whether officials should reveal their dirty hands to the public after the decision and whether we should hold the official accountable, understood in a narrow retributive sense, by punishing them in some form or another.

We can roughly carve up the debate on truth and accountability into three positions. First up we have a position that requires public officials to reveal their dirty hands, who then ought to be held accountable through some form of punishment. S.L. Sutherland argues that unless officials reveal their dirty hands, there will exist “an unbridgeable gulf between those who lead and those who are led,” turning the public into “a passive collection of individuals” (Sutherland, 1995, p. 483) instead of a meaningful political actor, which excludes “the prospect of the reciprocity between citizens and leaders” (1995, p. 486). She adds that to legitimize political actors avoiding this process of accountability would be to disregard the rule of law, and the checks and balances that the public has to limit the power exercised by the officials that act for them. She concludes that in cases of dirty hands situations in democratic societies, we need to ensure that “public judgment takes the place of the autonomous phase of self-judgment” (1995, p. 491) of the dirty-handed agent.

There is a further position that could be taken that agrees that officials ought to make public their dirty hands, but that they should not be held accountable through a form of punishment for their actions. While a phase of accountability in the sense of explaining their behavior seems to be necessary to uphold the democratic process, it might be thought that to actually hold officials liable, for example, by punishing or sanctioning them, would be both counter-productive and immoral. One might think that if the official is able to explain herself, the need to take any further action and to hold her liable simply vanishes. Tamar Meisels argues that when a public official has dirtied his hands for us and in our names, the act of holding him liable is morally dubious; “punishing him for what we ourselves would have wanted him to do is no longer an irony or a paradox; it is simply wrong” (Meisels, 2008, p. 173). She goes on to say that if an official can successfully show that she acted under the constraints of necessity (put on her both by external circumstances as well as by the demands of citizens), she should be excused for her action. Not excusing her and, indeed, punishing her is for Meisels “analogous to the indefensible case of punishing the innocent” (2008, p. 173). When an official can give a satisfactory explanation for her dirty hands there is, according to Meisels, no reason for us to punish her.

Finally, there are those who think that a dirty-handed public official should not reveal her actions to the public. Naturally, this position also means that the official should not, and in fact could not, be held accountable for her actions. These accounts put forward the view that dirty-handed officials, instead of revealing their guilt, should be wearing “clean gloves” (Bellamy, 2010) in public. To wear clean gloves is not to admit to one's dirty hands and to keep them secret. Bellamy in his examination for the need to wear these clean gloves argues that “political legitimacy […] depends on both parties [i.e. officials and citizens] keeping faith with ideal liberal democratic forms and norms even while selectively disregarding them. […] So [officials] have good reason to be cautious” (2010, p. 426). Bellamy sees the nature of politics as inherently conflictual and follows from this that democratic practice can never live up fully to the liberal ideals on which it is built. In order for the basis of democratic politics not to be undermined by this, it is therefore important for all the parties involved to keep up the faith in these ideals even though we know that they can never be fully realized. Revealing dirty hands could pose a dangerous disruption to democratic politics because the public official is torn between a non-ideal political reality and the idealistic picture held by the public that requires the official not to show anything that is ordinarily conceived of as a vice. Taken to the extreme, this position could argue that there are certain dirty-handed actions that should never be revealed (e.g., because keeping them secret is crucial to national security). I find this claim unconvincing, though. Even if the secrecy of a dirty-handed decision is, and remains for a long time, politically necessary, ultimately circumstances will change sufficiently so that the information can safely be revealed. There will come a point at which it is appropriate for the clean gloves to come off. It does mean though, that in some cases, so much time will have passed that the dirty-handed action will have turned into a historical injustice. The decision to keep dirty hands secret for an extended time presents the agent with an additional dirty hands problem; their hands will be doubly dirty.¹¹

What all of these three positions have in common is that they are mainly concerned with the way in which truth-telling and accountability either uphold or endanger important democratic and procedural values. In the case of accountability there also seems to be some concern about what the dirty-handed agent does or does not deserve given that they were damned if they did and damned if they did not. However, thinking back to the cover-up in the wake of Loughinisland massacre there seems to be something fundamental missing in these discussions, i.e. a concern for the victims and what may be owed to them. When the barman Aidan O'Toole, for example, recounted the events of the night in 1994 over 20 years later, he was still on the verge of tears. One of the survivors of the massacre, Brendan Valentine, painfully recounted how “for twenty years on […] you put the head on the pillow at night and you reminisce about things. Never goes away. You were living and these critters were dead” (Gibney, 2017). At the reading of the Police Ombudsman Report the lawyer of the Loughinisland families noted that it had been “twenty-two years that these families have waited patiently, with dignity, they have campaigned ferociously. Everybody knows the pain and trauma that you have gone through” (Gibney, 2017). What is absent in much of the discussions in the dirty hands literature is a sustained engagement with what kind of reparations are owed to the victims and those that have been left worse off for the harm that they have had to endure.

Of course, truth-telling and bringing perpetrators to justice can be important steps in achieving reparative justice.¹² The reason why I think that, despite debates on these two issues, the dirty hands literature has not sufficiently engaged with the idea of reparative justice is that, when talking about the need for officials to reveal their dirty hands and be subject to punitive measures, the focus appears to be less on restoring dignity and equal standing to the victim and more on the upkeep of democratic procedures. The focus is therefore more on the dirty-handed agent and procedural values rather than on the victims. Because reparations are marked by a specific concern for the victim, current discussions on dirty hands have not given sufficient consideration to reparative justice explicitly. This is not to say that the work done on both truth-telling and accountability in the wake of dirty hands is not important. There are clearly interesting philosophical problems to be solved in this area; the claim of this paper is simply that we should balance this by giving some attention to those that are negatively affected by the dirty-handed decision and the question of what we can do to repair the harm done to them. In fact, as I will show in the final section of this paper, there are interesting philosophical puzzles to be found in this part of the dirty hands world as well.

There are a handful of mentions of reparations (beyond truth-telling and punishment in the above-mentioned sense) or restitution, compensation and apology more specifically,¹³ throughout the dirty hands literature, but these tend to be brief without offering any sustained discussion. Martha Nussbaum states that “the recognition that one has ‘dirty hands’ […] informs the chooser that he may owe reparations” (Nussbaum, 2000, p. 1009) because her conduct, “while in a sense inevitable, was also unethical” (2000, p. 1017). Cristina Roadevin puts forward a similar argument and concludes that “this is why [the dirty-handed agent] owes reparations to the victim” (2019, p. 131). Later on in her paper she elaborates on this by saying that repairing the harm to victim may, for example, be done “by offering compensation” (2019, p. 136). Peter Digeser also employs the language of compensation when he argues that, in response to a dirty-handed action “justice demands compensating the victim and responding to the actor who has done wrong” (1998, p. 709), where the latter ought to take the form of punishment. Steve de Wijze states that, as part of an official's process of atonement for a dirty-handed act, they may have to pay reparations to the victims. Reparations, according to him, respond to two distinct issues: “the material harm done and the moral wrong committed” (2013, p. 891). In an earlier paper he has also talked about ensuring that “there is restitution to those harmed” (2004, p. 468). With regards to reparations, he acknowledges two problems when considering the dirty hands context. Firstly, he argues that while it is important that dirty-handed officials are held to account, in some situations revealing their dirty hands may not be appropriate until long after the violation occurred. As a result, victims may be “subject to a further injustice by having their legitimate claims forgotten or ignored for a long period” (2018, p. 143). Secondly, because public officials act for us and in our names, citizens may be, to a certain extent, complicit in their actions. There is therefore a further question to be asked about the way in which the cost of reparations ought to be spread to citizens as well. Neil Levy takes up this last point and states that, because citizens share in the responsibility for the actions of the officials that represent them, “society should compensate the wronged, if possible” (2007, p. 50). The idea of an apology as a form of reparation for dirty hands is only taken up on very few occasions in the literature. Roadevin asserts that a dirty-handed agent may have a moral obligation “to apologize to the victim” (2019, p. 136). In his discussion of the way in which dirty-handed agents ought to atone their actions, de Wijze considers apologies as one of the elements in this process: “atonement occurs when the offender undergoes expiation for her wrongdoing which involves four stages: repentance, apology, reparation and penance” (2013, p. 890).

While there seems to have been some limited acknowledgement of the need for reparative justice in the wake of dirty hands, discussions have stayed at a very superficial level so far. In the remainder of the paper I want to look at the suitability of apologies as one way of making partial reparation to the victims of dirty hands because, as I will show, they present us with a particularly interesting puzzle.

Before going into the nature of apologies, let us briefly consider who we think could legitimately apologize for a dirty-handed decision. Going back to the Loughinisland case will help us consider some of the possibilities. One option would be for individual police officers to apologize for any actions they took within their official capacity that enabled the covering up of the involvement of a police informant (e.g., warning the suspect of their arrest, not following up a particular lead, not asking certain questions during an interrogation). While such individual apologies would most likely not be without meaning for the victims and their loved ones, they cannot, in and of themselves, account for and respond to the entire wrong committed and harm done to the victims. Instead, such an apology only covers the actions of a single individual within the group that enabled the implementation of the dirty-handed decision. We might therefore think that an apology should be offered, either in addition to or instead of any individual apologies, by an authorized representative of the group that, as a whole, made and implemented the dirty-handed decision. Michael Murphy (2011, p. 50) has argued that the choice of representative will be crucial in shaping the meaning of the apology because it signals how seriously it is being taken by the group in question and how much respect it is willing to accord to the victims of injustice. In the case at hand, the most likely candidate to offer such a representative apology would be the lead investigator.

The lead investigator could apologize not only for the particular actions that they took within their official capacity, but for the decisions and actions of their team as a whole that enabled the involvement of a police informant to be covered up. An apology on behalf of the whole investigative team would signal to the victims an acknowledgement that what happened to them was not an isolated incident caused by the actions of a single individual but a consequence of the calculated decisions of a whole group of police officials. An apology in the name of all of those who were directly involved in the implementation of the dirty-handed decision may be able to send a stronger signal to the victims and their loved ones that what happened is being officially acknowledged and dealt with than in the case of a single police officer apologizing only for their own actions.

There is, however, a lingering worry in the Loughinisland case that even an apology from the lead investigator on behalf of the whole investigative team would not get to the root of things. It is unclear whether the decision to protect the informant in this case was made by the lead investigator or by someone higher up in either the police or government. If that was the case, victims might rightfully expect those in political power to acknowledge and apologize for their involvement as well. And even if, in this particular case, the decision had been made by the lead investigator alone, we might expect that a representative of the UK government should recognize and apologize for the sometimes devastating consequences of their overarching policy on the use of police informants during the Troubles. Murphy (2011, p. 51) has argued that it is important that injustices are recognized both at the micro (e.g., the suffering caused by leaving survivors and families of an atrocity in the dark about what happened) and the macro level (e.g., the systemic issues surrounding the use of police informants in a violent conflict). In the dirty hands case in question here, this macro level of the systemic and institutional nature of the problem can only be fully acknowledged and apologized for by the government itself.¹⁴

Who ought to apologize in the wake of a dirty-handed action is not straightforward and, ideally, may require a variety of actors and institutions to apologize for the part that they played in the wrongdoing. One of the most important considerations in who ought to apologize is listening to what the victims and their loved ones require to feel that the fractured relationship between them and those that got their hands dirty can, at least partially, be repaired. Murphy (2011, p. 51), for example, suggests that it can often be helpful to include the victims, survivors or their close ones in this process to negotiate the apology because it is a sign of respect and reaffirms their agency by engaging in mutual dialogue.

Before we can discuss the suitability of apologies as a form of partial reparation in the wake of a dirty-handed decision, we need to consider what is required for something to constitute an apology. Ernesto Verdeja has argued that the elements of an apology can be, broadly speaking, put into two different categories; one is that of “moral redress” while the other is that of “practical redress” (Verdeja, 2010, p. 567).¹⁵ The extent to which an apology provides moral redress depends on whether the perpetrator, in a manner appropriate to the context, acknowledges the harm inflicted on the victim, accepts responsibility for that harm, and expresses remorse for having been the source of that harm, thereby acknowledging the victim's rightful claim for respect and reaffirming their moral standing in the community. This is what is required for us to call a particular act an apology. Let us imagine that I am walking on the pavement with a cup of coffee in my hand when a reckless cyclist turns a corner without looking and crashes into me, causing me to spill my coffee. Now the cyclist says: “I'm really sorry about bumping into you and making you spill all your coffee. This was totally my fault; I should have been looking more carefully where I was going. I feel really bad about this. I hope you're ok!” This utterance would appear to satisfy the requirements necessary to be called an apology.

Usually, however, we are looking for more than an utterance that provides moral redress; we also want to see a form of practical redress by which we can judge the sincerity of the apology. Without such a forward-looking part, apologies will often ring hollow. The extent to which an apology is accompanied by suitable practical redress depends on whether the perpetrator, in a manner appropriate to the context, aims to compensate the victim, vows not to commit a similar transgression in future and works on transforming themselves or the relevant institutional set ups, thereby repairing some of the material harm and providing reassurance to the victim. In the case above, for example, the cyclist may offer to buy me a new coffee and promise to be more careful in future—which appears believable because they leave the scene of the accident using the designated cycle lane rather than the pavement for pedestrians. Were the cyclist to ride away from the scene of the accident just as recklessly as before, we would probably judge their apology to be hollow and insincere.¹⁶

When we think about using official apologies as a reparative response in the wake of a dirty-handed decision we are faced with a complication, though. While dirty-handed agents take responsibility for their actions and experience “tragic remorse,” they did the best they could, given the constraints in which they were acting. After all, they chose the lesser evil. As a result, the dirty-handed agent cannot promise that they would not commit a similar harm in future, in fact they may even be committed to doing exactly the same again should the tragic occasion arise. de Wijze points out that in a normal apology we would expect the offender to “acknowledge her wrongdoing and simultaneously disown it thereby dissociating the true self from the guilty self” (2013, p. 891). Does this mean that official apologies in the wake of dirty hands are doomed to be insincere because they do not offer the required practical redress? After all, if the cyclist in the previous example were to say to me: “I'm really sorry about bumping into you and making you spill all your coffee. This was totally my fault; I should have been looking more carefully where I was going. I feel really bad about this. I hope you're ok! But, just to let you know, I'd do the same again in future,” we would think that the cyclist's apology is not actually sincere.

In dirty hands scenarios, however, I do not think that this has to be case. The reason why we usually think that someone who fails to promise not to reoffend is insincere in their apology, is because we take this as evidence that the remorse they expressed was not genuine. It relies on the assumption that it is impossible to feel, what D. Z. Phillips and H. S. Price have called, “remorse without repudiation” (Phillips & Price, 1967) about the act in question. And in the case of the cyclist, this assumption appears to make sense; when they tell us that they are sorry but they will continue to drive recklessly in future, this appears to suggest that the remorse they expressed in their apology is not actually sincere and that they do not genuinely care about the harm that they have caused. But in dirty hands cases, the story is more complicated than this. Once we allow for the possibility of situations in which it can be wrong to do what is right and in which we can dirty our hands by doing what we ought to do, we also need to allow for the possibility that the agent can be genuinely remorseful about the action they committed while at the same time taking the position that they would act in the same way in future, should a similar tragic situation occur. In dirty hands cases the fact that the agent is unable to promise not to reoffend in future is therefore not a sign that the remorse they express as part of their apology is not genuine. Apologies in the wake of dirty hands are therefore not doomed to be insincere, even though the agent is unable to promise that they will not act similarly again in future.

Now someone might object that while apologies are appropriate for the binary of morally just and unjust actions, it is simply not a concept suitable for more complex situations in which we can get our hands dirty by doing what we ought to do. Maybe what is more appropriate in such instances is a showing of regret or heartfelt sympathy (in Verdeja's words a kind of moral redress) combined with either restitution or a form of compensation (in Verdeja's words a kind of practical redress). After all, when an agent did the best they possibly could in the circumstances, is an alternative response such as this not more appropriate than a conventional apology?¹⁷

Once we take seriously the story that dirty hands theorizing tells us about the reality of moral conflict, I think it becomes clear that an alternative to an apology would not be a fitting response in dirty hands situations. When we acknowledge that we commit a genuine moral wrong when getting our hands dirty, emotional responses such as regret or heartfelt sympathy are not morally sufficient. As de Wijze has argued, “to feel mere regret about this state of affairs would fail to do justice to the serious moral violation” (2004, p. 464). Instead, a form of remorse (what de Wijze has called “tragic-remorse”) is the most appropriate response to acknowledge one's guilt in consciously choosing to violate a core moral value. If we accept this, we are firmly in the territory of apologies which require, as argued earlier, that perpetrators acknowledge the harm they cause, take responsibility and express their remorse for having caused the harm. I therefore believe that apologies, rather than some alternative course of action, are an appropriate response to getting one's hands dirty, even if one did the best one could in difficult circumstances.

This conclusion is further supported if we consider the way in which apologies for dirty hands are able to fulfill the full range of social functions that we would expect more ordinary apologies to realize. Firstly, an apology for the harm caused as a result of one's dirty-handed actions can fulfill important functions for the perpetrator. Now we might think that this should not be our primary concern, especially given the purpose of this paper is to shift some of the focus from dirty-handed agents to the victims of their actions. For completeness' sake, however, it will be useful to note that the function more ordinary apologies can play for offenders, they can fulfill in the case of dirty-handed agents, too. Apologies can help restore at least some of the offender's moral integrity and enable them to reclaim their standing in the moral community by acknowledging their transgression and reaffirming the common moral values we hold dear (Gill, 2000, p. 17). Additionally, an apology can help the offender to relieve some of their feelings of guilt and shame for what they have done by reassuring them that they have taken action, both symbolic and practical, to help those that have been left worse off as a result of their action (Lazare, 2005, pp. 107–113). That an apology can play these functions for the apologizer may be thought particularly important in the case of dirty hands scenarios in which the agent was limited to a choice between lesser evils and tried to act the best way they could give the limited options available to them.

Secondly, and more importantly, an apology in the wake of dirty hands can fulfill important functions for the victims. Nicholas Tavuchis has argued that the most important function of any kind of apology in the public sphere is that it can set the official record straight (Tavuchis, 1991, pp. 70–72). This is particularly important in the case of dirty hands because they will often involve secrecy, lies and deception (Thompson, 1987, p. 32). Such means can result in a violation of the democratic process and a secondary dirty hands problem. Take the following example: in 1993, it was leaked that Prime Minister John Major had secretly been in contact with the Provisional IRA over the ongoing conflicts in Northern Ireland. At the time, these negotiations were part of a wider attempt to end the continuing violent struggles in the country and bring peace. We can assume that, in the first instance, Major had to make a decision between the conflicting values of not negotiating with a paramilitary organization on the one hand, and doing everything in his power to facilitate the peace process on the other. What was so interesting about the leak at that point in time was that it came only weeks after Major had declared to parliament that talking to the Provisional IRA “would turn my stomach” (Bevins et al., 1993). In addition to the primary conflict he faced, Major also had to make the following choice: either he was to publicize these negotiations and risk upsetting the delicate balance that had been struck, or he was to keep them a secret while at the same time putting on the mask of someone who wholeheartedly opposed negotiating with a paramilitary group. Secrecy, lies, and deception such as this will frequently accompany dirty-handed decisions in democratic politics. The apology of a dirty-handed actor can therefore fulfill the important function of setting the record straight, and of making the denial or minimization of the harm inflicted on victims more difficult to achieve (Murphy, 2011, p. 55). As a result of this it can also help to afford victims the status of truth-tellers (Harvey, 1995, p. 62) by acknowledging that their version of events accurately reflects what really happened.

Another vital function of apologies is that they can provide reassurance to the victims and the wider public that the offender is committed to our common moral and societal norms and vows to uphold them to the best of her ability. This will also give some assurance that they are unlikely to commit similar transgressions in future (Gill, 2000, pp. 16–17). Apologies in the wake of dirty hands can most definitely fulfill the first part of this function. When a dirty-handed agent apologizes for the harm their actions caused, they can use this as an opportunity to express their heartfelt tragic-remorse, which shows a commitment to the forgone value. The experience of tragic-remorse is a sign that the agent is giving due consideration to the value that was lost out on because it reminds them of the different moral requirements that were at stake. Things are, as mentioned earlier, slightly trickier when it comes to reassuring others that the dirty-handed agents will not commit similar actions in future. Because she chose the lesser evil, she cannot promise not to do the same again, should a similar tragic situation arise. Nonetheless, I think that an apology in the wake of dirty hands can provide some important reassurance to the victims and the wider public. This is because by expressing her tragic-remorse, the agent can make her behavior intelligible to others. This idea can be found in Bernard Williams's account of the importance of emotions when he argues that “in some cases the relevant unity in a man's behavior, the pattern into which his judgments and actions together fit, must be understood in terms of an emotional structure underlying them” (Williams, 1999, p. 222). Without an apology that brings to light the details of the difficult choice the agent had to make, victims and the public may justifiably believe that the agent violated an important moral value without good reason, thereby showing a complete disregard for the particular norm in question. When, however, the dirty-handed agent apologizes and explains the nature of the situation in which they found themselves, and expresses their genuine tragic-remorse for what they had to do, they can evidence that they hold the moral value in question very dear despite violating it on this occasion. This can provide at least some reassurance to victims and the public in as much as they know that similar transgressions would only arise in a very limited set of tragic situations and are therefore unlikely. An expression of tragic-remorse can also indicate to the victims that the dirty-handed agent is paying, at least part of the price, for the harm that she chose to inflict on others (Lazare, 2005, pp. 53–54).

Apologies also reaffirm the moral standing and status of the victims. The original injustice was in effect a communication that the offender did not regard the victims as deserving of the same rights and respect as the rest of the moral community. While an apology is not able to undo the harm, what it can do is recall this implicit message. By acknowledging that the action in question was wrong, and by acknowledging the harm this caused to the victims, the offender recognizes that their victims deserve to stand on an equal moral footing to themselves and the rest of the community. Apologies can therefore restore some of the dignity to victims and can help them to recover a sense of self-worth and confidence (Gill, 2000, p. 16; Govier & Verwoerd, 2002, p. 69; Thompson, 2008, p. 34). When a dirty-handed agent apologizes she can exhibit that, even though the all-things considered better option meant violating the rights of their victims, she is nonetheless committed to seeing them as equal members in the moral community deserving of the same respect and dignity as everyone else.

Finally, an apology can also be a tool that supports victims in seeking practical redress for the injustice they have suffered (Murphy, 2011, p. 63; Nobles, 2008, pp. 29–30). By bringing public attention to the way in which important moral and social standards have been the violated, and to the harm done to the victims, as well as by putting the offender on the record as admitting responsibility, an apology will be able to strengthen the case for having to mitigate or make up for the injustice in tangible ways. As mentioned earlier, a variety of writers (de Wijze, 2013; Digeser, 1998; Levy, 2007; Roadevin, 2019) have hinted at a need for reparations in the wake of dirty hands, so apologies could play the important function of initiating a process of providing practical redress to those who have been harmed by a dirty-handed decision.

What I hope to have shown in this section is that, even though dirty-handed agents cannot make a promise not to reoffend, this does not entail that their apology has to be insincere, and that it can in fact realize the full range of social functions that we would expect a more ordinary apology to fulfill.¹⁸ Obviously, apologies are only one amongst many other measures that can advance reparative justice. Adequately responding to the harm caused by a dirty-handed decision may also involve, for instance, conducting official enquiries, paying compensation or establishing certain commemorative practices. What exactly is required will be dependent on the particular case in question.

The purpose of this paper has been twofold. First, I argued that the dirty hands literature has missed a crucial consideration in its discussions on what we ought to do in the wake of a dirty-handed action. In particular, I pointed out that by focusing almost exclusively on whether the agent ought to publicly reveal her dirty hands, and on whether we should punish her for her actions in order to uphold democratic and procedural values, barely any attention had been paid to what may be owed to those that have been harmed or left worse off as a result, and whether any form of reparation may be owed to them. Second, I explored the suitability of one common way in which perpetrators can partly satisfy the demands of reparative justice—apologizing—in the dirty-hands context. The worry was that if dirty-handed agents did the all-things-considered best they could, then they could not vow not to repeat the action in future which would make any apology for dirty hands, at best, insincere. I argued that this worry is misguided once we allow for the reality of genuine remorse without repudiation. I concluded by showing that apologies for dirty hands can fulfill the full range of social functions that we expect more ordinary apologies to realize. As such, apologies can be an important part of providing reparations to the victims of dirty hands. While it is certainly important to consider the way in which we can uphold and promote important democratic values and processes in the face of dirty-handed decision-making, I hope that this paper can open up a wider discussion about the need for a genuine concern about reparative justice in the wake of dirty hands.