From rehabilitation to penal communication: The role of furlough and visitation within a retributivist framework

Retributivism is one of the most prevalent theories in contemporary penal theory. However, despite its popularity it is frequently argued that too little attention has been paid to the implications of retributivism for prison management and prison life, including prison visits and furlough. More so, it has been questioned both whether the various forms of retributivism found in the philosophical literature on criminal punishment have anything to say about what prison life ought to be like and whether they are able to criticize deeply contested rules and practices, such as those that deny inmates contact with family-members for the sake of prison discipline. In this paper, we argue that prison visits and furlough have a crucial role in a prison system based on retributivist principles. In particular, we argue that the communicative theory of punishment has important theoretical resources for proving a strong and compelling rationale for both furlough and visitation on retributivist grounds. Besides exploring this rationale, we also discuss the practical implications of this view for the penal policy.


Introduction
There is a voluminous scholarship on retributivism and its implications for legal decision-making and criminal sentencing. Penal philosophers and criminologists have all contributed to the growing body of literature on retributivism, elaborating on its main concepts, including penal censure, crime severity and proportionality (du Bois-Pedain and Bottoms, 2019; Duff, 2001;Duus-Otterstr€ om, 2013;Tonry, 2011;von Hirsch, 1976von Hirsch, , 2017). Yet, despite its popularity among penal theorists and policy-makers, it is frequently argued that too little attention has been paid to the implications of retributivism for prison management (as well as other postsentencing practices, like parole) (Dagan and Roberts, 2019;Hayes, 2018;Kerr, 2019;Lippke, 2007;O'Hear, 2011). This includes prisoners' opportunities for visits and furlough. Despite being an attraction for extensive empirical scholarship (see e.g., Cheliotis, 2008;Cochran and Mears, 2013;Duwe and Clark, 2013;Mitchell et al., 2016), few penal theorists have discussed what role, if any, prison furlough and visitation has within a retributivist framework (one notable exception is Lippke, 2007).
Humanitarian reasons aside, the most frequently evoked justification of both prison furlough and visitation is that of offender rehabilitation (Boudin et al., 2013;van Zyl Smit and Snacken, 2009). Even if they are very different, prison furlough and visitation programs are mechanisms that enable inmates to maintain important social bonds with people on "the outside" during their imprisonment, which, in turn, can improve inmate behavior and reduce recidivism. Empirical studies suggest that visitation programs might decrease the risk of recidivism as well as the prevalence of prison misconduct (Cochran and Mears, 2013;Duwe and Clark, 2013;Mitchell et al., 2016) and have found that furlough has a positive impact on offender reintegration (Cheliotis, 2008). In general, allowing prisoners to retain meaningful contact with the outside world might help mitigate many of the harmful effects associated with prison life and social isolation (Cochran and Mears, 2013).
Despite these positive effects, it is not clear whether either prison furlough or visitation programs are compatible with retributivism. To the contrary, critics have questioned whether retributivism has anything at all to say about the type of rules and practices that deny inmates contact with family-members (for example, for the sake of prison discipline). For instance, Whitman (2003) argues that instead of criticizing such practices, the retributivist's language of blame is (and indeed has been) easily exploited in order to support deeply contested and problematic penal policies of this sort. More recently, Kerr (2019) argues that although contemporary forms of retributivism -such as those defended by Jean Hampton (in Murphy and Hampton, 1988) and Antony Duff (2001) -help identify a reason to punish, they say very little about the type of prison conditions that are morally justified, including the appropriate rules and practices surrounding visits and furloughs.
These observations might lead penal theorists, as well as policy makers, to abandon retributivism. Given the importance of furlough and visitation programs one should simply endorse rehabilitationism, or some other forward-looking theory of punishment. Yet, as we argue in this paper, this does not necessarily have to be the case. To the contrary, we argue that it is possible to provide a plausible and compelling justification for both furlough and visitation programs from a retributivist perspective. Our argument builds on the contemporary and influential version of communicative retributivism developed by Duff (2001Duff ( , 2003. In this view, criminal punishment, if justified, should be understood as a form of secular penance, aiming at repentance, moral reform and reconciliation. According to Duff (2001), punishment should ideally seek to engage with the offender as an autonomous moral agent and as a member of the political community. We argue that the communicative version of retributivism provides a compelling case for both liberal visitation programs and prison furloughs, in part because this is the type of prison regime that is most likely to serve the ends of secular penance, but also because of how visitation and furlough serve as important reminders of how the prisoner has not lost his or her status as a member of the political community. In developing this argument, we also discuss what the criteria for both prison furlough and visitation programs ought to be, given that the communicative theory is to be implemented.
The rest of the paper is structured as follows: In the subsequent section, we provide an overview of some of the central components in Duff's communicative theory of punishment, with an emphasis on its implications for imprisonment (The communicative theory of punishment). We then describe some of existing visitation and furlough programs in some Western jurisdictions (Visitation and furlough programs). These two sections set the stage for our analysis of visitation and furlough programs from the perspective of the communicative theory of punishment. In our normative analysis, we then explore the role of furlough and visitation from the perspective of the communicative theory as well as the practical implications of this view for policy (From rehabilitation to communication: reforming prison furlough and visitation). The paper ends with a conclusion (Conclusion).
The communicative theory of punishment Duff's (2001) theory, often labeled the communicative theory of punishment, has been described as 'the most powerful retributive theory of recent times' (Matravers, 2011: 68). It shares a basic assumption with several contemporary retributivist theories, namely that punishment is justified as deserved censure. However, unlike other forms of retributivism -such as the just desert view famously defended by Andreas von Hirsch (1976 -the communicative theory is distinct in its emphasis on how criminal punishment is largely a communicative enterprise. In Duff's terminology, criminal punishment should be understood as a species of secular penance: 'it is a burden imposed on an offender for his crime, through which, it is hoped, he will come to repent his crime, begin to reform himself, and thus reconcile himself with those he has wronged' (Duff, 2001: 106).
Understood as a form of secular penance, criminal punishment communicates to the offender the wrongfulness of his or her conduct. In doing so, it should ideally make the offender repent his crime, but also provide the chance for the offender to communicate this repentance to both the victim and to the wider society. A repentant offender can do so by taking the censuring message communicated by his punishment seriously through self-reform, and by his or her willingness to undergo an appropriate penitential punishment as a form of apologetic reparation to those he has wronged. These include both the victim and wider community whose values have been flouted (Duff, 2001).
There are several aspects of the communicative theory that are important to recognize here. First, while one aim of punishment, according to the communicative theory of punishment, is moral reform, this does not mean that the punishment should go on until this end is successfully achieved. In order to respect the wrongdoer as an autonomous moral agent, he should have the freedom to remain unpersuaded by the message of the punishment. Regardless of the prisoner's conduct, once he has completed his sentence, 'we treat him as (as if) reformed' (Duff, 2013: 190). Also, given how the justifying aim of criminal punishment is to communicate deserved censure, the amount and mode of punishment should also reflect the wrongfulness of the wrongdoing. Thus, a version of the proportionality principle is built into this view.
It is of special interest to our discussion in this paper to note that Duff's communicative theory is grounded in a broader political theory about the role of criminal justice in a community based on liberal values, such as autonomy, privacy and freedom (Duff, 2001). In such a community, Duff (2001) argues, punishment should be inclusionary and treat offenders as autonomous moral agents and as members of the community who, even though they have violated its core values, can be committed to and recommit to them. For this reason, he is also largely skeptical towards imprisonment as a mode of punishment. The most salient aspect of imprisonment, Duff argues, is how it excludes and removes the individuals from the normative community. It must therefore be reserved for the sort of criminal offender who through his criminal act 'has not just damaged or threatened, but has broken, the normative bonds of community' (Duff, 2001: 150;emphasis in original). To such an offender, Duff argues, his imprisonment communicates how he has 'made it impossible for us to live with him in the ordinary community of fellow citizenship unless and until he has undergone this penitential punishment' (Duff, 2001: 150). It should be observed, that it is not always clear what this means. For example, Duff does not provide an explicit list of the types of wrongs he believes merit imprisonment, but merely suggests that it 'depends on the community's understanding of what kinds of wrong make community impossible' (Duff, 2001: 151). Nor does he provide much detail for how common particularities of many prison regimes are integrated with the communitive theory of punishment (see Kerr, 2019). Instead, Duff merely suggests that 'there is of course much more to be said about what kind of prison regimes are suitable to this purpose' (Duff, 2001: 150). Still, it is clear that Duff's theory has important implications for how prison facilities are designed and run. If the aim of punishment is to persuade offenders about the wrongfulness of their conduct and thereby promote both moral reform and reconciliation, then prison should also be compatible with these ends. Of course, whether prison can serve these ends at all is not something that can be merely assumed. To the contrary, there is some empirical evidence suggesting that at least long-term prison sentences are at risk of falling short of achieving the aims for which, according to Duff, criminal punishment should be imposed in response to criminal wrongdoing. Rather than facilitating secular penance, long-term prison sentences can 'create emotional reactions and pressures to cope that overshadow any communication inherent in the sanction' (Schinkel, 2014: 593).
Like any other influential penal theory, the communicative theory has its critics. For example, it is sometimes argued that the aim of inducing repentance and moral reform is intrusive and inappropriate as an aim of state punishment in a liberal society (Bennett, 2008). We don't want to dismiss this criticism as irrelevant. Yet, we should say that we believe that the emphasis on how punishment can serve as a form of penance is what makes the communicative theory attractive. It suggests that we ought to treat fellow citizens as autonomous agents who can do wrong, but who can also take responsibility for their wrongs through suitable forms of apology and moral reparation (Duff, 2003: 303).
In what follows, we treat Duff's theory as a fruitful starting point for discussing the place of visitation and furlough programs within a largely retributivist framework and for addressing what an appropriate policy concerning visitation and furlough should look like. Before we continue this discussion, however, we will first provide a brief overview of the role of visitation programs and prison furlough in at least some legal contexts, focusing on the US and Europe in particular. Our aim here is not to provide an exhaustive list of all possible policies, but merely to provide a sense of the range of different existing policies.

Visitation policies
Even if visitation programs are part of almost any criminal justice system, their form might vary a great deal. In general, European prison law recognizes the importance of being able to maintain meaningful contact with people outside of prison. As van Zyl Smit and Snacken (2009) point out, prisoners' right to family visits relates to article 8 of the European Convention on Human Rights, which states that everyone has the right to have their private and family life and correspondence respected. The European Prison Rules (2006) further specify the obligations that prison authorities have to ensure that this right is respected in the otherwise inherently restrictive and coercive conditions of imprisonment (see Coyle, 2006). The European Prison Rules (2006) dictate that '[p]risoners shall be allowed to communicate as often as possible by letter, telephone or other forms of communication with their families, other persons and representatives of outside organisations and to receive visits from these persons' (r.24(1)) and that '[t]he arrangements for visits shall be such as to allow prisoners to maintain and develop family relationships in as normal a manner as possible' (r.24(4)). They also prescribe that '[p]rison authorities shall assist prisoners in maintaining adequate contact with the outside world and provide them with the appropriate welfare support to do so' (r.24 (5)).
Even if all European countries allow prisoners to have contact with their families and friends, we should bear in mind that the conditions under which visits take place vary enormously between jurisdictions (Coyle, 2006;van Zyl Smit and Snacken, 2009). Also, as van Zyl Smit and Snacken (2009) point out, it should also be recognized that even though family visits are encouraged because of their reintegrative as well as their rehabilitative potential, the conditions under which visits take place in European prisons may sometimes jeopardize rather than facilitate these positive effects.
Some of the most liberal visitation programs are found in Central and Eastern Europe and in Scandinavia. Here prisoners may receive prolonged and unsupervised visits from their spouses, partners and families, at regular intervals, for up to three days (van Zyl Smit and Appleton, 2018). In some other jurisdictions, family visits might be understood as a privilege. For example, in England and Wales, prisoners are entitled to receive a visit 'twice in every period of four weeks, but only once in every such period if the Secretary of State so directs' (English Prison Rules, 1999, s.35 (2)). However, the prison governor may allow a prisoner an additional prison visit 'as a privilege or where necessary for his welfare or that of his family' (English Prison Rules, 1999, r.35 (3)).
In the US, the importance of visitation programs is not matched by any significant constitutional protection. To the contrary, the American Supreme Court has held that prisoners' visitation rights might be very limited and that visitation privileges might be withdrawn for a limited period in order to effect prison discipline (Boudin et al., 2013). There is no constitutional bar, for example, to a policy under which prisoners with substance-abuse violations are disallowed any visitors for two years. In general, the American Supreme Court has noted that 'freedom of association is among the rights least compatible with incarceration' and upheld strict visiting regulations on the grounds that they had a rational relationship to 'legitimate penological interests' (Overton v. Bazzetta, 539 U.S. 126, 131 (2003); Block v. Rutherford, 468 U.S. 576 (1984); see also Boudin et al., 2013;Cochran and Mears, 2013).
In a study on prison visitation policy in the US, Boudin et al. (2013) note that American states vary widely concerning the length of visits allowed, as well as the number of times friends and family may visit within a given period. According to Boudin et al. (2013), some US states, such as Alaska, explicitly recognize that visitations are required to preserve strong family and community ties and increase the likelihood of the prisoner's rehabilitation and that visits should only be denied when necessary for public protection or to maintain order and security within the prison. Boudin et al. (2013) further note that twenty-eight American jurisdictions 'have a floor for the minimum number of days or hours of visitation that must be made available' (p. 161). In several states, there are also ceilings on visitation hours. One example is Oregon, where only one visit per day per visitor on weekends and holidays is allowed. Another example is Utah, where no more than two hours per visit per day is allowed. The perhaps most liberal visitation policy is found in New York, where up to six hours of visits 365 days per year as well as overnight visits approximately every two months, are allowed. In contrast, North Carolina is the state with the perhaps most restrictive policy, with a ceiling of no more than one visit per week for up to two hours (excluding legal and clergical visits) (Boudin et al., 2013).
Boudin et al. also note that because 'states uniformly consider visiting a "privilege", policies often limit prisoners' access to visitors as a sanction and may reward good behavior with greater access to visitation' (Boudin et al., 2013: 161; see also Mitchell et al., 2016). Moreover, in approximately half of US states, prisoners at certain security classifications will be subject to limits on visitations. In general, higher-security prisoners are allowed fewer visiting opportunities than minimum security inmates. One notable exception, however, is New York, which according to Boudin et al. (2013) is the only state that provides more visitation opportunities and more flexible timing of visits for prisoners in highersecurity settings. Boudin et al. (2013) notes that only a few of US states allow for overnight family visits or prolonged visits.

Furloughs policies
Like visitation, furlough programs (often also called "temporary release" or "prison leaves"), vary between jurisdictions (Cheliotis, 2008;Du¨nkel et al., 2018). Some of the most liberal furlough policies exist in Germany. There, as part of "release planning" and German prison law "relaxation" policy, prison authorities may grant prison furloughs for release preparations, subject only to limitations if there is a risk for a future offence or escape. There is no upper time limit for such prison furlough program and, in theory, a prisoner may stay every weekend with his family if that is recommended for his or her reintegration process (Pruin, 2018). Recently, German law enabled up to six months of temporary release for certain kinds of prisoners (Pruin, 2018).
In Greece, furloughs are granted to prisoners who have served one-fifth of their sentence. The duration of prison leave is normally one to six days and in total up to 45 days per year. The general leave conditions include the obligations not to re-offend and to report to the police the prisoner's place of residence and requires that the prisoner has the capability to sustain him or herself during the temporary release (Anagnostaki, 2018).
The English Prison Rules (1999, r.9(3)) may allow temporary release of a prisoner so that he or she can engage in employment or voluntary work, receive training which cannot reasonably be provided in the prison, to assist the prisoner in maintaining family ties or to assist transition from prison life to freedom. Yet, a prisoner should only be released under such conditions if 'the Secretary of State is satisfied that there would not be an unacceptable risk of his committing offences whilst released or otherwise of his failing to comply with any condition upon which he is released' (r.9(4)).
Finally, American federal law allows, among other types, at least the following three forms of furloughs. (1) Day furlough is a 'furlough within the geographic limits of the commuting area of the institution, which lasts 16 hours or less and ends before midnight' and that is aimed to 'strengthen family ties or enrich institution program experiences' (U.S. Department of Justice, 2011, s.3(c) (1)). (2) Overnight furlough is a 'furlough which falls outside the criteria of a day furlough'. The lenght of an overnight furlough is ordinarily 3 -7 days and may only be extended for specific medical, educational or vocational reasons. (U.S. Department of Justice, 2011, s.3(c)(2)). (3) Social Furlough is a day or overnight furlough used primarily for purposes related to release planning, (re)establishing family and community ties, participation in religious, educational, recreational, civic activities (U.S. Department of Justice, 2011, s.3(d)).
Furthermore, under American federal law policy, prison authorities will not grant a furlough to a prisoner if he or she is 'convicted of a serious crime against a person' (U.S. Department of Justice, 2011, s.7(b)(1)). That is also the explicit policy in some US states. In Washington, for instance, any person convicted of the crime of aggravated first-degree murder shall be sentenced to life imprisonment without the possibility of release or parole. Such a prisoner is generally not allowed to participate in any sort of release or furlough program (Washington Revised Code, 2015: RCW 10.95.030).

From rehabilitation to communication: Reforming prison furlough and visitation rules
As we have seen already, prison furlough and visitation programs are frequently motivated by reference to offender rehabilitation and reintegration. Yet, while we share this commitment to rehabilitation, we should also recognize that it is not the only aim that a criminal justice system strives to secure. Among other things, criminal justice is also concerned with fairness, justice and accountability. In particular, we take it that part of the aim of any criminal justice system is to express deserved censure (Bennett, 2019;Duff, 2001;von Hirsch, 2017). In doing so, it recognizes criminal action as wrongdoing. As a result, criminal justice is, in part, also about vindicating victims of crime and to get the wrongdoer to acknowledge that what was done to the victims was wrongful and unjustified (Bennett, 2019;Duff, 2001;von Hirsch, 2017).
In this section, we argue that the communicative theory of punishment, with its emphasis on criminal punishment as a penal dialogue, can give us a plausible moral justification for liberal prison furlough and visitations programs within a framework that at the same time justifies punishment as deserved censure. In particular, we argue that furlough and visitation together serve important expressive and communicative functions that are favorable from the perspective of the communicative theory of punishment. In making this argument we also discuss the implications of this view for penal policy.
Respect for prisoners as members of the community As we noted above, the communicative theory of punishment portrays criminal punishment as a form of secular penance. It is 'a burden imposed on an offender for his crime, through which, it is hoped, he will come to repent his crime, begin to reform himself, and thus reconcile himself with those he has wronged' (Duff, 2001: 106). As such, punishment is not merely imposed on the offender, but seeks to engage with him in a way that respects him as an autonomous moral agent and as a member of the normative community.
As previously detailed, this view disfavors the use of incarceration. As a mode of punishment, imprisonment excludes the offender from the normative community, most evidently, because of how it physically removes him from the community. According to the communicative theory of punishment, this makes imprisonment a harsh form of punishment suitable only for more serious offenses. That said, we should bear in mind that imprisonment, according to this view, should not be a matter of complete exclusion. A prison sentence could, and indeed should, ideally constitute a form of secular penance and should be designed in such a way that repentance can be induced, deepened and strengthened (Duff, 2001). Whether it does so will of course largely depend on the prison conditions. Duff (2001) does not explore in much detail the sort of prison regime that his theory demands. However, he suggests that the prison regime should, to the greatest extent possible, treat its subjects as members of the normative community and provide them with opportunities to repair the normative relationships that were damaged by the criminal wrongdoing (Duff, 2001). While Duff never says so explicitly, we suggest that both prison furloughs and visitations programs are indeed crucial towards this end. What else could be as important for treating and recognizing prisoners' status as members of the normative community than allowing them to have direct and meaningful interactions with other members of the community? Being allowed to have visits from family, as well as other potential visitors, serves as a continuous reminder of one's social bonds to other individuals in the community. In fact, visits from other people besides family members is something that should be encouraged, for example by providing support to organizations that organize and initiate prison visits as part of voluntary community service. As for families, it has been suggested that prison visitations provide 'an opportunity to maintain or repair relationships disrupted by incarceration' as they might serve to 'settle the past, discuss the present, and plan for the future' (Tasca et al., 2016: 460).
These observations provide the basis for our first argument in defense of the importance of both furloughs and visitations. While prison could be administrated in a range of different ways in terms of how well it provides opportunities for reform and reconciliation (by providing access to education and vocational training for example), it is hard to deny that visitation and prison furlough are crucial for maintaining and communicating the message that the prisoner is a member of the normative community. Both prison furlough and liberal visitation programs send the message to the prisoner that his punishment, although deserved, should not to be understood as an irrevocable judgment of complete exclusion and that the prisoner is not only a prisoner, but also a member of the political community to whom, as Duff puts it, we owe respect and concern (Duff, 2003: 305). In sociological terms, the prisoner's "master status" (Becker, 1963) should not be that of an excluded prisoner or the only status that society should afford him.
Some might argue, contra our above argument, that prisoners are in fact not members of the political community. For example, those who endorse a version of social contract theory might argue that by committing a crime one violates the social contract and therefore forfeits one's usual status of an equal fellow citizen (at least temporarily). In this view, the message of exclusion associated with a prison sentence might even be appropriate, as it forcefully communicates the seriousness involved in breaching the social contract. In response, we should note that there are both pragmatic and moral reasons for favoring the communicative view. By treating prisoners as fellow citizens, rehabilitation and reform are probably more likely. And as Duff has argued, our concern for our fellow citizens should arguably not depend on whether they abide by the law (Duff, 2014). Therefore, insofar as we aspire towards a criminal justice system based on the communicative theory of punishment, it seems as if liberal visitation and prison furlough ought to be understood as integrated and evident parts of the prison regime.

Facilitating secular penance
As we saw above, the communicative theory of punishment suggests that imprisonment should not only recognize the prisoner as a member of the normative community. It should also be structured and administrated in a way that it can serve as a secular penance. According to Duff (2003), penance is 'intrinsically inclusionary, not exclusionary: it is required of and undertaken by the sinner as a member of the community -as someone who violated values that are her values as a member of the community, and whose relationships with the community can be repaired in this way' (p. 300). One important aspect of this is that the offender is not a mere recipient of the message that his punishment is held to communicate.
Rather, his punishment is an attempt to engage with him as an active participant in a penal dialogue.
We have already noted that there is empirical evidence suggesting that longterm prison sentences are ill suited for this end (Schinkel, 2014). A few philosophers have also raised skepticism about whether imprisonment can constitute a form of secular penance at all (Brownlee, 2011;Cochrane, 2017). Part of the worry here is not only that prison life is overly coercive, but also that the harsh message of exclusion that imprisonment inevitably communicates renders it unlikely to elicit repentance in the offender, or to promote reconciliation. We believe that this type of criticism is indeed worth taking seriously. However, we do not believe that this means that we ought to abandon the communicative theory, or that we should abolish imprisonment. Instead, we think that it becomes even more important to think critically about the conditions in which undertaking a prison sentence might serve as a secular penance.
To some extent, communicative theorists might argue that prison life already allows for a penal dialog of the sort favored by the communicative theory. For example, as O'Hear (2011) suggests in a discussion on the communicative theory of punishment and indeterminate sentencing, we could understand a prisoner's efforts to comply with the rules of the prison regime as a form of secular penance. In his view, abiding by the rules of the prison regime might constitute the appropriate burdensome work by which the prisoner expresses his recognition of the wrong he has done. Surely, as O'Hear (2011) points out, not all existing prison regimes might function this way. To the contrary, this goal is feasible only under more humane prison conditions where the offender is treated with dignity and respect. Yet, it is not clear that this is sufficient to communicate convincingly that undertaking the prison sentence is part of a penal dialogue in which the prisoner is encouraged to take part not only as a passive recipient, but also as an active participant and an autonomous moral agent. To this end, there should be at least some room to show that one has indeed changed and to respond positively to the penal censure. That is, there should be at least some possibilities for the prisoner to freely display that he or she has indeed taken seriously the message that their punishment is intended to communicate. In an important way, prison furloughs, but also many visitation programs, do exactly that. They provide important room and freedom for the prisoner to act responsibly and to show that he is committed to both reform and to reconciliation. Thus, allowing for prison furloughs, as well as unsupervised visits, send the important message to the prisoner that we -as a community -have not given up on him and that he can regain our trust. By allowing him to engage with the law-abiding community (for a limited and conditional period) the prisoner is provided with an important opportunity to display how he too wishes to repair 'the mutual trust on which the community depends' (Duff, 2001: 101).
It should be observed that our argument is not that every prisoner will de facto perceive their furloughs or their opportunities for visits as chances to show that they are active participants in a penal dialogue and how they have taken the message communicated by their punishment seriously. To the contrary, we suspect that a majority of prisoners perceive neither the opportunities for visits nor furloughs this way. Even so, our point here is that if a prison sentence ought to serve as a form of penance, as the communicative theory suggests it should, then we also need to communicate properly to the prisoner that it is both possible for him to regain our trust and to repair his relationship with the wider community. We also need to clearly communicate that we, as a community, see the offender as an equal moral agent capable of all this. Our point is that visits and prison furloughs are among the type of practices by which we express this most clearly within the prison context. In fact, how else can we effectively seek to persuade the prisoner that his punishment is an appropriate form of reparation in which he is expected to partake?
In sum, prison furlough and visitation programs might, on the communicative theory, be understood as important practices that allow for a meaningful penal dialog. Offering visits and prison furloughs soften the message of exclusion inherent to imprisonment as a mode of punishment. It expresses a community's message regarding the prisoner's ability to change and to repair his relationships with the victim and the community. Even more so, both visitation and furlough reinforces the common idea that prison should not be thought of as a place of mere warehousing or exile, but as a place where inmates actively lay a foundation for reconciliation and, hopefully, a successful reintegration back into the community.

Practical implications
A communicative account of visitation programs and prison furloughs is not only a theoretical exercise. As we argue in this section, it also has significant implications for criminal justice policy. While visitation programs and prison furloughs often are related to concerns about risks and security, as well as rehabilitation and reintegration, simply reducing opportunities for visits and furloughs to a matter of mere risk assessment seems problematic from a communicative approach. On a rehabilitationist and preventive approach, the criteria for furlough and visitation will largely be to focus on the assessment of possible security risks (including prison escapes) and prospects of rehabilitation. Moreover, a rehabilitationist or preventive approach might also take the prisoners behavior while in prison, as well as the nature of his or her crime or criminal history, into consideration when deciding whether the prisoner should be granted visits and/or prison furlough. In contrast, a communicative approach suggests that there are strong moral reasons for adopting a more principled approach as the guiding policy for both furlough and visitation programs. This view also favors more liberal visitation programs, especially in comparison to the practices found in many US states as well as in other Western jurisdictions discussed above. While we obviously cannot offer complete guidelines for policy-making here, the principles we have in mind can be summarized as follows.
First, from the perspective of the communicative theory, visitation programs and prison furloughs should be seen as fundamental aspects of a morally justified prison regime to which prisoners should be entitled, and not as "privileges" or "rewards" for abiding with the rules of the prison regime. According to this view, visitation programs and prison furloughs should not be reduced to mere means for promoting rehabilitation and reintegration to society or good prison behavior. As we argued above, both prison furlough and visitation programs send a recurring message to the prisoner that his punishment is not an irrevocable judgment of exclusion and that the prisoner is not only a prisoner, but also a fellow citizen. In this sense, prison furlough and visitation programs are crucial for establishing a prison regime that treats and recognizes its prisoners as members of the community, as the communicative theory so strongly requires. Hence, both visiting programs and prison furlough should be understood as integrated parts of the prison system in order for the prison regime to be justified in the first place. This further implies that the right to visits and prison furloughs is the default and that this right should be denied only in exceptional cases. For example, in contrast to some of the jurisdictions mentioned above, it is not clear that minor breaches of the prison rules should call for the revocation of visiting opportunities or prison furlough.
However, we should bear in mind that there are prisoners who are dangerous and who might impose a substantial risk to members of the public. In those cases and at least for those offenders, it might indeed be morally permissible to deny prison furlough. The concern for public safety might then simply override both the importance of facilitating secular penance as well as the general requirement to respect ordinal proportionality in sentencing. Despite this, we should also bear in mind that not all prisoners are inherently dangerous (see e.g., Coid et al., 2009). Neither are they necessarily so for the full sentencing period. Therefore, even if both visiting or furlough opportunities must sometimes be denied due to robust concerns about risk, it is important to remember that the default position still is that the revocation of visitation or furlough rights should not be permanent. 'All or nothing' policies should be strongly discouraged in favour of an individualized assessment in which we seek to balance the possible conflicting goals.
Second, from the perspective of the communicative theory, allowing for prison furlough only during the years just before release from prison seems too restrictive. This is the practice in some US jurisdictions, however. The American federal law, for example, enables furloughs, usually, only for inmates within two years of their anticipated release dates under certain conditions (U.S. Department of Justice, 2011, s.7(a)). In contrast, according to the communicative theory that we are examining here, visits and permission to leave prison should arguably be available throughout the whole sentence, not only towards the end. Indeed, empirical research suggests that visitations during the first months of the punishment may help prisoners to cope with newly disconnected societal ties and the sudden pains of imprisonment. This could have a more positive effect on the prisoners' behavior both while in prison and post-release (Cochran and Mears, 2013). That is important not only from rehabilitationist point of view, but also from a communicative theorist's perspective.
Third, following the communicative theory of punishment, rights to visits and to prison furlough should not necessarily be decided with reference to the seriousness of the offense. In some US states, for example, the seriousness of the offense is an important criterion for the possibility of having visits and for being granted prison furlough. As we have noted already, some US jurisdictions will not usually grant a furlough to a prisoner who is convicted for an offence against a person (U. S. Department of Justice, 2011). However, from the perspective of the communicative theory of punishment, it is important to remember that a prison sentence, even if it is inevitably sending a message of social exclusion, must also be compatible with recognizing and treating the prisoners as a member of the normative community. This is so, even if the crime for which the prisoner is sentenced was a serious one. Therefore, insofar as the communicative theory of punishment is plausible, there is reason to allow for visitation programs and prison furloughs to become integrated parts of the prison sentence, even in the case of more serious forms of wrongdoing. This does not mean that prisoners might not still be denied furlough, which could be warranted for reasons of public safety or for maintaining security within the prison. However, even if considerations of risk are important, one cannot deny prisoners the possibility of furlough simply by reference to the seriousness of their offenses, as a sole criterion, at least not on the communicative theory of punishment.
Fourth, if our analysis is correct, then this also has implications for the environment and for the conditions under which visits take place. For example, overly closely supervised visits might be experienced as very stressful for those being supervised and might hinder, rather than facilitate, the type of meaningful social contact that motivates visits in the first place. This might be particularly important in the case of family visits, where some of the visitors are children (see van Zyl Smit and Snacken, 2009). Scharff Smith (2015) notes that the intense security control that arises in situations where a prisoner only has permission for supervised visits, as often is the case for remand prisoners in Scandinavia for example, might be experienced as stressful and awkward for the children of those prisoners (see also Aiello and McCorkel, 2018). Comfort (2007) points out that visiting rooms in US prisons might be ill-equipped and rarely accommodating to children's needs. Restrictive conditions can also make it difficult for some prisoners to interact with their family. It might also make them feel guilty and stressed by having their loved ones exposed to an overly restrictive prison setting (Turanovic and Tasca, 2019). This is unfortunate because of how it might affect the experience as well as the likelihood of visits. Thus, depending on the conditions under which visits take place, there is a risk that the rules surrounding the visitation regime may undermine the possibility for meaningful visitations. In turn, it might negatively affect the chances for the prison to serve as an appropriate communicative sanction. Therefore, from the perspective of the communicative theory of punishment, prolonged and unsupervised family visits should indeed be made available to prisoners and their families. In general, it encourages us to think critically about the conditions under which it is possible for prisoners to maintain and retain contact with important people on the outside and for their social bonds with, for example, family members to remain intact.
We suspect that some might find these suggestions overly demanding or purely utopian, especially in a context where the current prison system falls short of satisfying either some or any of the suggestions that we have made. However, we should immediately say that we do not believe that this objection shows that there is anything wrong with either the communicative theory of punishment or the practical implications that we have argued follow from it. To the contrary, the above analysis depicts a normative ideal for which any prison system could strive towards. Thus, what we have done here is to point to a direction for what needs to be changed, which is not necessarily the same as ensuring change. What our argument shows is that for those prison systems that currently fall short of this normative ideal, there is a strong moral imperative to take important measures in order to move closer to this ideal. As Duff (2001) notes, 'the task of justifying criminal punishment . . . is the task of so transforming the content and context of criminal punishment that it can become what it ought to be' (p. 201).

Conclusion
Retributivist theorists have devoted much of their focus on the implementation and implications of retributivist theories for criminal sentencing and the sentencing phase. In this paper, we have expanded the scope of retributive theory by focusing on two important aspects of prison life, prison furloughs and visitation programs. We have analyzed these two practices from the perspective of the communicative theory of punishment. In this view, prison furloughs and visitation programs serve two important communicative functions. Firstly, allowing for both prison furloughs and visitations to a larger extent serves as a continuous reminder that although the offender has committed a wrong serious enough to warrant temporary exclusion from the community, he or she is still a member of the same community and deserves to be treated and recognized as such. This is largely because prison furloughs and visitations allow the prisoner to continue to have direct and meaningful interactions with other members of the community. Secondly, both prison furlough and visitation programs are crucial for the prison sentence to qualify as a form of secular penance, most notably because of how they allow important room for the prisoner to be able to act as a responsible and trustworthy agent. In doing so, it makes it possible -at least in theory -that the prisoner can show freely that he or she has taken the message expressed by their punishment seriously and that he or she has indeed changed. Thus, both visitation programs and prison furloughs should be understood as crucial aspects of any prison regime if one accepts the idea that the prison sentence should constitute a form of apologetic reparation, as communicative theorists hold that it should.
In our normative analysis, we have also discerned a number of important implications for penal policy. First, from the perspective of the communicative theory, visitation programs and prison furlough should be viewed as fundamental aspects of a justified prison regime and not as "privileges" or "rewards" for abiding with the regime rules. This further implies that the right to visits or furloughs should only be denied in exceptional cases and that it is not clear that minor breaches of the prison rules should result in visiting opportunities or prison furlough being denied. Second, both visitation programs and prison furloughs should be available throughout the whole sentence, and not only towards the end of the sentence. Third, following the communicative theory of punishment, it is not obvious why opportunities to have visits and prison furloughs should be denied or limited merely because the prisoner in question is sentenced for a graver or more serious crime without individual assessment. Last, a commitment to the communicative theory of punishment entails that there should be a strong commitment to creating and fostering a prison environment that facilitates meaningful visits. Thus, one ought to make sure that visits, for example from families, can take place in a family and child-friendly environment and, to as great an extent as possible, without being closely supervised.