A philosophical account of repentance

Abstract Repentance is central to the doctrinal philosophies of all major religions. From a theoretical point of view, however, the practice poses significant challenges. As the past cannot be altered, the guilt that obtains from already committed sins appears forever fettered to the sinner. In this article, I explore this conundrum and discuss a number of solutions that have been proposed in religious traditions. I show how these solutions fail to satisfy from both theological and philosophical perspectives. Finally, I propose a novel approach that, I believe, solves the problem.

past. In the following discussion, I will elaborate on this issue, focusing on the tension between the aforementioned claims.
In this article, I will suggest an original solution to this tension, one that allows us both to accept the core of these three claims and to avoid inconsistency. I will first discuss some solutions proposed in the religious and philosophical literature, and identify their various weaknesses. Next, I will present a unique solution and show how it manages to avoid inconsistencies. Finally, I will attempt to show how my suggested solution captures our pre-theoretical intuitions about the nature of repentance and its applications.
A few preliminary remarks are in order. First, many of the philosophers and theologians dealt with in this article are Jewish, and their writings refer mainly to Jewish theologyand practice. This reflects the fact that, among religioustraditions, I am most familiar with Judaism. That being said, I assume parallel examples from other traditions. Second, although my discussion references religious texts and focuses on the religious context of repentance, I relate to repentance in the moral, as well as religious, context. There may indeed be important differences between the moral and the religious spheres, but I leave such considerations aside.

Solution 1
The first and primary solution is to reject the first claim above, according to which 'the past cannot be altered'. In Jewish religious literature, we can find two methods that adopt this approach.
The first method is simply to point out cases in which it seems that the past misdeed has been repented for. For example, suppose person X stole money from person Y and subsequently not only regrets his action but returns the stolen moneyand perhaps even compensates Y for the inconvenience and suffering he caused him. In this case, it appears that X has succeeded in repenting for his misdeed. 3 This solution strikes me as unsatisfactory for three reasons: 1. Even if such counter-examples show us that we should be careful about accepting the first premise, they do not accomplish what we set out to do. After all, had there truly been no misdeed? The past, after all, remains the same. This difficulty leads us to the next point. 2. Does the deed of repentance (returning the money, etc.) cancel out the previous misdeed, or does it merely make up for it by balancing the guilt with some merit? It seems more intuitive to argue that X is simply performing a good deed that 'balances out' his prior transgression so that he ultimately emerges as righteous, or at the least, returns to his pre-transgression state. 3. Even if we accept that there are certain cases in which one can alter the past, and therefore we could cleanse ourselves of guilt, it seems unlikely that this can be applied to every kind of religious/moral misdeed. Can we make amends for a transgression in which no one has suffered (assuming such a case exists)? Or one in which someone has been killed? 4 Therefore, even if this approach, in rejecting the first premise, succeeds in highlighting certain situations in which repentance compensates for past misdeeds, it still fails to explain how. Nor does it explain how repentance is possible in other situations. 5 The second approach, which rejects the first premise found in Jewish religious sources, posits that repentance has the power to change the past. In this view, repentance performed in t 1 can alter or change the reality in t 0 .
Several statements in Jewish tradition can be thus understood. For example, a well-known Talmudic principle states that a repentance has the power to make a person's 'intentional sins count for him as merits' (B. Yoma 86b). In other words, his sins somehow become good deeds. A similar sentiment can be found in Rashi's (Rabbi Shlomo Yitzchaki, 1040-1105) commentary on the Talmud, when he declares that 'whoever repents out of love, his transgressions are uprooted from the source' (Yoma 86a). However, as mentioned previously, these statements do not have to be interpreted in a way that allows them to support the veracity of the claim that one can change the past. Instead, they can easily be read as some kind of metaphor that aims to express the ability of a man to change his moral status via repentance.
Nevertheless, Kabbalistic and Hasidic philosophies chose to adopt the radical interpretation of the above remarks. For example, the Lubavitcher Rebbe (Menachem Mendel Schneerson, 1902Schneerson, -1994 wrote explicitly that 'repentance out of love also changes the past' (Schneerson (1991), 56). Rabbi Adin Steinsaltz (1937Steinsaltz ( -2020, a prominent exponent of Hasidic and Kabbalistic thought, summarized this view of repentance as follows: The broader significance of repentance includes the idea that a person controls his movement in all dimensions, including the dimension of time. Without repentance, the entire system of time operates in a single direction: a deed performed at a certain point in time cannot be undone, a deed performed is an immutable fact; the past is therefore the fixed element of reality. Nevertheless, the power of repentance enables the time system to also operate in the opposite direction, i.e. the past, and not only the future, can be changed. Repentance is essentially not only a change from a specific turning point onward but also a change of the pastits erasure or transformation into another value. Repentance then is, in essence, complete control of time, including of the seemingly unalterable past. (Steinsaltz (2011a), 69-70) 6 An instructive example of this view can be found in a Midrash (textual interpretation written by ancient Judaic authorities) that deals with the sin of the biblical King David. 7 The second book of Samuel describes how David seduced a married woman, and then sent her husband to his certain death in battle. The narrative subsequently tells how David repented of his sin, a process that is sometimes even presented as the paradigmatic model of repentance -'David, son of Yishai, who raised the yoke [ulla] of repentance' (B. Moed Katan 16b). And yet, the Talmud (Shabbat 56a) also features a declaration according to which 'anyone who says that David sinned [with Batsheva] is nothing other than mistaken'. How is this possible? The Bible explicitly states that David sinned.
Furthermore, the very presentation of David as the ideal of repentance is in itself an indication that he sinned. According to the Midrash, it is precisely because he repented that it cannot be said that David sinned. According to both the Midrash and the Hasidic-Kabbalistic view presented above, repentance alters the past and 'erases' the misdeed. If so, the claim that David sinned is untenable because upon repenting his sin, he erased it, such that it never occurred in the first place. According to this view, if a person repents, 'rewinding the film' will reveal no trace of the sin.
Although this solution preserves the positive attributes of repentancethe possibility of altering the past and cleansing oneself of guiltit remains philosophically problematic. This is because it violates a basic principle held by philosophers (and scientists), according to which the past is fixed and unchangeable. 8 The abandonment of this principle would take more persuasion than rabbinic literature or Kabbalistic references, and I doubt this is even possible. 9 Although the possibility of 'backward causation' and 'changing the past' has been discussed in the philosophical literature (Faye, 2018), few have advocated the idea that the past can be changed in a way that would make the possibility of repentance real. This is not to imply that this view is false, but it does mean that its acceptance would require a more convincing argument than those I have found in the literature.
In this regard, it is worth mentioning that Sam Lebens and Tyron Goldsmith have presented a metaphysical framework 10 in which it is possible for God to change the past whenever a person truly repents (Lebens and Goldschmidt, 2017). Without getting into details, I would just say that I find their suggestion, and more specifically their metaphysical world-view, implausible. 11 Instead, in what follows, I will suggest a more plausible theory, which has the main benefits of Lebens and Goldsmith's theory and avoids most of its downsides.

Solution 2
Another solution, which deals with the difficulties raised by the previous one, was proposed by the medieval Jewish philosopher Moses Maimonides (1138-1204). 12 Maimonides was one of the firstand remains one of the few -Jewish thinkers to formulate a comprehensive model of repentance. In discussing repentance, Maimonides wrote: Among the paths of repentance is for the penitent to constantly call out before God, crying and entreating; to perform charity according to his potential; to separate himself far from the object of his sin; to change his name, as if to say 'I am a different person and not the same one who sinned;' to change his behavior in its entirety to the good and the path of righteousness; and to travel in exile from his home. Exile atones for sin because it causes a person to be submissive, humble, and meek of spirit. (Hilkhot Teshuva, chapter 2, second verse) 13 According to the above, 14 repentance does not entail a process of eliminating past guilt, or of amending the misdeed itself, but specifically involves changing (or more precisely, determining) the present and the future. According to this view, the person who has repented is not the person who 'amended' the past, but the person who 'amends' the future. Maimonides further clarifies this idea: Accordingly, throughout the entire year, a person should always look at himself as equally balanced between merit and sin and the world as equally balanced between merit and sin. If he performs one sin, he tips his balance and that of the entire world to the side of guilt and brings destruction upon himself. [On the other hand,] if he performs one mitzvah, he tips his balance and that of the entire world to the side of merit and brings deliverance and salvation to himself and others. (ibid., chapter 3, fourth verse) Maimonides asks us to view the world as balanced between innocence and guilt; our role is to try to tip the scales in favour of the former. He does not even entertain the notion of 'erasing the sin' in order to influence the balance positively, but only raises the possibility of adding good deeds. It is reasonable to assume that this is due to Maimonides' general conviction that a person cannot change his past. All he can do in order to change the balance of the scales is to change his sinful ways and begin performing good deeds. According to this view, repentance is the performance of enough good deeds to tip the balance back to where it was prior to the sin. I will call this approach the 'even up' method.
This solution is much more intuitive but, nonetheless, remains incomplete. It is intuitive because it evades the problematic nature of repentance. It views repentance not as requiring the alteration of the past but a change of the present and future. In other words, this view rejects the third premise by redefining the term 'repentance' as an act that does not require changing the past.
Nevertheless, this approach has several problems. First, it presents the process of repentance as entirely unrelated to past sins. This disengagement means that the process of repentance does not need to include atonement for the misdeed itself, but only the performance of a good deed in place of it. Thus, for example, a thief is not obligated to return what he stole in order to repent (although he may be obligated to do so for other reasons). His repentance only requires the donation of a sufficient sum to charity and the commitment not to steal again. Such an approach seems problematic, to say the least. 15 The second problem is that this view of repentance seems incomplete and unsatisfactory because a person performing a transgression can never eliminate the associated guilt. All that can be done is to merit more good deeds, though these will fail to erase the misdeed or the guilt it carries. Maimonides himself accepts, as it stands, the argument presented at the beginning of this article. For him, we should not wonder at the fact that a person performing a misdeed will bear the resultant guilt forever. Just as the past is irrevocable, so too is the bearing of guilt.
Is this how most of us approach repentance? Do we believe that a person is destined to bear the weight of guilt for his misdeeds forever, even if he regrets them and does everything possible to atone for them? I do not believe so. Perhaps this is a case of wishful thinking, and there really is no way to help us eliminate our guilt. If so, we must accept the view of repentance as proposed by Maimonides. I believe otherwise, namely, that we can accept the first and the third claims and reject the second. In other words, although we cannot change the past, we are capable of eliminating guilt.

Solution 3
I shall now present my own solution for the problem at hand. 16 On the one hand, it allows us to preserve the intuition that repentance is possible, in the sense that a person can eliminate his guilt for a past sin. On the other hand, it does not require the ability to change the past. My solution relies on a specific understanding of personal identity. In general, we can say that there are two types of views about personal identity. 17 The first considers personal identity as a binary feature: either I am identical to the person who I was at some other point, t, in the past, or I am not. According to this view, there is a 'core' self that, if lost, makes us a different person from the one we were previously.
The second view suggests that personal identity is a non-binary characteristic. According to this approach, the answer to the question 'Am I the same person I once was?' can be gradable (e.g. 'sort of' or 'almost', etc.) and not only binary answers ('yes' or 'no'). One main reason to maintain the second approach, from a naturalistic point of view, is that it emanates from the view that a person is a collection of biological and psychological traits. Therefore, if these traits change over time, the person will also change over time. However, even on an unnaturalistic (and even theistic) world-view, this view can be justified.
To clarify this point, I will consider the question of whether I am identical to the person I was when I attended school. The first approach tends to suggest an affirmative answer -I am, after all, still me. On the other hand, according to the second approach, the answer is much more complex. Since leaving school, I have changed (e.g. my body's constitution has changed, as have my psychological characteristics, etc.). The answer to the question whether I am the same person will, therefore, be something like 'not exactly, but there are similarities'. The first approach seems more intuitive, because we feel as though we continue to be the same person we were when we were younger and not someone who was just very similar to us. If we think about it for a moment, however, there is something very reasonable in the thought that we are not identical to the person we were some years ago.
I suggest that this more complex approach towards the conditions of permanence relating to one's 'self' offers a solution to the repentance puzzle. 18 Suppose at any given time (t), a person is not the same person as at (t−x). In that case, he can bear no moral/religious guilt for a transgression committed by his prior selfas he was, after all, not the same person (naturally, the details of this, especially concerning the change a person needs to undergo in order to be exempt from guilt, require clarification). 19 This solution does not claim that we can alter our past, but provides us with a way to shed our past guilt. In other words, this view accepts the argument presented at the outset of this discussion, but applies it to a specific personality, and recognizes that an individual has the capacity to separate from the past by creating a new personality. According to the complex approach to personal identity, as we have seen, a person changes constantly. Repentance, according to this perception, is an active endeavour towards selfchange aimed at eliminating that part of 'me' bearing the guilt. 20 Who has not exclaimed, 'I am not the person I used to be!' Such an understanding of self-identity allows us to accord these statements an altogether serious meaning, without the need for grandiose theological-metaphysical assumptions that the modern person finds difficult to accept.
Yet, this line of thought raises a fundamental problem. If the person who repents is not the same person he was before, why will he be motivated to repent? This process, if it succeeds, will wipe him out, and he will cease to exist.
Before trying to cope with this difficulty, it is important to note that according to the complex approach, it is possible to separate the different parts of one's self. This would mean that, at least in principle, a person can choose to discard a part of his 'self' that bears the guilt for a past misdeed while preserving the 'self' that acts in a virtuous manner. This is a simplistic presentation, and further study is required to understand the exact workings of repentance according to this approach and its implications for the view of 'self'. In part, this requires developing a better understanding of how guilt or merit is apportioned and, subsequently, how a change in 'self' can change the applicability of merit.
Thus, embracing the view that different parts of oneself can be considered 'independent' offers a reply to the claim that people will have no interest in their future selves after repentance, because that future self will not really be 'me' at all. This direction of thought indicates that it is precisely the process of repentance that allows us to improve ourselves and create a new 'me', devoid of the flaws and moral failures inherent in the former 'me'. Of course, this will no longer be the exact same 'me'. However, if we accept the non-binary approach to the self, that the 'me' continually changes, it seems reasonable to assume the existence of a motivation to change for the better. Furthermore, as Parfit argued, the relation a person deems important to maintain with future versions of 'me' is not with a continuation of myself but rather a continuation of my characteristics, memories, passions, etc. (Parfit, 1987) (he terms this the 'R Relation'). If we adopt Parfit's stance, it appears reasonable to assume that it is specifically the process of repentance that preserves the important parts of the present 'me'. 21 Now, in the light of this approach, we can better understand the words of Maimonides, cited above: Among the paths of repentance is for the penitent to constantly call out before God, crying and entreating; to perform charity according to his potential; to separate himself far from the object of his sin; to change his name, as if to say 'I am a different person and not the same one who sinned;' [emphasis added] to change his behavior in its entirety to the good and the path of righteousness; and to travel in exile from his home. Exile atones for sin because it causes a person to be submissive, humble, and meek of spirit. (Hilkhot Teshuva, chapter 2, fifth verse) 22 Maimonides is, in fact, describing a process of changing identity. Repentance, in this reading, entails an identity change of the most profound kind. This is further developed in Maimonides' own words about the manner of ascertaining whether or not a person has repented: [Who has reached] complete Teshuvah [repentance]? A person who confronts the same situation in which he sinned when he has the potential to commit [the sin again], and, nevertheless, abstains and does not commit it because of his Teshuvah alone and not because of fear or a lack of strength. For example, a person engaged in illicit sexual relations with a woman. Afterwards, they met in privacy, in the same country, while his love for her and physical power still persisted, and nevertheless, he abstained and did not transgress. (ibid., chapter 2, first verse) A person knows that he has repented when he finds himself in the same situation in which he sinned previously, but now successfully overcomes his urge and does not 'stumble'. In other words, in the past, the person possessed a disposition whereby in situation A, he was unable to control himself. Following a process of repentance, he lost this disposition and gained another, whereby, in the same situation, he is able to exercise selfcontrol. According to Maimonides, such a change allows one to cleanse oneself of the guilt one previously bore. This is not because it cancels out the past, or because it 'exempts' the person from the misdeed, but because it testifies to a profound change of self, so much so that one is not considered the same person. This new person does not bear the moral guilt of the person responsible for the past transgression.

Discussion of the solution
Having presented this solution, which I regard as the most successful of the available choices, I shall briefly expand on the characteristics of this view of repentance. In general, people hold certain 'expectations' or pre-theoretical intuitions about repentance. In this section, I will present these intuitions and examine my solution in their light, comparing them to the 'even up' approach attributed to the first interpretation of Maimonides. This is because the 'even up' approach is, in my view, the most convincing opponent.
Our first and primary expectation from any account of repentance is that it will include a correlation between the severity of guilt, and the difficulty and length of the appropriate repentance process. One who spoke impatiently to a neighbour would need to invest much less effort in his repentance than one who committed a severe felony.
As we have seen, this is a central feature within the approach attributed to Maimonides in the first interpretation above, according to which repentance seeks to even the balance between bad and good deeds by doing the latter. Following this idea, if the wrongdoings are insignificant, all one needs to do in order to repent is an insignificant good deed. However, if the bad deed is a severe one, one needs to do a truly outstanding great deed or a host of smaller good deeds in order to achieve repentance.
I believe that this 'expectation' regarding repentance also gets fulfilled in my approach. If repentance is an active endeavour to separate 'me' from the 'me' that sinned, it seems reasonable that a more severe transgression will entail a more demanding process of repentance. There are two main reasons for this. First, the very fact that I found myself in a situation of committing a serious wrongdoing is evidence of my inferior moral/religious status; atonement for it, therefore, requires a more in-depth and fundamental change. Second, it is reasonable to assume that committing a sin influences the perpetrator in some way. One such view can be found in the Talmud, which suggests that the very performance of a misdeed exerts a negative influence on a person's moral perception: 'When a person transgresses and repeats his transgression . . . [he] no longer senses that it is a sin' (B. Kidd. 40a). Therefore, it seems that according to the view of repentance I have presented here, there is indeed a clear correlation between the severity of the transgression and the rigour of the repentance procedure required to cleanse oneself of guilt. This is because if a person truly desires a fundamental change, whereby he separates the identity of his present (or future) 'me' from the 'me' that sinned, he must first reach a situation in which he would not even have committed the sin in the first place.
The second 'expectation' from repentance is that we expect, or hope, that repentance is available for anyone, regardless of the severity of his situation. 23 It seems that this is not exactly the case, according to the 'even up' approach. This is because in some cases, when someone has committed an egregious moral/religious sin, it is impossible, within his lifetime, to do enough good deeds in order to even out and overcome the guilt-merit balance. My proposal leads to the same conclusion. If repentance is an active endeavour to separate 'me' from the 'me' that sinned, and if this process requires primarily psychological changes (although perhaps also biological), it would seem that, at least in principle, no person is prevented from adopting these changes. Nonetheless, it is certainly possible that our psychological or biological nature does not allow for such drastic changes. Therefore, even if the gates of repentance are, in principle, open to all, some people will be unable to undertake this process in practice. It seems that this conclusion is, to a large extent, also consistent with our intuitive understanding of this point.
The third 'expectation' from repentance is that we expect repentance not to be granted in cases of cynical misuse, namely, that a person will not be able to repent solely to cleanse himself of guilt, but then continue sinning. A famous Mishna in Tractate Yoma (8,9) states that 'one who says: "I shall sin and repent, sin and repent," they do not allow him to repent'. In other words, the repentance of a person who exploits the repentance mechanism in order to continue or even increase his sinful behaviour, without any acceptance of guilt, is deemed unacceptable. The Mishna here is drawing on a religious-theological view whereby God does not accept the repentance of a cynic. The repentance proposal presented above also denies the possibility of cynical misuse, as a person declaring that he will 'sin and repent' is not attempting to separate and change his 'self' but, in fact, to preserve himself and his actions. By contrast, the 'even-up' approach does not succeed in considering this expectation. To repent for morally/religiously sinning, one need only equalize one's merit-sin balance. If so, repentance is ripe for perversion because nothing prevents a person from deliberately engaging in a cycle of sin and repentance.
The fourth 'expectation' from repentance is that we expect that there will be a meaningful connection between a sin and its corresponding repentance process. If someone steals a car, our intuition tells us that his repentance process did not relate, in some way or another, to that theft.
As we saw, the 'even-up' approach fails to fulfil this expectation. If someone steals a car, he does not need to return the car, or even stop stealing. All he needs to do is act in a way that will even things up. According to my approach, however, it seems that if someone wants to repent for a sin, such as stealing, he needs to act positively in a way that the 'part of him' that led him to this specific sin will vanish such that he will not continue to contain the part of himself responsible for that sin. I am not sure what such a change would include, but it is reasonable to think it must include a change of the disposition and moral beliefs that led him to steal. Hence, according to this view, there is a deep connection between the sin and the repentance process.
The fifth and last 'expectation' from repentance that I will explore in this article is that we expect 'collective entities' such as states, organizations, or nations to be capable of repenting. Although both the 'even-up' approach and my approach can cope with the repentance of 'collective entities', I believe that my approach provides a better and more profound explanation. This is because the complex picture of 'me' transforms a person into a kind of collective entity. Parfit put it thus: I claim that a person is not like a Cartesian-Ego, a being whose existence must be all-or-nothing. A person is like a nation. In the true account of identity over time, these two kinds of entities go together . . . If this is true . . . the account of their identity over time would, in its essential features, be similar. (Parfit (1987), 275) 24 According to this approach, a person's identity is determined by a variety of facts. Similarly, 'collective entities' consist, among other things, of multiple and constantly changing elements. This parallel can explain the repentance process of a 'collective entity' as the same process for individual repentance. For example, suppose Germany wants to atone for its historical misdeeds. In that case, it is not enough that most of its citizens change. Rather, the elements, in virtue of which Germany bears its guilt (whether they be specific individuals, political platforms, or moral conventions and ideologies, etc.), must be separated from the country's former identity. Such a description is reasonable, as a country successfully separating itself from the cause of its guilt would indeed be truly cleansed of all guilt.

Summary
I have presented the difficulty in the idea, regarding repentance, that one can erase guilt without erasing the misdeed itself. I have surveyed four solutions, at least two of which fail to stand up to critical scrutiny. The last, and in my view most satisfying, solution claims that the process of repentance does not erase the misdeed itself but rather only severs the connection between the 'me' that performed it, and the present 'me'. As noted, I believe that this solution conforms to pre-theoretical intuitions regarding the repentance process. I also think that this solution fits the phenomenology of repentance. Consider the penitent's cri de coeur: 'I am not the person I once was; I have changed!' Of course, these can be taken as non-literal assertions. Yet, in my view, they express a basic phenomenological insight. The internal feeling of repentanceat least mineinvolves a profound change of personal identity. When I feel as if I have repented from a certain behaviour, I feel as though I have changed, and I am no longer the same person. Moreover, when I look back on the past, I feel alienated from the person who performed the sinful action.
I do not claim that I have offered anything profoundly new with regard to the essence of repentance. On the contrary, my sole intent was to provide a philosophical account of the feasibility of repentance and its modus operandi. The prevalence of this perception of repentance, if it is indeed prevalent, even without the philosophical account, only serves to strengthen my argument and proves that the explanation I have provided is largely intuitive. Thus, as indicated in the title of this article, I have sought to provide an initial philosophical analysis of repentanceits possibility, its process, and its meaning.
This account may have implications for other moral and religious questions, such as the proper times or conditions for forgiveness, a statute of limitations for sins, the connection between regret and reduction of legal punishment, and more. These questions have been widely discussed in the philosophical literature. Yet, the view of repentance herein presented may enrich these discussions and perhaps even open up intriguing avenues of inquiry.