Undoing one’s past

In contemporary research on memory, the idea of mental time travel (MTT) has been con-nected, at the functional level, with planning and imagining what might occur in one’s future. Episodic memory impacts on our capacity to move imaginatively towards possible scenarios ahead. Consequently, Gerrans and Kennett (2010, 2016) urge us to agree that MTT is essential to moral agency. In this paper, we suggest that if we conceive the specific varieties of MTT as something more than remembering one’s past and imagining one’s future, then the capacity of undoing one’s past both by episodic counterfactual thinking and the emotion of regret must be considered essential to moral agency on equal terms.


Introduction
In "Neurosentimentalism and Moral Agency" Philip Ger ans and Jeanette Kennett endorse the crucial role that imagination and memory play for capa le moral agents. They op ose the soca led "neurosentimentalism" , namely, a kind of metaethical sentimentalism built upon empirical evidence, availa le in studies with neurological patients impaired by significant damages to the ventromedial prefrontal cortex (vmPFC), patients who are compromised both at the level of the tacit affective processes and of moral agency. These are the affective processes which, therefore, Filosofi a Unisinos -Unisinos Journal of Philosophy -19(1):97-102, jan/apr 2018 according to advocates of neurosentimentalism, such as Joshua Greene, Jonathan Haidt and Jesse Prinz, the competence in moral judging crucia ly depends on.
As most of you surely know, in the footsteps of the groun breaking results by Antonio Damasio and Antoine Bechara, there is a lot of discussion, both in neurosciences and moral psychology, on the cor ect interpretation of the experimental data exploring the capacities for moral judging and decision-making in vmPFC patients.
For our cur ent goals, it would be less demanding to discuss the big picture of the theoretical dispute against neurosentimentalism than to point out the pro lematic dissociation between moral agency and moral judgement in the neurosentimentalistic camp, a dissociation su ge ed by the idea that it is only possi le to make moral judgements either by the ap lication of a rule or by automatica ly responding to tacit affective processes, without the need, in both cases, to menta ly project ourselves into fitting future scenarios.
The moral agency could be exemplified by the capacities to plan and make decisions, to commit ourselves to the attitudes of evaluating ends and means, to act in accordance with what we are valuing, to respond to moral demands and to justify chosen courses of a ion after deliberation.
Kennett and Ger ans offers a fruitful hypothesis to explain why we must avoid dissociating moral judgement and agency, or, in their words: "A moral agent needs to be a le to conceive herself as a temporal extended entity as a necessary condition for moral reflection and decision-making" (Ger ans and Kennett, 2010, p. 588).
With the hypothesis in plain view, we can move on to introduce the leading motivation behind it: Mental time travel (MTT) is essential to moral agency because episodic memory and imagination are tools to conceive ourselves as tempora ly extended entities. Furthermore, to reinforce the point about agential competence we must a d another component of Ger ans and Kennett's strategy. The MTT capacities to imagine future scenarios and recover past ones are under voluntary control: "It is this voluntary, executive, a ect of mental time travel which is so important for the capacity for planning, and which makes executive a ion dependent on maturation of the frontal systems" (Ger ans and Kennett, 2010, p. 599).
Neurotypical subjects voluntarily build upon reco lected and constructed past scenarios to decide what to do ahead via simulations of possibilities into the future.
The connections between the capacity for moral agency and MTT could be highlighted in the commitments we assume after short or long-term planning: We exercise the capacity for mental time travel whenever we revise for this year a class we gave last year-remembering what   worked and what didn't-whenever we reflect on what kind of career or job would  best suit us, whenever we plan a holiday or a  shopping trip, arrange a meeting, organize  a party, or commit ourselves to a course of  study, an exercise program, or a marriage (Gerrans and Kennett, 2010, p. 601).
Around six years after the first paper on the subject, the authors re arted the discussion in Mental Time Tra el, Dyna ic Evaluation, and Moral Agency in a slightly different focus, rea ing to a critique by Zarpentine (2017) on the cor ect interpretation of experimental data about vmPFC patients' capacities for diachronic agency, or, in Zarpentine's conception, dynamic evaluation. One of the positive results from the discussion was to clarify the fo lowing point of Ger ans and Kennett's strategy: " [...] what matters for diachronic agency/ dynamic evaluation is the ability to feel the future. Indeed our main point was to ar ue that if people lose this ability they have less of a self to project and hence are compromised as agents" (Ger ans and Kennett, 2016, p. 5).
Taking stock of Ger ans and Kennett's position on the topic give us at least two questions for the next sections: (1) For what reasons research on episodic memory, along the lines of the MTT quasi-paradigm 2 , did the unexpected move from focusing on remembering the past to mainly discussing the anticipation of future scenarios? (2) If the "ability to feel the future" is a functional presup osition to moral agency, why not ar ue for another phenomenological component of we l-functioning moral agency as being the ability to feel the undoing of one's past?
Mental time travel and one's future scenarios When you write a paper for a dossier, it is possi le to presup ose some of the information already given by the previous participants. I am not a le to describe the reception of the mental time view of memory more accurately than André Sant' Anna, e ecia ly in section 2 of his contribution in the present issue -Mental time tra el and the philosophy of me ory -but I would need to reco lect two of the points rightly highlighted by him.
The first one is about the phenomenological a ects of "autonoesis" as described by Tulving: "[t]he act of remembering [...] is chara erized by a distinctive, unique awareness of reexperiencing here and now something that hap ened before, at another time and in another place" (Tulving, 1993, p. 68). The "what-it-is-likeness" of reexperiencing or anticipating autobiographical events gives us the capacity to travel in subjective time, and the conceptual derivations from Tulv-

99
Filosofi a Unisinos -Unisinos Journal of Philosophy -19(1):97-102, jan/apr 2018 ing's autonoesis a low phrases like "feel the future" or "feel the past" when we ta k about the MTT varieties.
The second point directs us to the first question made at the end of the introduction and refers to MTT operations as tools to imagine one's future scenarios rather than just recollect or re-experience the past.
One of the most innovative results of the widespread acceptance of the MTT framework in memory research has been a re-conceptualization of episodic memory, through its functional role in future-oriented planning and decision-making. Rather than just reco lect the past, episodic memory helps to project one's future, working side by side, as we ar ue, with the capacity for undoing one's past.
Daniel Scha er above a l, championed the idea of a proective brain, whose main functional task would be to anticipate future scenarios. The impact of this view is highlighted in the editor's introduction to a new co lection of essays on the topic edited by Michaelian et al. (2016, p I argue that seeing memory as a cognitive system for remembering the past may not be the best way of making sense of its function. Instead, I offer a picture of memory as an integral part of a larger system that supports not only thinking of what was the case and what potentially could be the case, but also what could have been the case. More precisely, building upon the work of Schacter and colleagues (e.g., Schacter 2001;Schacter and Addis 2007), I claim that remembering is a particular operation of a cognitive system that permits the flexible recombination of different components of encoded traces into representations of possible past events that might or might not have occurred, presumably in the service of constructing mental simulations of possible future events.
In a dition to De Brigard's and Scha er's views, the very same general position has been proposed by Klein for the last sixteen years or so (e.g. Klein et al., 2002;Klein et al., 2010Klein et al., , 2011Klein, 2013Klein, , 2016, exploring the biological pressures on evolving episodic memory capacities: "[...] memory, as designed by natural selection, is not simply be capa le of imagining the future; rather imagining the future is its evolved function, its raison d' être" (Klein, 2013, p. 233).
The main reason for the interdependence of (1) episodic future thinking, (2) episodic memory and (3) episodic counterfactual thinking is what we may ca l the "same brain area assumption" , namely, the widespread acceptance of the neuroimaging evidence for the overlap ing a ivation of the same brain areas in these three ecific MTT processes. The fo lowing quotes are bona fide examples of the "same brain area assumption": [...] it seems safe to conclude that episodic future thinking and episodic counterfactual thinking both engage regions that are also recruited when people remember specific past experiences from their everyday lives. On a general level, the overlap of this core-network with the default network is consistent with theoretical perspectives that have emphasized the role of this network in supporting various kinds of mental simulations (Schacter et al., 2014, p. 16).
It is a reasonable assumption that counterfactuals share many cognitive and neurological processes with past and future memory. To construct a counterfactual, key elements from past experiences need to be remembered (like episodic past thinking) and, crucially, some elements need to be recombined so that a novel imagined scenario can be constructed (like episodic future thinking). An abundance of neurological studies during the last decade have demonstrated that episodic past and future thinking share several areas in the brain. Addis et al. (2009;Schacter and Addis, 2007) proposed a core memory brain network engaged during remembering and imaging of past and future events that includes the hippocampus, posterior cingulate/retrosplenial midline, inferior parietal lobule, lateral temporal cortices and the medial prefrontal cortex (Van Hoeck et al., 2013, p. 556, emphasis added).
In view of the structural similarities, at the level of the physical realization of the three main MTT processes in the brain, it seems defenda le to assume their functional coordination, hypothesizing the adaptive value of episodic future thinking as the chief target of the entire MTT cognitive system. This hypothesis is at the core of the MTT quasi-paradigm and leaves an open door to explore the connections between moral agency in particular, agency in general, and the MTT processes. A l things considered, it seems equa ly defenda le that, as moral agents "[...] we predominantly stand in the present facing the future rather than looking back at the past" (Su dendorf and Corba lis, 1997, p. 147).
This hypothesis has some theoretical benefits as we l, for example, it helps to come to grips with the peculiar pervasiveness of misremembering in our daily lives, the forthright discussion in De Brigard (2013) as tries to demonstrate.
The same general hypothesis has been expanded in another fruitful direction by Hoerl and McCormack (2016), Filosofi a Unisinos -Unisinos Journal of Philosophy -19(1):97-102, jan/apr 2018 discussing both the role of the counterfactual emotion of regret and the evolutionary function of episodic memory in future-oriented decision-making. In their words: "Our ar ument wi l be that episodic memory, because of its backward-looking element, underpins the ability to experience regret, and that this is an important way in which episodic memory impacts on and sup orts adaptive future-oriented decision making" (2016, p. 242).
To examine this ar umentative strategy in general and in Hoerl and McCormack's piece wi l automatica ly demand from us an answer to the second question made in the introduction, on the ability to feel the undoing of one's past as a necessary component of the we l-functioning moral agency. That wi l be the task for the next section, but by now we need to increase the understanding of the multiple ways a competent moral agent faces the future.
One recently proposed taxonomy for future-oriented mental time travel capacities helps to disclose the moral agency dynamics, at least in one of the directions in which we conceive ourselves as "temporal extended" entities. According to Szpunar et al. (2016), we can classify future-oriented mental time travel, or pro ection, in four basic modes: simulation (construction of a detailed mental representation of the future); prediction (estimation of the likelihood of, and/ or one's reaction to, a particular future outcome); intention (the mental act of setting a goal); and planning (the identification and organization of steps toward achieving a goal state) (Szprunar et al., 2016, p. 21).
The four operative modes equa ly draw on semantic and episodic memory contents. We wi l refrain from discussing the semantic forms of FMTT keeping an eye on the episodic forms only. We choose that way on phenomenological grounds, hence in the episodic FMTT "the subject has a pre-reflective sense that he is "pre-living" a possi le future" (Michaelian et al., 2016, p. 6). In our view, that is another ecial feature of moral agency to be a ded in a comprehensive theory.
For the episodic contents, simulation is the "construction of a mental representation of a ecific autobiographical future event" , prediction is the "estimation of the likelihood of [...] one's rea ion to a ecific autobiographical future event" , intention works "setting a goal in relation to a ecific autobiographical future event" , and for planning we have the demand for "organization of steps needed to ar ive at a ecific autobiographical future event" (Szprunar et al., 2016, p. 22).
For the sake of the ar umentative strategy developed in the next section, we need to presup ose the cor ectness and explicative power of this taxonomy. Understanding the phenomenological a ects of the four episodic modes of pro ection is crucial to what wi l fo low. As we wi l see in the next section, the "what-is-it-likeness" of FMTT and its functional connections with moral agency wi l be partia ly replicated for the counterfactual episodic thinking.
In order to feel the temporal extension of the self into the future, we need to feel the simulation, prediction, intention and planning, working upon episodic contents. In order to feel the undoing of one's past, we need to pay attention to the phenomenological dimensions of the counterfactual e iso ic simulation or the construction of a mental representation of an autobiographical counterfactual event, and counterfactual e iso ic prediction or the estimation of the likelihood of one's rea ion to an autobiographical counterfactual event.

Episodic counterfactual thinking, regret and moral agency
In both fictional literature and science, the question as to what episodic counterfactual thinking is for has been asked now and then, keeping up with skeptical wor ies, sometimes with genuine curiosity, sometimes only nurturing pure perplexity. It is not exactly an easy task to formulate an explanation for the evolutionary pressure to think about what you and me could have done or chosen differently. Everybody knows what must be the meaning conveyed by common sayings like "don't cry over spi led mi k" , "what's done is done" , or, a little more fi uratively, the proper rhetorical push of phrases like "to rake over old coals" . A l these phrases and sayings point to the same old and venerated metaphysical view on the immutability of past events.
Nevertheless, it is plausi le to find a convergence of opinions on a usual su ect for that skepticism or perplexity: the emotion of regret. In view of the immutability of the things already done, what could be a function for the counterfactual emotion of regret?
Let's re-start e a lishing a meaning for the word: "Regret [...] is by definition an emotion directed toward the past: one regrets a choice one has made, typica ly believing that if one had chosen differently a better outcome would have obtained" (Hoerl and McCormack, 2016, p. 241-242). Therefore, regret is a counterfactual emotion, to be more ecific, an emotion functiona ly dependent on the occur ence of an upward counterfactual conditional thought. The undoing of a ects of one's past-the "if I had chosen differently" antecedent part-is fo lowed by the simulation of a possi le upward consequence-the "a better outcome would have obtained" consequent part.
In the psychological literature, it is not difficult to find authors hypothesizing functions for the emotion of regret in decision-making processes, planning, or in motivation and control of future behavior (see Zeelenberg and Pieters, 2007). Neal Roese proba ly has been one of the main advocates of a functional theory of counterfactual thinking in general and counterfactual emotions, like regret, in particular. Back in the early nineties, he wrote: "[...] people may strategica ly use [...] upward and a ditive counterfactuals to improve performances in the future (a preparative function)" (Roese, 1993, p. 806). In his last book on the subject, using a more popular Filosofi a Unisinos -Unisinos Journal of Philosophy -19(1):97-102, jan/apr 2018 tone, we also read: "Regret feels bad, but it is utterly essential for healthy living. Understanding and harnessing our own regrets can make you better. [...] Regret is an example of a negative emotion that spurs people to pro lem-solving and personal betterment" (Roese, 2005, p. 2). Hoerl and McCormack (2016, p. 245), as already mentioned, ar ue in a more straightforward fashion, using the emotion of regret as a link between decision-making about the future and episodic memory. Their ar ument offers a simple explanation of the role played by episodic memory in thinking and deciding about the future, through the fo lowing schematic steps: (1) "the emotion of regret plays a crucial role in future-oriented decision-making"; (2) "one can only regret what is in the past"; (3) regret "requires a type of mental simulation that is itself intrinsica ly bound up with a capacity to recollect one's past"; (4) therefore, the capacity to reco lect one's past (episodic memory) "plays a crucial role in future-oriented decision-making" (Hoerl and McCormack, 2016, p. 245).
Trying to sup ort the premise (1), Hoerl and McCormack (2016, p. 246) say, endorsing Zeelenberg and Pieters (2007): One way to argue for a close relation between regret and future-oriented decision making might be to point to its close relation with agency and personal responsibility: indeed, Zeelenberg and Pieters (2007) argue that regret is the only negative emotion that has a special relation with one's own choices, such that it is only experienced if one believes one has made a poor or suboptimal choice. However, for present purposes we want to focus on the temporal character of regret. Zeelenberg and Pieters (2007, p. 8) capture this character in their claim that "[r]egret bridges the past and the future in the present." What they mean by this is that although regret is experienced in the present, it is always past-directed; at the same time, although regret is directed toward our past, it has the capacity to exert a strong influence on our future (Hoerl and McCormack, 2016, p. 246).
To meet our present needs it is sufficient to propose the emotion of regret as the main phenomenological component of the undoing of one's past; therefore, it is equa ly sufficient to propose that the capacity for regret is essential to moral agency, e ecia ly because of the temporal scope of it, crossing past, present and future directions.
The emotion of regret, a ditiona ly, shares at least two a ects with the commonsensical attribution of responsibility: we do not feel regret or are held responsi le for (a) what could not have been done otherwise, and (b) for what has not been done by ourselves. We do not regret the incapacity to fly without some ap aratus, as we are not responsi le for that. We do not regret other people's choices, as we are not, in most cases, responsi le for them.
Recovering the previous discussion on the dynamics of moral agency, we fina ly propose the general capacity for episodic counterfactual thinking as one of the tools we must use to properly conceive ourselves as temporal extended entities. Considering the taxonomy of four basic modes of FMTT or pro ection, we think it is sound to ar ue for the functional role of two counterpart mental processes, namely, (1') counterfactual e iso ic simulation, or the construction of a mental representation of an autobiographical counterfactual event, and (2') counterfactual e iso ic prediction, or the estimation of the likelihood of one's rea ion to an autobiographical counterfactual event. 3 It is not difficult to perceive that the a ivation of regret, for example, depends on both (1') and (2').
The temporal extension of the moral selves, in the direction of episodic counterfactual scenarios, a ds a modal dimension to moral agency. It is not enough to travel back and forth in the subjective time to have a complete picture of moral agency. The we l-documented contribution of regret in decision-making processes shows that we must take seriously the capacity of undoing one's past as an essential a ect of capa le moral agency.
A ditiona ly, we must recognize that the "same brain area ar ument" equa ly works here to integrate the emotion of regret, decision-making, episodic future thinking and episodic memory.
In a paper by Sacha Bourgeois-Gironde one can find a fine example of this: The integration of regret in decision theory has been supported by recent neurobiological investigation. Present studies on the neural correlates of regret take advantage of previous observations on the role of the orbitofrontal cortex in the processing of reward and its role on subsequent behaviour. Rolls (2000) has evidenced the incapacity of orbitofrontal patients to modify their behaviour in response to negative consequences. Ursu and Carter (2005) have demonstrated how the anticipated affective impact of a choice was modulated by the comparison between the different available alternatives. These reasoning patterns, con-Filosofi a Unisinos -Unisinos Journal of Philosophy -19(1):97-102, jan/apr 2018 sisting of anticipating contrasts between actual outcomes and counterfactual ones (counterfactual in the sense that those outcomes are the ones that I would have got had I taken an alternative course of action), are reflected in the orbitofrontal cortex activity. More precisely, the impact of potentially negative consequences of choices is essentially represented in the lateral areas of the orbitofrontal cortex, whereas the medial and dorsal areas of the prefrontal cortex are more specialized in the impact of positive consequences (Bourgeois-Gironde, 2010, p. 250).

Final remarks
I think it is fair enough to claim that the idea of mental time travel has e a lished a solid beachhead in moral psychology, bringing effects for metaethics. And that this is great news for the re ective fields. After Ger ans and Kennett (2010Kennett ( , 2016 we have very good reasons to nourish second thoughts about the theoretical possibility of only making moral judgments either by the ap lication of a rule or by automatica ly responding to tacit affective processes. The fu l capacity for moral agency create a demand to menta ly project ourselves into future and past scenarios. And, we would like to a d, to menta ly project ourselves into counterfactual episodic scenarios. The functionality of the emotion of regret in decision-making is only the most pervasive evidence for an expected counterfactual dimension of the temporal extended moral self.