The philosophical literature on personal identity is vast, but a dominant view attributed to John Locke posits that psychological continuity and, specifically, memories are at the heart of personal identity: if “the soul of a prince, carrying with it the consciousness of the prince’s past life, enters and informs the body of a cobbler, as soon as deserted by his own soul, everyone sees he would be the same person with the prince, accountable only for the prince’s actions” (Essay, II, 27.17). That is, if two people were to switch bodies, the “real” person would be the person with their own mind (or “soul”), not their body. This view of personal identity as dependent on psychological continuity suggests that if one were to severely disrupt parts of a person’s mind (rather than their body), they might cease to be the same person after the change.

Traditionally, accounts based on psychological continuity have focused on memories as the key ingredient for identity persistence: Person X at Time 1 can be identified as Person Y at Time 2, if X and Y share an autobiographical memory (Locke 1694/1975). While philosophers have identified important flaws with psychological continuity accounts based solely on memories (Reid 1785/1975), empirical evidence suggests that such a memory-focused view does track folk intuitions (Nichols and Bruno 2010). But do laypeople really think that memories are what is most critical for identity persistence? Across five experiments, Strohminger and Nichols (2014) explored this, testing how other parts of the mind are perceived with respect to identity persistence. They found that morality was perceived as most central to identity. That is, when a person changed in terms of traits like honesty, empathy, or virtuousness, they were rated as more of a different person than when they changed in terms of their memories, preferences, or desires. This finding—henceforth “the moral self effect”—has been replicated by a number of independent groups (see Everett, Skorburg, and Livingston 2022 for a recent review), being seen in eight- to ten-year olds (Heiphetz et al. 2018) and Buddhist monks in India (Garfield et al. 2015). Moreover, it does not appear limited to hypothetical, abstract judgments, being seen in perceived identity persistence for patients with different neurodegenerative diseases (Strohminger and Nichols 2015) and people with addiction (Earp et al. 2019), and even having consequences for how we think about moral duties toward others (Everett, Skorburg, and Savulescu 2020).

Especially important for the present aims, Heiphetz, Strohminger, and Young (2017) found that changes to widely shared moral beliefs are perceived to be more disruptive to identity than changes to controversial moral beliefs. To explain this, they suggest a community hypothesis: perceived identity change is based on the extent to which changed attributes contribute to communal ties. Widely shared moral beliefs (e.g., “murder is wrong”) bind one to a community, and so changes to these moral beliefs are perceived as more disruptive to identity than changes to controversial moral beliefs (“abortion is wrong”) which may not bind in the same way. These results fit nicely with a related body of research showing that moral character (both one’s own and the character of others) is a central element of social perception (Brambilla and Leach 2014; Goodwin, Piazza, and Rozin 2014).

In this chapter we focus on a key question about this moral self effect: Does the importance of morality for identity persistence depend on the target of evaluation? Existing empirical work suggests that morality is just as important when people judge their own identity persistence as when they judge others. For example, some of the earliest work on the moral self effect demonstrated that moral change was more important than memories regardless of whether it was presented in a first- or third-person perspective (Prinz and Nichols 2017). More recently, no difference was observed in a series of studies directly comparing the moral self effect for self and a hypothetical other (“Chris”), though there were stronger effects of changes in certain moral traits on the other compared to self (Heiphetz et al. 2017). And yet, there are reasons to think that perhaps the importance of morality for identity persistence could depend on who we’re thinking about.

At least two (not mutually exclusive) lines of reasoning suggest that the importance of morality for perceived identity could vary across targets—especially as the distance becomes greater and participants move from thinking about themselves or friends to people they actively dislike or feel negatively about. First, people consistently overestimate their own morality compared to others (Epley and Dunning 2000), judge their friends as being better than the average person (Brown 1986), and think morality is a more important trait for the in-group than the out-group (Leach, Ellemers, and Barreto 2007). If people believe that they themselves and those close to them are more moral than average, when participants imagine themselves becoming more immoral, this should lead to a large degree of perceived identity change. This, of course, is what previous work on the moral self effect has demonstrated. However, when thinking of a stranger (assumed to be morally average) or someone that the participant does not like (assumed to be less moral than average), we might expect the moral self effect to be attenuated: in fact, disliked targets may even to some extent be defined by their immorality. Second, according to the community hypothesis (Heiphetz et al. 2017), morality is seen as more important for identity persistence because it binds people together, and so when evaluating someone we do not wish to be bound to (i.e., someone we dislike or don’t get along with), morality might cease to be so important. To explore the generalizability of the moral self effect, we had participants evaluate changes to morality, warmth, competence, memories, and preferences for a number of different targets: the self (Study 1), a friend (Study 2), a stranger (Study 3), or a foe (Study 4). In answering this primary question, we take the opportunity to answer two further questions.

First, we ask, does the direction of change matter, and does this interact with the target of evaluation? While the vast majority of studies on the moral self effect have considered the impact of losing memories or losing morals on perceived identity persistence, it seems plausible that gaining memories or gaining morals might have different consequences. Indeed, focusing on morals (rather than memories), work on the “true self” suggests that the direction of moral change could matter. A growing number of studies have suggested that people typically regard others’ true selves as being fundamentally good (Bench et al. 2015; De Freitas and Cikara 2018; Newman, Bloom, and Knobe 2014; Strohminger, Knobe, and Newman 2017), so that as people become more moral they are perceived to get closer to their true self.

In the context of identity persistence, Tobia (2015) draws on the well-worn (if potentially apocryphal) case study of Phineas Gage: a railroad worker who experienced brain damage in a horrific accident, after which he was reported to have become cruel and impulsive—so much so that “he was no longer Gage.” In his work, Tobia gave participants one of two versions of this story. In a moral deterioration condition, participants saw the “standard” case of Phineas Gage, where he was kind before the accident, but cruel afterward. In another moral improvement condition, holding the magnitude of the change constant, participants saw a vignette where Gage was described as cruel before the accident, but kind afterward. In both conditions, Tobia asked participants to judge whether Phineas Gage was the same person as before the accident, and found that Gage was less likely to be judged as identical to his pre-accident self when the change was in a “bad” direction (deteriorating from kind to cruel) than when the change was in a “good” direction (imp...