Abstract
Humans rely heavily on their prosocial relationships. We propose that the experience and display of prosocial emotions evolved to regulate such relationships through inhibiting individual selfishness in service of others. Two emotions in particular serve to meet two central requirements for upholding prosociality: gratitude motivates maintenance of ongoing prosocial interactions, and guilt motivates repair of ruptured prosocial interactions. We further propose, and review developmental evidence, that nascent forms of these two emotions serve their respective functions from early in ontogeny. The remarkably early emergence of these prosocial emotions allows even very young children to participate in and benefit from prosociality. We discuss the implications of and challenges to this conclusion and identify pressing future directions for this work.
The Development of Prosocial Emotions
No other animal plays non-zero-sum games as tirelessly as we do. Much of your emotional life is natural selection’s way of getting you to play. Gratitude for favors rendered and guilt over neglecting a friend help you start or sustain potentially win–win games. (Wright, 2000, p. 59)
Humans depend profoundly on prosocial interactions with one another, that is, interactions in which one acts to benefit another, often at a cost to oneself (Tomasello, 2016). As Darwin (1874) observed, humans benefit greatly from helping those in need, exchanging goods and services, reciprocating favors, assisting one another with child care, and so forth. When such prosocial interactions are directed towards kin, they can be explained by kin selection wherein an individual’s genetic fitness is measured not only by the fitness of the individual and their offspring, but also by the fitness of other genetically related individuals (Hamilton, 1964). Strikingly, however, humans also form and rely on prosocial, reciprocal relationships with nonkin (Seyfarth & Cheney, 2012; Sober & Wilson, 1998). This ultraprosociality is believed to account for our survival and success as a species (Tomasello, 2009, 2016). But relationships with nonrelatives are also susceptible to cheating and defection that, if left unchecked, threaten the relationships and their immense benefits. Humans have thus evolved a wide range of psychological adaptations that promote our prosocial relationships (Bjorklund, 2018; Tooby & Cosmides, 1992). On the proximate level, these adaptations include emotions that regulate social relations, safeguard against selfishness, and promote our prosocial relationships (Frank, 1988).
Yet little attention has been paid to the development of these emotions. Infants and young children across cultures demonstrate remarkable prosocial behaviors, even towards nonrelatives (Callaghan et al., 2011; Kärtner, 2018; Warneken & Tomasello, 2009b). This suggests that from early in development, humans are equipped with psychological attributes including emotions that enable children to engage in and profit from prosocial interactions (Bjorklund, 2018; Fessler & Haley, 2003; Warneken, 2015). What is the nature of these emotions and how do they motivate young children to be prosocial rather than to act purely out of self-interest? Answers to these questions are crucial for gaining a fuller understanding of the roots of human prosociality. Here we are concerned with what we consider to be two fundamental requirements for upholding prosociality. First, once a prosocial interaction has been initiated, it must be maintained. Second, when a prosocial interaction breaks down, it must be repaired. Without these requirements being met, our vital social relationships will collapse. Our aim here is thus to identify in early ontogeny the emotions that motivate children to maintain and repair their prosocial interactions.
Prosocial Emotions
Prosocial behavior faces at least two challenges. One, we must inhibit our selfish tendencies in order to fulfill our commitments to our partners, and two, we must identify individuals who are committed and faithful partners so that we can enter into and maintain beneficial, long-term relationships with them (Frank, 1988; Keltner, 2009). We propose that emotions are critical to solving both challenges. This view aligns with a functional approach to emotions more generally wherein emotions are seen as adaptations that motivate us to behave in ways that help us solve challenges of adaptive import (Campos, Campos, & Barrett, 1989; Keltner & Haidt, 1999; Nesse, 1990).
First, our experience of certain emotions informs us about, and motivates us to respond to, specific social situations. For example, anger informs us about the injustice of an interaction and prepares us to rectify it (Solomon, 1990). Second, others’ emotional displays help us infer their mental states and predict their future actions, thus allowing us to effectively respond to and form expectations about social partners (see Keltner & Haidt, 1999). For example, others’ displays of distress inform us about their suffering, which moves us to help them. The experience and display of emotions are thus critical for regulating our social lives.
We apply this view of emotions to the two fundamental requirements for safeguarding prosocial relationships: (a) once a prosocial interaction has been initiated, it must be maintained, and (b) when a prosocial interaction breaks down, it must be repaired. We propose that gratitude enables us to meet the first requirement, and guilt enables us to meet the second. Because they are key to motivating prosocial behaviors that are so essential for our survival and success as a species, we subclassify these emotions as prosocial emotions.
It is well established, and we review in what follows, that among adults, gratitude and guilt promote relationship maintenance and repair, respectively. However, crucial insights into these prosocial emotions come from studying their early ontogeny. A developmental perspective provides unique opportunities to study the components of a system or mechanism, and to see when and how these components emerge and interact before they are fully mature and operational in adults (de Haan & Gunnar, 2009; Gottlieb, 2007). A developmental approach to prosocial mechanisms is thus indispensable to understanding human prosociality more generally. In this article, therefore, we review the research on when in human ontogeny nascent forms of gratitude and guilt begin to serve their vital prosocial functions. Note that our goal here is not to study the phenomenology of early gratitude or guilt but to understand the mechanisms underlying early prosociality. In other words, our focus is on the early prosocial functions of these emotions rather than on the early experience and nature of the emotions themselves.
Our focus on gratitude and guilt is based on our proposal that these emotions allow children to flexibly maintain and repair ongoing prosocial relationships in ways that other, arguably more basic, emotions do not. Most prominently, sympathy (or concern) is a vital, early emerging prosocial motivator. Starting in the first year and increasingly during the second year, children respond to those in need or distress through facial and verbal expressions of concern and prosocial actions such as helping and comforting, and their degree of concern predicts their degree of prosociality (e.g., Eisenberg & Fabes, 1990; Vaish, Carpenter, & Tomasello, 2009; Zahn-Waxler, Radke-Yarrow, Wagner, & Chapman, 1992). During the second and third years, sympathy becomes increasingly sophisticated such that children sympathize with victims even when victims display no overt distress and sympathize less with those who display unjustified distress (Vaish, 2016). From early on, then, sympathy robustly makes us care about and enhance the welfare of others. Importantly, although sympathy promotes prosocial behavior towards those in need, it does not help us identify when our valued relationships require attention and work and is thus not specifically tuned to monitoring and maintaining our prosocial relationships as gratitude and guilt are. We will thus not focus on the early prosocial functions of sympathy (but see, e.g., Davidov, Zahn-Waxler, Roth-Hanania, & Knafo, 2013; Eisenberg, Spinrad, & Morris, 2014; Hoffman, 2000; Vaish, 2016, for reviews).
More recently, developmentalists have stressed that prosocial behaviors emerge not only in negative contexts (e.g., in response to distress) but also in positive contexts. In particular, infants’ and children’s prosocial and collaborative behaviors are often accompanied by positive emotions such as engagement, affection, and enjoyment (Hammond & Brownell, 2018; Rheingold, 1982). Parents also encourage infants to help, and praise and thank their children for helping; this positive social feedback scaffolds early prosocial behavior (Dahl, 2015, 2018; Dahl et al., 2016).
Engaging in prosocial behavior also elicits positive affect. By age 2, children show the “warm glow” of giving: a positive emotion after acting prosocially. Notably, children display more positive affect after giving away their own reward (paying a cost) than after giving away others’ rewards (paying no cost) or receiving a reward (Aknin, Hamlin, & Dunn, 2012). Likewise, children show more positive affect after sharing spontaneously than upon request (Lennon & Eisenberg, 1987), and after they benefit themselves or someone else than after they benefit no one (Hepach, Vaish, & Tomasello, 2017b). Prosocial behavior thus elicits positive emotions, which in turn reinforce prosocial behavior (Aknin, van de Vondervoort, & Hamlin, 2018). Important as these positive emotions are for promoting early prosociality, they, like sympathy, do not specifically help children monitor and maintain their prosocial relationships.
Moreover, basic emotional processes such as sympathy and enjoyment are arguably sufficient to maintain prosocial behavior within the family context, where relationships are based on kinship rather than expectations of reciprocity and mutual benefit (Osinski, 2009; Stewart-Williams, 2007), and where prosocial behavior is naturally supported by the strong evolutionary forces of kin selection (Hamilton, 1964). Kin relationships may thus not require complex emotional mechanisms that carefully track favors and slights or allow for exact bookkeeping. It is likely in the larger group context—where relationships with nonkin are based on mutually beneficial, reciprocal interactions guided by principles of fairness and where we must be more vigilant about finding and keeping trustworthy partners—that we rely more heavily on complex prosocial emotions such as gratitude and guilt (Roberts, 2005; Rotkirch, Lyons, David-Barrett, & Jokela, 2014; Tomasello, 2016). These emotions allow us to monitor and regulate more vulnerable but still vital relationships with nonrelatives. As we aim to shed light on the early prosocial emotions that safeguard these prosocial relationships, our focus here is on gratitude and guilt.
Gratitude: Maintaining Prosocial Relationships
Gratitude is a positive emotion that arises from the perception that one has benefited through the good intentions of another (McCullough, 2002). 1 For centuries, social theorists have recognized the importance of gratitude for creating and sustaining social relations (see Harpham, 2004). It is argued that gratitude has proven so useful for group living that it has become foundational to human nature (McAdams & Bauer, 2004). We consider here how the experience and display of gratitude in early childhood regulates children’s prosocial interactions.
The Experience of Gratitude
The experience of gratitude urges a grateful person to respond prosocially in turn (Algoe, 2012; Bartlett, Condon, Cruz, Baumann, & DeSteno, 2012; McCullough, Kimeldorf, & Cohen, 2008). Gratitude thus turns (selfish) receivers into givers, thereby upholding the cycle of reciprocity that sustains human cooperation (Bonnie & de Waal, 2004; Trivers, 1971). Empirical work with adults provides evidence for this reciprocal function of gratitude (see Ma, Tunney, & Ferguson, 2017). For instance, participants who were grateful to an experimenter for helping them avoid a tedious task were more likely to subsequently help the experimenter than participants who were not grateful (Bartlett & DeSteno, 2006). Importantly, gratitude resulted in more prosociality than simply a positive mood or a positive evaluation of benefactors (Tsang, 2006a, 2006b; Tsang, Schulwitz, & Carlisle, 2012).
Yet little is known about the development of gratitude and its prosocial functions. Some work suggests that by age 4, children know that gratitude is a positive emotion but not what kinds of situations elicit gratitude; this understanding only emerges by age 10 (Harris, Olthof, Terwogt, & Hardman, 1987; Russell & Paris, 1994). Further, the experience of gratitude has been argued to emerge only in middle childhood. For instance, Graham (1988) presented 6-, 8-, and 10-year-old children with hypothetical scenarios in which a child is chosen for a team by a captain who was either required or not required to choose the child. The 8- and 10-year-olds, but not the 6-year-olds, inferred that the child would be more grateful and more likely to reciprocate if the captain picked him voluntarily and thus intended to benefit him (see also Poelker & Kuebli, 2014). Gleason and Weintraub (1976) also found that few 6-year-olds said “thank you” when given gifts, whereas most 10-year-olds did so. Further, children’s gratitude becomes increasingly sophisticated between 7 and 14 years, with older children’s gratitude increasingly aimed at engaging and connecting with benefactors (Wang, Wang, & Tudge, 2015). These studies suggest that gratitude emerges around 6–7 years and develops slowly thereafter. However, they do not provide sensitive measures of gratitude at younger ages as they rely on hypothetical scenarios, which are challenging for young children, or on children’s use of verbal routines, which may not indicate how children feel but what they are socialized to say (Poelker & Kuebli, 2014).
Other work has relied on interactive situations and behavioral measures and revealed gratitude-like patterns in remarkably young children. One line of work shows that children engage in direct reciprocity from early on. For instance, 3-year-olds but not 2-year-olds share more resources with a partner who previously shared resources with them than with one who did not (Warneken & Tomasello, 2013; see also Levitt, Weber, Clark, & McDonnell, 1985). When forced to choose between a person who helped versus did not help them, even 2-year-olds preferentially help their benefactor (Dunfield & Kuhlmeier, 2010). Direct reciprocity thus emerges early in ontogeny.
However, these studies leave unclear whether children’s reciprocity is in fact motivated by gratitude. Children may respond positively to all generous individuals, even if the individuals were generous towards others and not the child. Indeed, toddlers and preschoolers prefer and help those who are prosocial towards third parties more than those who are antisocial (e.g., Dahl, Schuck, & Campos, 2013; Kenward & Dahl, 2011; Vaish, Carpenter, & Tomasello, 2010). Alternatively, children may experience a positive mood whenever they receive benefits from another individual, regardless of whether or not the other individual intended to benefit them. Positive mood promotes adults’ prosocial behavior (Carlson, Charlin, & Miller, 1988), and may do so among young children as well. But if gratitude motivates early reciprocity, then children should take both of the aforementioned factors into account (whether the benefactor benefited the child and whether the benefits were provided intentionally), such that they are particularly appreciative and prosocial towards a benefactor who intentionally benefited them. This is based on the proposal and evidence that gratitude is experienced upon receiving a benefit through the good intentions of another, and urges more prosociality than simply a positive mood or a general positive evaluation of benefactors (McCullough et al., 2008; Tsang, 2006a).
A recent study teased apart these mechanisms (Vaish, Hepach, & Tomasello, 2018). Three-year-old children either received benefits from a helpful puppet (who intentionally handed them resources they needed) or from an indifferent puppet (who placed the resources near them but did not convey an intention to help). In two other conditions, 3-year-olds watched as another individual received benefits from the helpful or the indifferent puppet. When children could subsequently share with the benefactor, they did so significantly more if the puppet had intentionally helped them than in the other conditions. In other words, children’s reciprocity was not simply motivated by a positive mood, as children who received benefits from the indifferent puppet did not reciprocate as much. Children’s reciprocity was also not motivated by a general positive evaluation of benefactors, as children who observed the benefactor intentionally help someone else did not reciprocate as much. Rather, children were sensitive to both of the key factors associated with gratitude. The pattern of their reciprocal sharing suggests that the core elements of gratitude may be present and motivate reciprocity by age 3.
This study did not include children younger than 3, but in other research, 2-year-olds did not help or share reciprocally (Warneken & Tomasello, 2013), seemingly precluding the possibility of gratitude at this age. However, perhaps an emerging sense of gratitude is present before age 3 but is only apparent as a change in children’s underlying motivation to help. In particular, gratitude is argued to motivate us to be “instrumental in promoting” the well-being of our benefactors (A. Smith, 2006), that is, to actively and selectively help those who have helped us before. It is possible that being helped changes even young children’s reciprocal motivation if not their reciprocal behavior.
To test this, Hepach, Vaish, Müller, and Tomasello (2019a) examined both the reciprocal behavior and the underlying reciprocal motivation of children younger than 3. This study used pupil dilation as an index of children’s prosocial motivation. Before delving into the findings, we briefly describe pupil dilation and its relevant interpretations (see Hepach & Westermann, 2016, for a review). Systematic changes in pupil diameter are linked to activation of the sympathetic branch of the autonomous nervous system (Loewenfeld, 1993). Pupil diameter increases in response to emotionally charged images or sounds and to stimuli that hold motivational significance (Bradley, Miccoli, Escrig, & Lang, 2008; Nieuwenhuis, De Geus, & Aston-Jones, 2011; Partala & Surakka, 2003). Psychologically mediated effects on pupil diameter are evident from infancy (Laeng, Sirois, & Gredebäck, 2012; Sirois & Brisson, 2014).
Recent research further indicates that pupil dilation indexes children’s prosocial motivation: children between 1.5 and 5 years show increased pupil dilation upon seeing someone in need of help, and the greater the increase, the likelier and faster children are to subsequently help that person (Hepach et al., 2019a). More pertinently, toddlers’ pupil dilation decreases similarly when they themselves provide help and when they see someone else provide help (Hepach, Vaish, & Tomasello, 2012). Toddlers are thus not primarily motivated to get credit for their helpful actions but rather to see the person in need be helped. However, in the study in question—on the prosocial motivations created by early gratitude—Hepach, Vaish, Müller, and Tomasello (2019b) asked whether gratitude-like situations modify this motivation such that children become motivated to help their benefactors themselves.
Children aged 1.5 and 2 years first either received help from an experimenter or did not need and thus did not receive help (but were familiar with the experimenter). Then the experimenter needed help and children were either able to help or another adult intervened to provide help. The rate at which children attempted to help did not differ across conditions, replicating prior work that 2-year-olds do not help reciprocally (Warneken & Tomasello, 2013). However, the underlying motivation to help was differentiated. If the person in need was simply a familiar experimenter (not a benefactor), children’s internal arousal decreased both when they or the other adult provided help. This fits prior findings that, generally speaking, young children want to see a person in need be helped, regardless of who helps (Hepach et al., 2012). Crucially, if the person in need was their benefactor, children’s arousal remained high if they could not actively help but rather saw the other adult provide help. Thus, prior to age 3, receiving help did not increase toddlers’ reciprocal behaviors but did alter their underlying motivation to help. A follow-up study found that being helped did not change children’s prosocial motivation towards a new individual: their motivation to reciprocate was modified specifically towards the benefactor (Hepach et al., 2019b). In other words, when a person in need had previously benefited them, children wanted to both actively and selectively help that person.
During the second and third years of age, then, children begin to recognize when they have been the recipients of someone’s goodwill and are motivated to uphold that prosocial interaction. This pattern is consistent with an early sense of gratitude, evident both in changes in young children’s internal arousal by 1.5–2 years of age and in their overt reciprocity by 3 years of age. We note, however, that these studies did not provide a direct measure of gratitude; rather, the patterns of children’s helping motivation and helping behavior were consistent with the patterns predicted by gratitude, and each study accounted for some compelling alternative explanations such as positive evaluation of generous individuals or positive mood resulting from receiving benefits. Still, it is difficult to claim that children in these studies were motivated by full-fledged gratitude. This is a point we return to later.
The Display of Gratitude
Why do grateful individuals display their gratefulness to others? What do gratitude displays convey to our benefactors and to bystanders? Theorists have proposed that displays of gratitude indicate that one has noticed a kindness, one appreciates the kindness, and one is likely to reciprocate in the future (McCullough et al., 2008). As such, gratitude displays convey that the grateful person is aware of the reciprocity norms, is unlikely to act selfishly, and is committed to her prosocial relationships (cf. Keltner, Horberg, & Oveis, 2006). It follows, we argue, that gratitude displays should elicit affiliation and cooperation from benefactors and bystanders alike (see Algoe, Dwyer, Younge, & Oveis, 2019).
Empirical work with adults supports this proposal. For instance, when a beneficiary thanks a benefactor, the benefactor views the beneficiary more positively and is more willing to help the beneficiary and others than when a beneficiary does not thank the benefactor (see McCullough, Kilpatrick, Emmons, & Larson, 2001). Adults who witness gratitude being expressed to a third party are also more helpful towards the grateful person (Algoe et al., 2019). Expressions of gratitude within a relationship are also associated with improved long-term relationship quality and greater relationship maintenance behavior (Algoe, Fredrickson, & Gable, 2013; Algoe, Haidt, & Gable, 2008; Lambert & Fincham, 2011).
When in development do gratitude displays begin to serve these important prosocial functions? To our knowledge, only one study has examined this question thus far (Vaish & Savell, 2018). Four- and 5-year-old children watched videos of a benefactor giving gifts to two beneficiaries. One beneficiary displayed gratitude, whereas the other showed happiness but not gratitude. The gratitude display involved acknowledging the benefactor’s generosity, expressing appreciation for the gift, and affirming the relationship. The grateful beneficiary did not say “thank you” as it is a heavily socialized verbal routine; children may thus like beneficiaries who say “thank you” because the beneficiary behaved politely or normatively rather than because they value the underlying sentiment of gratitude (Poelker & Kuebli, 2014). The 5-year-olds preferred the grateful beneficiary and expected the benefactor to prefer her as well, and they distributed more resources to the grateful beneficiary. The 4-year-olds showed a similar but weaker pattern of responses. Gratitude displays thus elicit affiliation and cooperation from observers, and do so by the preschool years.
This is, of course, only the first study and many questions remain open. Most importantly, do gratitude displays have the same effects within first-party interactions, that is, when children are the benefactors towards whom gratitude is displayed? Such first-party interactions are the quintessential prosocial relationships that need to be maintained, and so we may predict that children will respond positively to gratitude displays in such interactions as well. Indeed, children may do so more robustly and at an earlier age than in third-party interactions, given that beneficiaries’ displays of gratitude may evoke stronger positive responses in benefactors than in bystanders. More work is needed to expand our understanding of the functions of gratitude displays in development.
Guilt: Repairing Prosocial Relationships
Guilt is an aversive, self-conscious emotion (evoked by self-reflection and self-evaluation) that follows the realization that one has harmed another individual (Nelissen & Zeelenberg, 2009). 2 Guilt was long viewed as an intrapsychic phenomenon that had very little to do with one’s social world. In classic psychoanalytic theory, for instance, guilt is conceived as a response to one’s own unacceptable impulses. More recently, though, it has become clear that guilt is an interpersonal phenomenon that is inherently linked to the relationships between individuals (see Baumeister, Stillwell, & Heatherton, 1994).
The Experience of Guilt
When a transgressor experiences guilt, she focuses on and feels subjective discomfort about the harm that she caused, and crucially, she is motivated to make amends by aiding or otherwise compensating the victim (Hoffman, 1982; Keltner, 1995). Such reparative behavior serves to restore a sense of justice, repair possible rupture of the relationship, and promote social attachment (Baumeister et al., 1994). Guilt is thus tuned to identifying and repairing damaged social relationships. Research with adults provides support for these vital functions of guilt. For instance, undergraduates who felt guilty after behaving uncooperatively in a decision-making game were subsequently more likely to behave cooperatively towards their partner than undergraduates who did not feel guilty (Ketelaar & Au, 2003; see also Brock & Becker, 1966; Carlsmith & Gross, 1969).
The experience of guilt serves these prosocial functions from early on. By as early as 2 years of age, children who accidentally cause minor harm (e.g., break a person’s favorite toy) show signs of experiencing guilt such as accepting responsibility and trying to repair the damage (Drummond, Hammond, Satlof-Bedrick, Waugh, & Brownell, 2017; Kochanska, Casey, & Fukumoto, 1995; Zahn-Waxler & Kochanska, 1990). However, it is unclear whether these studies measured guilt or only one of its components. Guilt consists of two basic components: sympathy (feelings of concern) for a victim of harm and the realization that one caused that harm (Hoffman, 1982; Regan, 1971). Each component separately can motivate reparative behavior, but neither is by itself guilt. For instance, when children harm someone, their attempts to repair could be motivated purely by sympathy for the victim—without the realization that they caused the harm. Yet sympathy is not guilt, as it lacks the second crucial component of guilt: the awareness of having caused the harm or distress (Hoffman, 1982). Thus, although it motivates prosocial behavior towards those in need, it is not specifically tuned to identifying and repairing one’s own damaged prosocial relationships. The crucial question, then, is whether guilt further hones young children’s prosocial motivation in a way that is specifically geared towards repairing their ruptured relationships.
Some recent work has begun to address these questions. One study crossed the two key components of guilt: sympathy and awareness of having caused an outcome. Vaish, Carpenter, and Tomasello (2016) examined 2- and 3-year-old children’s reparative behavior after they caused unintentional harm: a ball they placed on a marble run accidentally rolled out and knocked down an intricate tower that an experimenter had built. This guilt condition was compared with a sympathy condition in which children saw another person accidentally cause the identical harmful outcome, as well as conditions in which children or another person caused the same outcome but in a nonharmful context (the tower did not belong to anyone). The study revealed that by age 3 (but not age 2), children displayed more physical and verbal repair (e.g., spontaneously attempting to rebuild the tower or saying that they could rebuild it) in the guilt than in the other three conditions. Thus, by age 3, guilt resulted in greater reparative motivation than only sympathy or only the urge to repair nonharmful damage. Importantly, as would be expected based on prior work (Vaish et al., 2009; Zahn-Waxler et al., 1992), 2-year-olds did show an effect of sympathy: they acted more prosocially in the two conditions in which the experimenter was harmed than in those in which she was not harmed. However, 2-year-olds did not show the specific effect of guilt that 3-year-olds did. Guilt thus motivates reparative behavior by at least age 3.
Another study took a different approach to the question of the early prosocial functions of guilt (Hepach, Vaish, & Tomasello, 2017a). The study asked: do children who have caused someone harm feel compelled to repair the harm themselves, that is, because they recognize that they harmed the relationship and therefore they must fix it? This study again measured changes in pupil dilation to index children’s prosocial motivation. Two- and 3-year-olds either caused someone accidental harm (spilling water on an experimenter’s work) or watched another adult cause the same harm. Then children could either repair the harm by wiping up the spill or watched as a different person (another adult) repaired the harm. Three-year-olds (and to some extent even 2-year-olds) showed decreased arousal when they could repair the harm that they had caused, but their arousal remained high if the other adult repaired the harm that they had caused. In contrast, if someone else caused the harm, then children’s arousal decreased to a similar extent regardless of whether they or the other adult repaired it (as in Hepach et al., 2012). Thus, as bystanders, children sympathize with victims of harm and are motivated to see a person in need be helped, regardless of who provides the help. However, when children cause the harm, their motivation for helping changes such that they want to provide the help, perhaps to show their commitment to the partner and restore the ruptured relationship. Already by age 2, nascent experiences of guilt may change children’s motivation from wanting to see others helped to wanting to actively repair, though the motivation may not yet manifest itself in children’s observable repair behaviors (Vaish et al., 2016).
Thus, by 3 and to some extent even 2 years of age, children recognize when they have caused harm and are motivated to repair that harm. Our take on this pattern of findings is that a sense of guilt helps repair disrupted prosocial interactions from remarkably early in ontogeny (see also Vaish, 2018). Again, however, these studies did not provide a direct measure of guilt; rather, the pattern of children’s helping behavior and motivation was consistent with the prosocial functions of guilt, and each study accounted for some of the most compelling alternative explanations, such as sympathy for victims more generally. We return later to discuss this point.
The Display of Guilt
Transgressors not only experience guilt but also display it facially and verbally. Such displays of remorse are thought to appease the victim as well as bystanders by signaling that the transgressor is also suffering, she did not mean harm and is not generally a harmful person, and she plans to make amends and do better in the future (Castelfranchi & Poggi, 1990; Keltner & Anderson, 2000; McGraw, 1987). As a result, remorseful transgressors should be judged as more reliable social partners and deserving of more forgiveness, affiliation, and cooperation than unremorseful transgressors (Darby & Schlenker, 1982, 1989; Goffman, 1967). Indeed, research with adults shows that both victims and bystanders evaluate remorseful transgressors more positively and punish them less than they do to unremorseful transgressors (Ohbuchi, Kameda, & Agarie, 1989; O’Malley & Greenberg, 1983). Guilt displays thus move victims and bystanders to allow the repair of ruptured relationships with transgressors who demonstrate their cooperative commitment.
These reparative functions of guilt displays emerge early in development. One line of research has examined children’s responses to apologies, which are a common stand-in for guilt. For instance, in response to stories in which a moral transgressor did or did not apologize, children between 4 and 9 years blame and punish the apologetic transgressor less, and forgive and like her more than the nonapologetic transgressor (Darby & Schlenker, 1982, 1989; C. E. Smith, Chen, & Harris, 2010). Children also think that stories in which a transgressor apologizes are more just, and that a victim who received an apology will feel better (Irwin & Moore, 1971; C. E. Smith et al., 2010; Wellman, Larkey, & Somerville, 1979). Transgressors’ apologies thus appease preschool-age children.
However, caregivers regularly prompt their children to apologize, even when children might not feel guilty (C. E. Smith, Noh, Rizzo, & Harris, 2017). Thus, children’s positive responses to apologetic transgressors may in fact be responses to hearing routinized and reinforced phrases such as “sorry” rather than to the transgressor’s remorse. Researchers have worked around this by taking conventional apology terms out of the transgressor’s remorse display. In one study (Vaish, Carpenter, & Tomasello, 2011), 4- and 5-year-olds watched two third-party transgressions. In one, the transgressor was remorseful without saying “sorry” (“I didn’t mean to do that. It’s my fault”). In the other, the transgressor was neutral but unremorseful. Five-year-olds showed a strong preference for the remorseful transgressor: they evaluated her more positively, preferred to interact with her, and gave her more resources. However, 4-year-olds did not show any systematic preference. To assess whether 4-year-olds’ performance suffered from task demands, Vaish et al. (2011) conducted a follow-up study in which the remorseful transgressor did apologize by saying “sorry.” Now, as in prior work (C. E. Smith et al., 2010), 4-year-olds showed a strong preference for the apologetic transgressor. Thus, by age 4, children respond positively to transgressors’ remorse when it includes explicit apologies, but about a year later, children recognize and respond positively to displays of the prosocial emotions that apologies represent, namely, guilt and remorse.
Importantly, for a ruptured relationship to be repaired, transgressors must appease and make amends not only with bystanders but also (perhaps especially) with victims. Yet victims are directly involved in and affected by the transgression and may thus find it more challenging than bystanders to respond positively to a transgressor’s remorse. This additional complexity makes it critical to examine the effects of guilt displays in first-person contexts. Here, again, some work has assessed children’s responses to explicit apologies. After suffering minor transgressions, 4- to 7-year-olds who receive an apology report improved feelings, evaluate the transgressor more positively, and act more prosocially towards the transgressor than children who receive no apology (Drell & Jaswal, 2016; C. E. Smith & Harris, 2012). A more recent study excluded explicit apologies (Oostenbroek & Vaish, 2019) and revealed very similar results as when children were bystanders: 5-year-olds were more forgiving of the remorseful transgressor, as seen in their positive evaluations, preference, and distribution of resources; 4-year-olds did not show any systematic effects, but when an explicit apology was included, 4-year-olds were also more forgiving of the apologetic transgressor. Thus, even in first-person contexts, transgressors’ displays of remorse motivate young children to forgive and reenter prosocial interactions with their transgressors.
In sum, by the late preschool years, children effectively use transgressors’ displays of guilt to identify reliable, committed social partners and thus restore valuable relationships. The display of guilt thus serves vital prosocial functions from early in ontogeny.
Discussion and Future Directions
The experience and display of prosocial emotions evolved to regulate the prosocial relationships that humans depend so profoundly on. Two emotions in particular serve to meet two central requirements for upholding our prosocial relationships: gratitude motivates maintenance of ongoing prosocial relationships, and guilt motivates repair of ruptured prosocial relationships. In contrast to previous theories based on research with adults (Haidt, 2001; Keltner, Haidt, & Shiota, 2006; Krebs, 2008), our argument is fundamentally developmental. Specifically, we propose that nascent forms of prosocial emotions serve their respective functions from early in development, and we reviewed recent developmental evidence in support of this claim. Together, these findings suggest that gratitude-inducing situations promote reciprocal behavior by 3 years of age, and alter children’s intrinsic motivation to reciprocate by age 2. Gratitude displays serve further prosocial functions by the preschool years (ages 4 to 5), when children begin to appreciate and behave more prosocially towards beneficiaries who display gratitude than towards those who do not. Similarly, guilt-inducing situations promote reparative behavior by age 3, and affect children’s intrinsic reparative motivation by age 2. By 4 to 5 years, transgressors’ displays of guilt further promote reparative behavior as children are appeased by, forgive, and are more prosocial towards transgressors who display guilt than towards those who do not. Recent advances in developmental research thus indicate that the remarkably early emergence of gratitude and guilt allow even young children to regulate their prosocial relationships.
It is worth briefly considering here why these prosocial emotions and motivations should emerge so early in development. Evolutionary theory and anthropological evidence indicate that young children’s prosocial behaviors (such as hauling water or collecting firewood) contribute nonnegligibly to their groups’ subsistence (Warneken, 2015). Accordingly, children are motivated to be prosocial from a young age (Warneken & Tomasello, 2009a). The research on prosocial emotions reviewed here suggests that around 2–5 years of age, when children start to be weaned (Dettwyler, 2004), become more independent from their caregivers, and increasingly interact with and rely on nonrelatives and peers, their prosocial motivations become remarkably sophisticated, consisting not only of concern for others’ welfare but also of more nuanced concerns for safeguarding their own prosocial interactions and distinguishing reliable, trustworthy partners from unreliable, untrustworthy ones. We propose that these rather advanced prosocial motivations mean that young children support not only their groups’ subsistence but also their prosocial relationships with nonkin that are so vital for large-scale human societies. Prosocial emotions allow young children to uphold prosociality itself.
In the remainder of this article, we discuss the implications of and challenges to this conclusion, and some of the pressing future directions we see for this work.
What Develops? A Novel Theoretical Proposal
It is well established that even very young children help and cooperate with others and do so across cultures, thus providing evidence for an early emerging and intrinsic prosocial motivation (see Callaghan et al., 2011; Hepach et al., 2012; Kärtner, 2018; Warneken & Tomasello, 2009a). During the second and third years of life, as children’s helping becomes more robust, their intrinsic prosocial motivation becomes increasingly flexible. Situations that evoke gratitude motivate children to not only want to see their benefactor being helped but to provide the help themselves, and their physiological arousal remains high when someone else provides that help. Similarly, guilt-inducing situations motivate children to actively remedy the harm that they caused, and their physiological arousal remains high when they cannot do so. This pattern of results suggests that children are sensitive to situations in which not actively helping could disrupt their prosocial relationships. Harming others and being helped thus change the motivation underlying toddlers’ helping.
At the same time, during the first 3 years of age, children’s helping behavior may not yet be affected by guilt- and gratitude-like situations. This changes around 3 to 4 years of age, when children increase their prosocial behavior towards those whom they harmed and those who meant them well. One explanation for this developmental progression is that children become increasingly adept at acting on their motivations and in knowing how and whom to help (i.e., what constitutes appropriate help) in various situations. These advances are likely the result of developing skills, gathering experience, and receiving social guidance (Warneken & Tomasello, 2009a). Additionally, children become increasingly aware between these ages that they are being seen and evaluated by others and that their behaviors convey important social information to others (Engelmann & Rapp, 2018; Vaish & Tomasello, 2014). For instance, 3-year-olds engage in some self-presentation behaviors such as lying after cheating (Evans & Lee, 2013) and increasing their prosocial behavior under cues of being observed (Kelsey, Grossmann, & Vaish, 2018), perhaps to present themselves in a positive light. Children of this age may thus begin to recognize the importance of showing one’s commitment to others, such as by repairing harm that one caused or reciprocating help one received. This may turn the intrinsic motivations to repair and reciprocate seen at 2 years of age to the overt reparative and reciprocal actions seen around 3 years of age.
The third crucial developmental step is children’s ability to respond to others’ displays of prosocial emotions such as gratitude and guilt. Preschoolers at 4 to 5 years of age adjust their prosocial behaviors not only in response to their own experiences of causing harm or receiving help, but also in response to others’ displays of gratitude and guilt. Thus, even before children start formal schooling, they have a burgeoning ability to repair and maintain their valuable cooperative relationships.
Why do the experiences of gratitude and guilt serve their prosocial functions earlier in development than the displays of these emotions? This discrepancy might partially be a methodological by-product. Since studies on children’s responses to gratitude and guilt displays involve asking interview questions (which young children find challenging), these studies have not generally assessed children younger than age 4. However, a study that did include 3-year-olds found that these younger children did not evaluate apologetic transgressors more positively than nonapologetic ones (Wellman et al., 1979). Thus, perhaps children’s appreciation of others’ gratitude and guilt displays does emerge around 4 to 5 years of age. Why might this be?
One proposal is that as gratitude and guilt have no single, identifiable emotional expression, young children may struggle to identify these emotions in others. Indeed, children’s emotional understanding develops greatly during the preschool years (Harris, 1989; Pons, Lawson, Harris, & de Rosnay, 2003). It may thus take until about age 4 to gather sufficient experience and develop the capacity to identify and respond to the familiar normative expressions of the emotions (e.g., apologies), and until age 5 to other, subtler signs. Additionally, preschool-age children have difficulty grasping that traits remain stable and predict behavior over time and across different contexts, and need more behavioral examples to make trait attributions (Boseovski & Lee, 2006; Heyman & Dweck, 1998; Rholes & Rubel, 1984). Thus, presenting more instances of grateful or remorseful displays (or the lack thereof) may help even younger children grasp the importance of these emotional displays.
In sum, we propose an early emergence and continuous developmental unfolding of children’s experience of and response to others’ prosocial emotions. Thus far, however, the research consists of separate studies carried out with children in one of the crucial age brackets: younger than 3 years of age, between 3 and 4, and older than 4. What is needed is an integrative research agenda that combines existing and new research paradigms to better understand the workings of prosocial emotions in early childhood.
Do Young Children Really Experience Gratitude and Guilt? The Importance of Converging Methods
We return now to an important issue alluded to earlier, namely, that developmental research so far has not directly tapped children’s experience of gratitude and guilt. There are several reasons for this. First, complex emotions such as gratitude and guilt are not associated with single, identifiable facial expressions. Rather, they are expressed multimodally, including facially, verbally, bodily, and especially through specific types of actions. For instance, gratitude is typically expressed through verbal expressions of appreciation and favors or gifts (Harris et al., 1987), and guilt through confessions, apologies, and repair (Keltner & Buswell, 1996; Zahn-Waxler & Kochanska, 1990). Moreover, toddlers and preschoolers have limited verbal capacities and rarely issue polite assertions such as “sorry” or “thank you” spontaneously. Such young children also struggle with introspection, thus making it difficult for them to label or describe what they feel in the way that adults can. It is thus challenging to say with certainty that young children do indeed experience these prosocial emotions.
An alternative approach, taken in several of the studies reviewed here, is to focus on observable, measurable behaviors (overt or implicit) and assess whether those are in line with the predicted functions of the emotions, such as enhanced reparative motivations in guilt-inducing situations. The more such predictions are supported, the surer we can be that we are tapping into the emotion in question. A shortcoming of this approach is that in their earliest forms, these emotions may not yet serve all the functions they serve in later development or adulthood. One example is the pay-it-forward function of gratitude. Among adults, gratitude towards one’s benefactor motivates not only direct reciprocity (paying back) but also upstream reciprocity (paying forward to new individuals; Bartlett & DeSteno, 2006; DeSteno, Bartlett, Baumann, Williams, & Dickens, 2010). The latter promotes prosocial interactions with novel individuals, thus increasing prosociality within the group (Nowak & Roch, 2007). Yet Hepach et al. (2019b) did not find evidence of an enhanced prosocial motivation towards a novel individual but rather a specifically enhanced reciprocal motivation towards the benefactor. Other work suggests that it is around 4 years of age that children begin to pay benefits forward (Leimgruber et al., 2014).
Is the conclusion, then, that what toddlers and young preschoolers experience cannot be nascent guilt or gratitude? We do not think so. Rather, we suggest that these emotions first emerge to serve their most essential prosocial functions (promoting direct reciprocity in the case of gratitude and repair in the case of guilt), and with development, their functions become more complex, nuanced, and culture-specific. These advances rely on a variety of cognitive and social factors such as theory of mind, executive functioning, and cultural norms. The earliest functions of prosocial emotions mold into the more sophisticated and flexible functions seen in older children and adults. However, an absence of these sophisticated and flexible functions in early development need not signal an absence of the underlying emotions. This distinction is critical if we are to take seriously the ontogenetic emergence and functions of these prosocial emotions.
These considerations open up a new avenue for future research, one that integrates the implicit measures established in early childhood into studies with both children and adults. Pupil dilation is one such measure. Unlike behavioral or verbal measures, which are often limited in the ages with which they can be used, pupil dilation can be measured and interpreted similarly across ages (Hepach & Westermann, 2016). Moreover, pupil dilation indexes prosocial motivation in children aged 18 months to 5.5 years (Hepach et al., 2019a). Another promising candidate measure is cardiovascular responding, which relates to children’s prosocial behavior (J. G. Miller, 2018). Children with higher respiratory sinus arrhythmia show more costly sharing behavior (J. G. Miller, Kahle, & Hastings, 2015), and children with increased heart rate or skin conductance show less helping behavior (Eisenberg, McCreath, & Ahn, 1988; P. A. Miller, Eisenberg, Fabes, & Shell, 1996). Crucial insights will come from implementing these psychophysiological measures in studies with older children and adults, who can also label and provide subjective reports about their emotional experiences and thereby help validate the measures. We may predict, for instance, that older children and adults who report more guilt about causing harm will show greater physiological arousal if they cannot actively provide help, compared with those who report less guilt.
Furthermore, the comprehensive study of prosocial emotions across development will require the incorporation of valence measures to support current measures. One possibility is to assess facial expressions and relate the (degree of the) expressions to children’s prosocial behavior and physiological arousal (see, e.g., Vaish et al., 2009; Zahn-Waxler et al., 1992). Though neither gratitude nor guilt have a specific facial expression, gratitude is considered a positive emotion and guilt a negative emotion. Assessing the valence of children’s facial expressions could thus help clarify the nature of their emotional experience. This approach is exemplified in the “warm glow” research, which shows that both adults and children show positive facial expressions after benefiting others (Aknin et al., 2018; Lennon & Eisenberg, 1987). A second possible measure is children’s posture. The most common way to measure posture is to apply coding criteria to video recordings (Heckhausen, 1988; Shiota, Campos, & Keltner, 2003; Tracy & Robins, 2004). More recently, research has assessed young children’s postural changes directly using depth sensor imaging technology to measure the location of skeletal joints, and revealed that children show an upright body posture—indicative of positive affect—after helping others (Hepach, Vaish, & Tomasello, 2017b). Thus, assessing facial expressions and posture has led to important insights about the emotional consequences of prosocial behavior. We expect these measures of affective valence, in conjunction with established behavioral and physiological measures, to lead to similarly important insights into the early experience and functions of prosocial emotions such as gratitude and guilt. More generally, using converging methods and triangulating across studies will be key to drawing more confident conclusions about how young children experience prosocial emotions and how those experiences change with age.
Placing Prosocial Emotions in Context
Expressions of gratitude and guilt arguably evolved to communicate one’s commitment to invest in and maintain cooperative relationships with others, and perhaps most importantly with our peers (Frank, 1988; Keltner, Kogan, Piff, & Saturn, 2014). Thus far, however, studies on children’s prosocial emotions have involved adult partners (or puppets controlled by adults). These studies have helped to establish the early emergence of prosocial emotions, but adult–child interactions are importantly different from peer interactions. Adults are by and large cooperative partners who scaffold, share with, comfort, and instrumentally help children (Bar-Tal, Raviv, & Goldberg, 1982; Hay, Castle, Davies, Demetriou, & Stimson, 1999). Peer interactions, in contrast, are often accompanied by negative affect such as frustration. Toddlers also find it challenging to coordinate their actions with each other towards mutually beneficial goals (Brownell, Ramani, & Zerwas, 2006). Early peer relationships may thus demand more regulation and maintenance work from the child than adult–child relationships. It is critical to examine whether and how prosocial emotions such as gratitude and guilt help regulate early peer interactions. The challenge, of course, is in measuring these complex emotions and their impact within young children’s peer interactions given their volatile, at times erratic, nature. A crucial first step is to establish experimental paradigms that elicit peer helping and collaboration, as some recent studies have done (Hepach, Kante, & Tomasello, 2017; Ulber, Hamann, & Tomasello, 2015). These scenarios may then be modified to create mild guilt- or gratitude-inducing situations.
It is also important to consider socialization factors in the development of prosocial emotions (see Brownell, 2016; Brownell, Svetlova, Anderson, Nichols, & Drummond, 2013). One key factor is parenting style. Some work on this exists for the development of guilt (but not, to our knowledge, of gratitude). Children of mothers who use more power-assertive discipline display less guilt after causing a mishap (Kochanska, Gross, Lin, & Nichols, 2002). More generally, inductive parenting (highlighting the victim’s perspective and the consequences of the child’s actions on others) is associated with increased empathic and prosocial responding than techniques involving power assertion (Hoffman, 2000). Parenting style also influences what parents expect of their children. Compared with authoritative and authoritarian parents, for example, permissive parents are less likely to prompt their children to apologize for transgressing (C. E. Smith et al., 2017). This variation could plausibly modify the degree to which children experience remorse or expect and value it in others. Beyond parents, children’s experiences with peers and siblings also impact their prosocial emotions and behavior, and the nature and extent of these impacts depend on children’s age, relationship quality, and wider family context (see Dahl, 2018; Hughes, McHarg, & White, 2018).
Further, most current empirical work has been carried out with Western children. To fully understand early prosocial emotions and their functions, however, we must study them across cultures and contexts. Nonverbal displays of emotions regulate social relations in every culture (Tracy, Randles, & Steckler, 2015). Indeed, some aspects of both negative and positive prosocial emotions are pan-primate. Across humans and other primates, for instance, lowered body posture signals subordination and raised body posture signals confidence and dominance, especially in the face of competition (Fessler, 2007; Keltner et al., 2014). On the other hand, a universal and deep-rooted capacity for prosocial emotions need not imply universal experiences or manifestations of those emotions. Prosocial emotions, like all social phenomena, are situated within cultural values and beliefs. One may predict that as children are socialized into the values, beliefs, and norms of their culture, their prosocial emotions and behavior will change to reflect this (see Hay & Cook, 2007; Trommsdorff, Friedlmeier, & Mayer, 2007; Warneken & Tomasello, 2009b). Accordingly, new research shows both cross-cultural similarities and differences in the expressions and functions of school-age children’s gratitude (Mendonça, Merçon-Vargas, Payir, & Tudge, 2018). More such research, including with younger children and converging measures, will provide invaluable insights into the development and functions of prosocial emotions.
Other Prosocial Emotions
Our focus on gratitude and guilt is based on the proposal that these emotions help us meet two essential requirements for upholding prosocial relationships: maintenance and repair of prosocial interactions. However, they are certainly not the only prosocial emotions. Indeed, all social emotions arguably evolved to help humans resolve social dilemmas in service of promoting social living (Fessler & Haley, 2003; Nesse, 1990). It is worthwhile to touch upon a few key examples in order to paint a broader picture.
Among the self-conscious emotions, guilt is one of a number of members. Others include embarrassment, or the abashment that follows transgressions of social conventions (Goffman, 1967; Keltner & Buswell, 1997); shame, which follows the failure to live up to expectations that define one’s ideal or core self (Tangney, 1992); and pride, generated by appraisals that one is socially valued or has produced a socially valued outcome (Tracy & Robins, 2004). By making us reflect on and evaluate our behaviors and selves according to social standards or values, self-conscious emotions inhibit behaviors that harm our social relations and motivate behaviors that promote our social success (Hoffman, 2000; Ross, 2017; Sznycer, 2019).
Embarrassment after violating social norms (e.g., forgetting the name of an acquaintance) motivates us to rectify our mistakes and be more vigilant of those norms in the future. Displays of embarrassment also appease bystanders: adults report high levels of affiliative emotions such as amusement and sympathy in response to others’ embarrassment, and help individuals who display embarrassment more than they help those who display no embarrassment (Keltner, Young, & Buswell, 1997; Semin & Manstead, 1981). Embarrassment thus remedies social transgressions and returns social interactions to smooth functioning. Yet we know little about the ontogenetic emergence of these functions of embarrassment. Around ages 2 to 3, children begin to show embarrassment about feeling conspicuous (Lewis, Stanger, Sullivan, & Barone, 1991), which aligns with the developmental emergence of nascent gratitude and guilt. However, we do not yet know whether experiencing embarrassment promotes children to remedy their norm violations and whether children value displays of embarrassment in others.
There is a similar dearth of research on the functions of shame. Unlike guilt, which focuses on one’s harmful actions, shame involves feeling that the whole self is a failure and thus leads one to avoid the victim and withdraw from social contact (Tangney & Dearing, 2002). As such, shame is considered an inhibitor of prosocial behavior. Yet newer work suggests that shame may predict reparative effort under some circumstances, such as when the perceived failure seems more reparable (Leach & Cidam, 2015). Children begin to show shame-like responses by age 2 (e.g., avoiding the victim, withdrawing; Barrett, Zahn-Waxler, & Cole, 1993; Drummond et al., 2017). Whether early shame also sometimes promotes repair remains unknown, however (though see Ross, 2017).
Finally, pride is a positive self-conscious emotion that encourages one to seek social proximity (Ross, 2017). Most research has examined pride in achievement contexts (e.g., success in difficult tasks), and shows that children display pride by age 3 (Alessandri & Lewis, 1996; Belsky, Domitrovich, & Crnic, 1997; Heckhausen, 1988) and recognize displays and elicitors of pride by preschool and early school age (Kornilaki & Chloverakis, 2004; Tracy, Robins, & Lagattuta, 2005). One recent study found that 3- to 4-year-olds showed more pride-like displays after spontaneous helping than after prompted helping (Ross, 2017). More research is needed to understand whether children’s prosocial pride motivates further prosocial behavior and what social functions pride displays serve.
Emotions that involve evaluating others’ actions (rather than one’s own) can also promote prosociality. In gratitude, we positively evaluate another’s actions towards oneself and seek to maintain that positive interaction. Interestingly, seeing prosocial actions towards others can also elicit a positive emotion: elevation. Elevation is a warm, uplifting feeling in response to seeing unexpected acts of kindness, and it motivates us to initiate new prosocial interactions (Algoe & Haidt, 2009; Schnall, Roper, & Fessler, 2010). Work on elevation in young children has just begun but promises to shed new light on early prosocial mechanisms (Hepach & Tomasello, 2019).
The prosocial emotions described here are complex as they require advanced social-cognitive capacities, including subscribing to certain social standards or norms and evaluating ourselves and others according to those standards or norms. This “normative turn” occurs between 2 and 3 years of age (Schmidt, Rakoczy, & Tomasello, 2019; Tomasello, 2018). It is thus around these ages that we generally see the first evidence of these complex prosocial emotions. Prior to this, more basic emotional processes such as sympathy, enjoyment, and “warm glow” are already at work promoting prosocial behavior from infancy on. Children thus engage prosocially from very early in development. With the emergence of complex prosocial emotions such as gratitude and guilt, they can not only engage prosocially but actively maintain their valuable prosocial relationships, particularly with nonkin. This allows them to reap the benefits of prosociality and safeguard against its vulnerabilities in large groups.
In sum, emotions are an indisputable part of human prosociality, accompanying prosocial behavior from its earliest ontogenetic beginnings. We hope we have demonstrated here both the importance and the feasibility of a developmental perspective to understanding prosocial emotions, and that this review will vitalize research into not only gratitude and guilt but into all of the rich emotions that support human prosociality.
Footnotes
Acknowledgements
We are grateful to Tobias Grossmann for helpful discussions.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Amrisha Vaish was partially supported by a grant from the John Templeton Foundation (Grant No. 55437) as well as by the Society for Research in Child Development’s Scholar-in-Residence program.
