Journal of Experimental Child Psychology 167 (2018) 78–92

Contents lists available at ScienceDirect

Journal of Experimental ChildPsychology

journal homepage: www.elsevier .com/locate/ jecp

The development of intention-based sociomoraljudgment and distribution behavior from a third-party stance

https://doi.org/10.1016/j.jecp.2017.09.0210022-0965/� 2017 Elsevier Inc. All rights reserved.

⇑ Corresponding author at: Key Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy ofBeijing 100101, China.

E-mail address: lij@psych.ac.cn (J. Li).

Jing Li a,b,c,⇑, Michael Tomasello c

aKey Laboratory of Behavioral Science, Institute of Psychology, Chinese Academy of Sciences, Beijing 100101, ChinabDepartment of Psychology, University of Chinese Academy of Sciences, Beijing 100049, ChinacMax Planck Institute for Evolutionary Anthropology, 04103 Leipzig, Germany

a r t i c l e i n f o a b s t r a c t

Article history:Received 3 August 2017Revised 25 September 2017

Keywords:Moral judgmentSocial preferenceDistributionIntentionIndirect reciprocityThird party

The current study investigated children’s intention-based socio-moral judgments and distribution behavior from a third-partystance. An actor puppet showed either positive or negative inten-tion toward a target puppet, which had previously performed aprosocial or antisocial action toward others (i.e., children wit-nessed various types of indirect reciprocity). Children (3- and 5-year-olds) were asked to make sociomoral judgments and to dis-tribute resources to the actor puppet. Results showed that 5-year-olds were more likely than 3-year-olds to be influenced byintention when they made their judgment and distributedresources. The target’s previous actions affected only 5-year-olds’intent-based social preference. These results suggest that chil-dren’s judgments about intent-based indirect reciprocity developfrom ages 3 to 5 years.

� 2017 Elsevier Inc. All rights reserved.

Introduction

The mechanism of indirect reciprocity and the ability to evaluate social behavior toward third par-ties are assumed to be the evolutionary and ontogenetic bases of human morality and the cooperative

Sciences,

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 79

networks that characterize human societies (Fehr & Fischbacher, 2004; Fry, 2006; Gintis, Henrich,Bowles, Boyd, & Fehr, 2008; Henrich et al., 2005; Nowak & Sigmund, 2005). Moral thinking is arguedto be the basis of children’s interpretations of social interactions (Wainryb, Brehl, Matwin, Sokol, &Hammond, 2005). Indirect reciprocity of moral acts means acting positively toward those who areprosocial to third parties and/or negatively toward those who are antisocial to third parties (Nowak& Sigmund, 2005). Young children may engage in indirect reciprocation of moral acts; for example,3.5- to 4.5-year-olds distribute more resources to a puppet that previously acted prosocially towardothers than to a puppet that previously acted antisocially toward others (Kenward & Dahl, 2011;Olson & Spelke, 2008), 5-year-olds punish antisocial adults in an anonymous context (Kenward &Östh, 2015), and 6-year-olds even inflict costly third-party punishment toward unfair proposals(McAuliffe, Jordan, & Warneken, 2015). It appears that children’s own moral and prosocial actionsare affected by the recipient’s previous (moral or immoral) behavior toward third parties.

People interact positively with helpful individuals and punish or interact negatively with harmfulindividuals (Fehr & Gachter, 2002). Selectively prosocial behavior based on the content of previousinteractions is regarded as highly beneficial for humans and for human societies due to its effective-ness in maintaining cooperation between unrelated individuals (Dunfield & Kuhlmeier, 2010; Fehr &Fischbacher, 2004; Trivers, 1971). People are motivated to punish individuals who have been harmfulin the past (Henrich, 2006). However, research on adults has found that, under some circumstances,harmful acts and antisocial individuals have not been punished but have instead been positively eval-uated. Such positivity toward harmful individuals might be because this harmful behavior may besocially evaluated positively if it is directed toward those who are themselves wrongdoers (Hamlin& Wynn, 2011).

Social evaluation in humans often goes beyond analyzing the immediate and local valence of abehavior directed toward another, whereas previous evaluations of a target, which is the globalvalence of the action, also need to be considered (as illustrated by the common phase ‘‘the enemyof my enemy is my friend”) (Aronson & Cope, 1968; Hamlin, Wynn, Bloom, & Mahajan, 2011;Heider, 1958). Even infants (8 months old) can understand that the meaning of an act is influencednot solely by its own value but also by that of its target (Hamlin et al., 2011). Hamlin et al. (2011)found that 8-month-olds selectively preferred individuals who had acted positively toward a prosocialperson and who had acted negatively toward an antisocial person. Infants (10 months old) lookshorter when the fair distributor is harmed by a third party’s antisocial actions compared with whenthe unfair distributor is harmed, suggesting that they are able to morally distinguish individualsaccording to their distribution performance (Meristo & Surian, 2013; Meristo & Surian, 2014). Dahl,Schuck, and Campos (2013) found that 17- and 22-month-old infants equally helped actors whohad previously acted prosocially or antisocially, whereas 24-month-old toddlers preferred to helpthe prosocial actor. These studies mainly focused on infants’ social preference and social behaviorbased on their understanding of actions.

However, as children grow up, they make mature moral judgments based not only on the outcomeof actions but also on the underlying intention (Cushman, 2008; Karniol, 1978; Piaget, 1932/1997;Turiel, 2006). Although the role of outcome in the identification of good social partners is mainlyemphasized in traditional models of reciprocity (Hamilton, 1964), others have argued that reciprocityis also influenced by the intention behind an individual’s previous behaviors (Dunfield & Kuhlmeier,2010; Falk & Fischbacher, 2006; McCabe, Rigdon, & Smith, 2003). It is important for individuals to pre-dict others’ future behaviors by tracking others’ valenced mental states. It seems reasonable to believethat individuals who currently display negative intention are more likely to lead to negative outcomesin the future than those who currently demonstrate positive intention (Hamlin, 2013).

It is argued that intent-based judgments first emerge in children’s evaluation of naughtiness andsubsequent judgments of deserved punishment (Cushman, Sheketoff, Wharton, & Carey, 2013). Inmaking moral judgments, 3-year-olds are sensitive to the intention behind an action when stimuliare carefully controlled to remove confounding factors and when the intentional information is expli-cit and salient (Armsby, 1971; Nelson, 1980; Nobes, Panagiotaki, & Pawson, 2009; Siegal & Peterson,1998; Yuill & Perner, 1988). A rephrased question is also beneficial by making 4- to 8-year-olds moresensitive to intention (Nobes et al., 2009). It is argued that 3.5-year-olds make moral judgmentsaccording to any negative aspects (either negative intention or negative outcome), whereas in the neg-

80 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

ative intention/positive outcome condition 4.5-year-olds make intention-based judgments in storieswhere damage is directed against a person (Moran & O’brien, 1983). Vaish, Carpenter, andTomasello (2010) found that 3-year-olds not only engaged in less helping behavior toward a harmfultarget than neutral and helpful targets but also engaged in less helping behavior toward a target whohad intended but failed to harm others than toward a neutral target. This suggested that not onlyactions but also the intention behind those actions may affect an individual’s prosocial behavior.

Moreover, intention-based moral judgment continues to develop between 3 and 11 years of age(Costanzo, Coie, Grumet, & Farnill, 1973; Cushman et al., 2013; King, 1971; Margoni & Surian,2017). It is argued that, when making moral judgments of others’ actions, children rely more heavilyon intent than on outcome as they get older; specifically, compared with younger children, older chil-dren are less likely to condemn accidental harm and are more likely to condemn attempted harmdespite a benign outcome (Heiphetz & Young, 2014). As stated before, even infants are able to considerthe target’s previous moral action when they make social evaluations (Dahl et al., 2013; Hamlin et al.,2011; Meristo & Surian, 2013). We were interested in whether children are able to consider the tar-get’s previous moral action when they make intention-based moral judgments and perform moralbehavior as they get older.

Children’s behaviors show an agent-neutral application of moral norms through the third-partyparadigm (Vaish et al., 2010). Children as young as 3 years are able to evaluate social acts against thirdparties by verbally identifying transgressions against social norms (Catron & Masters, 1993; Darley &Shultz, 1990; Ingram & Bering, 2010; Kenward & Dahl, 2011; Sanderson & Siegal, 1988). The capacityof intention understanding based on a third-party perspective might also be expressed in terms ofheuristics of the form ‘‘If X has displayed a positive intention toward Z in the past, approach X” and‘‘If Y has displayed a negative intention toward Z in the past, avoid Y” (Hamlin et al., 2011). In addition,observation of others’ social interaction helps individuals to escape future harm by systematicallyavoiding those who have demonstrated malevolent mental states and to gain future benefit by sys-tematically approaching those who have demonstrated benevolent mental states (Hamlin, 2013). Itis helpful for individuals to decide who to trust and who to avoid in future social interactions (Bull& Rice, 1991; Cushman, 2015; Noë, 1990).

In this study, these two lines of research—indirect reciprocity and intention-based judgment—werebrought together to assess when and how intention behind acts from a third-party stance influenceschildren’s moral judgment and how intention influences children’s own complex social behavior, suchas social preference and distributive behavior, toward the actors. It is argued that the complexity ofmotivational states leads to discrepancies between everyday experience and psychological theorizing(i.e., the cognition–behavior gap) (Kuhl, 1985). For example, children do not always act fairly to otherseven if they understand principles of fairness by 3 years of age (Blake, McAuliffe, & Warneken, 2014).The reasoned action theory holds that attitude toward the behavior and subjective norms plays animportant role in the cognition–behavior gap (Ajzen, 1985). Thus, both sociomoral judgments (herereferring to social preference and moral judgment) and moral behavior (here referring to distributionbehavior) were simultaneously considered in our study.

At least two studies have demonstrated a high positive correlation between intent-based moraljudgment and developmental changes in the theory of mind (Chandler, Sokol, & Hallett, 2001;Killen, Mulvey, Richardson, Jampol, & Woodward, 2011). Moreover, children are able to adapt theirunderstanding of theory of mind to interpret more complex moral situations as they grow older(Smetana, Jambon, Conry-Murray, & Sturge-Apple, 2012). It has been argued that intent-based moraljudgment cannot be generated until children are able to represent others’ mental states in sufficientlyrich detail (Cushman et al., 2013). Indirect reciprocity also seems to require the ability of ‘‘theory ofmind” (Whiten & Byrne, 1997). In view of this evidence, 3- and 5-year-olds were recruited for the cur-rent study. The study applied a ‘‘show-not-tell” approach, in which children watched videos of theinteraction between the agent and target and drew their own conclusions.

The current study aimed to investigate the development of intention-based sociomoral judgmentsand distribution behavior from a third-party stance in preschoolers. The actor interacted positivelyand negatively, based on both intention and outcome, with the target who previously performedeither antisocially or prosocially to others. Children were asked to make sociomoral judgments towardthe actor and to distribute resources between the actor and a bystander. Based on the empirical evi-

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 81

dence suggesting that intention-based moral judgment continues to develop between 3 and 11 yearsof age (Costanzo et al., 1973; Cushman et al., 2013; King, 1971; Margoni & Surian, 2017; Nobes et al.,2009), it was predicted that, compared with 3-year-olds, 5-year-olds would rely more on intentionthan on outcome when making sociomoral judgments and distributing resources. In addition, basedon the evidence that 4.5-year-olds, but not 3-year-olds, are much more likely to act positively towardthe helper when the resources are scarce (Kenward & Dahl, 2011) and 5-year-olds are able to sponta-neously engage in third-party punishment toward antisocial adults (Kenward & Östh, 2015), wehypothesized that, compared with 3-year-olds, 5-year-olds would be more likely to focus on anagent’s intention and to be affected by the third party’s previous action when making sociomoral judg-ments and distributing resources.

Method

Participants

Participants were 50 3-year-old children (25 girls; Mage = 3.57 years, SD = 0.16, range = 3.24–3.80)and 51 5-year-old children (25 girls; Mage = 5.52 years, SD = 0.16; range = 5.24–5.76). Of these partic-ipants, 25 3-year-olds and 24 5-year-olds were randomly assigned to the prosocial context condition,and 25 3-year-olds and 27 5-year-olds were assigned to the antisocial context condition. Childrenwere recruited from kindergartens in a medium-sized German city and were native German speakers.All children were tested by the same female experimenter.

Procedure and design

All children were tested in a quiet room in their kindergartens. They were familiar with the place.Thus, it was easy to warm up, especially for the younger children. The procedure is shown in Fig. 1.

During the warm-up phase, two practice videos were shown to teach children how to use the moralscale. In one of the practice videos, an elephant puppet hit another elephant puppet. In the other one, apenguin puppet shared flowers with another penguin puppet. The children were first introduced toone of the practice videos in order to attract their attention, and then the moral scale was shownand explained to the children. The moral scale showed a very sad face at one end of a line (definedas ‘‘completely wrong”), a very happy face at the other end (defined as ‘‘completely right”), a neutralface midway (defined as ‘‘I do not know or I am not sure”), and a sad face and a happy face betweenthe midpoint and extremes (defined as ‘‘a little bit wrong” and ‘‘a little bit right,” respectively). Thechildren were told that the sad faces meant that the protagonist did something wrong, and the happyfaces meant that the protagonist had acted rightly. Moreover, they were told that the degree of sad-ness and happiness referred to the degree of wrongness/rightness. Children were told that the veryhappy (sad) face meant that the agent was completely right (wrong) and that the moderately happy(sad) face meant that the agent was moderately right (wrong). Then the experimenter randomlypointed to the faces and asked children to answer what the faces meant. If children were not ableto answer them correctly, the experimenter would give examples involving the children themselves,including toys being snatched by friends as an example of doing wrong and helping friends as anexample of doing right. After the children had learned the moral scale, the practice video was playedagain and they were asked to make a moral judgment about the puppet by using the moral scale. Thenthe other practice video was played, and the children were also asked to make a moral judgment aboutthe puppet by using the moral scale. Only children who understood and readily used the scale movedforward to the formal test phase.

In the formal test phase, the children were first asked to watch one of the videos about moral sto-ries in the target puppet (second party). The target puppet performed either prosocial or antisocialbehavior (more details can be found in the online supplementary material videos). In the prosocialcontext condition, the target puppet shared candy with the other puppet. In the antisocial contextcondition, the target puppet hit the other puppet to get a toy. Then four scenarios of interactionbetween actor puppets (first party) and the target puppet were shown to children (more details can

Fig. 1. The procedure of the experiment. Children first watched the video showing the target puppet acting antisocially orprosocially. The target puppet is the monkey puppet, which hit the cat puppet for snatching a toy or shared food with the catpuppet. Then children watched four videos showing the interactions between the actor puppet and the target puppet. Therewere two cow puppets (cow with gloves and cow without gloves). One of the cow puppets acted as the actor puppet and theother cow puppet acted as the bystander puppet. The actor puppet interacted with the target puppet as a successful helper, afailed helper, a successful hinderer, and a failed hinderer, and the bystander puppet just watched the interactions. Afterwatching each video, children were asked to make judgments and distribute resources to the actor puppet.

82 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

be found in the online supplementary material videos). Four different combinations of intention andoutcome were shown in these four conditions (Hamlin, 2013). First was successful helper, where theactor’s behavior toward the target had a positive intention and a positive outcome. The target wantedto open a box of toys. The actor tried three times to help the target open the box, and the box wasopened after the third attempt. The second combination was failed helper, where the actor’s behaviortoward the target had a positive intention and a negative outcome. The target wanted to open a boxwith toys. The actor tried three times to help the target open the box but failed. The third combinationwas successful hinderer, where the actor’s behavior toward the target had a negative intention and anegative outcome. The target wanted to open a box with toys. The actor tried three times to preventthe target from opening the box, and at the end the box was still closed. The fourth combination wasfailed hinderer, where the actor’s behavior toward the target had a negative intention and a positiveoutcome. The target wanted to open a box with toys. The actor tried three times to prevent the targetfrom opening the box, but the box was opened after the third attempt. In these four conditions, abystander puppet was placed to the side and just watched the interactions between the actor puppetand the target puppet. After each condition, children were asked three questions. First, ‘‘Do you thinkthe actor puppet acted rightly or wrongly? Can you show me on the scale?” (moral judgment). Chil-dren were asked to show their judgment on the moral scale. Second, ‘‘Who do you like more [actorpuppet or bystander puppet]?” (social preference). Children were asked to point their finger to thepuppet they prefer more. Third, ‘‘How many stickers would you like to give to the actor puppet andbystander puppet?” (distribution behavior). Five stickers were given to children, and they were askedto allocate these stickers to the actor puppet and bystander puppet. All five stickers needed to be dis-tributed. Children were asked to put the corresponding stickers in front of the two puppets.

A 2 (Age: 3 years old or 5 years old) ⁄ 4 (Condition—type of actor: successful helper, failed helper,successful hinderer, or failed hinderer) ⁄ 2 (Social Context—type of target: prosocial or antisocial)design was conducted, in which condition was a within-participant variable and social context wasa between-participant variable. Different intentions and outcomes were involved in these four condi-tions, including positive intention (successful helper and failed helper), negative intention (successfulhinderer and failed hinderer), positive outcome (successful helper and failed hinderer), and negativeoutcome (failed helper and successful hinderer).

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 83

A puppet show was used to perform the whole story because even nonviolent antisocial interac-tions between adult strangers might have shocked many children into inactivity (Vaish, Carpenter,& Tomasello, 2009). Interactions between puppets were recorded as short movies. The puppetrywas of a dramatic standard. Drama can arouse emotional responses in children and adults comparableto those triggered by similar real-life events regardless of whether the events are perceived as real ornot (Pouliot & Cowen, 2007). Moreover, short movies with puppets can also make the process of inter-action more unified than live puppet performances. Rather than explicitly informing children aboutthe status of the potential actors and recipients, a ‘‘show-not-tell” approach (Kenward & Dahl,2011), in which children witness the behavior of potential actors and recipients and draw theirown conclusions, was applied in this study.

The procedure was counterbalanced, including (a) the sequence of the four conditions (there are 24kinds of sequence), (b) which cow puppet was the actor puppet (half of the conditions had the cowpuppet with gloves as the actor puppet, and half had the cow puppet without gloves as the actor pup-pet), (c) the position of the actor puppet in the video (right/left of the target puppet; there are fourversions of the video in each condition), and (d) the position of the actor puppet and bystander puppetwhen children were asked to choose (the target puppet is always in the middle of the actor puppet andthe bystander puppet, and there are four versions of the video in each condition). Children werealways asked to make a moral judgment toward the actor puppet first. The social evaluation (prefer-ence judgment) and resource distribution between the actor puppet and bystander puppet were bal-anced randomly. The study has been approved by the Max Planck Institute for EvolutionaryAnthropology Child Subjects Committee. The study was carried out with the written informed consentof the participants’ parents and in accordance with all applicable laws and rules governing psycholog-ical research in Germany.

Results

Children’s moral judgment in helping and harming conditions during the practice phase was ana-lyzed. All children recognized that helping others was right and harming others was wrong (Mharming =�1.77, SD = 0.42; Mhelping = 1.94, SD = 0.24), F(1, 100) = 5355.08, p = .000, gp2 = .98. There was a signif-icant difference in moral scale of the practice phase between 3-year-olds and 5-year-olds (M3-year-olds

= 0.13, SD = 0.03;M5-year-olds = 0.04, SD = 0.03), F(1, 100) = 4.76, p = .031, gp2 = .98. However, the interac-tion between age and condition was also significant, F(1, 100) = 11.39, p = .001, gp2 = .10. Post hoc sim-ple effect analysis showed that children displayed significant differences with regard to the moralscale between 3-year-olds and 5-year-olds in the harming condition of the practice phase, F(1, 99)= 10.70, p = .001, gp2 = .10, but not in the helping condition of the practice phase, F(1, 98) = 2.95, p =.089, gp2 = .03.

Children’s moral judgment in different conditions and social contexts

Children’s performance in the moral scale was measured in order to examine their moral judgmentabout right and wrong (shown in Fig. 2).

The mixed designs of 2 (Age: 3 years old or 5 years old) ⁄ 2 (Intention: positive or negative) ⁄ 2(Social Context: prosocial or antisocial) and 2 (Age: 3 years old or 5 years old) ⁄ 2 (Outcome: positiveor negative) ⁄ 2 (Social Context: prosocial or antisocial) were analyzed by the repeated-measures anal-ysis of variance (ANOVA).

The main effect of intention was found to be significant, and post hoc multiple comparisonsshowed that children were significantly more likely to approve of the actor puppet’s action in the con-dition of positive intention than in the condition of negative intention, F(1, 100) = 68.39, p = .000, gp2 =.41. The interaction between intention and age was significant, F(1, 100) = 38.46, p = .000, gp2 = .28, andpost hoc simple effect analysis showed that 5-year-olds were significantly more likely to approve ofthe actor puppet’s action in the condition of positive intention than in the condition of negative inten-tion, F(1, 100) = 105.57, p = .000, gp2 = .52, whereas 3-year-olds performed similarly in these two con-ditions, F(1, 100) = 2.12, p = .149, gp2 = .02. Compared with 3-year-olds, 5-year-olds were significantly

Fig. 2. Children’s performance with the moral scale for different conditions and contexts. ‘‘Prosocial” refers to the social contextof the prosocial target, and ‘‘Antisocial” refers to the social context of the antisocial target. The error bars refer to the standarderrors.

84 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

more likely to approve of the actor puppet’s action in the condition of positive intention, F(1, 100) =11.96, p = .001, gp2 = .11, and to disapprove of the actor puppet’s action in the condition of negativeintention, F(1, 100) = 25.65, p = .000, gp2 = .21. The main effect of social context was not significant, nei-ther was the main effect of age, and nor were other interactions.

The main effect of outcome was also found to be significant, and post hoc multiple comparisonsshowed that children were significantly more likely to approve of the actor puppet’s action in the con-dition of positive outcome than in the condition of negative outcome, F(1, 100) = 35.55, p = .000, gp2 =.27. The main effect of social context was not significant, neither was the main effect of age, and norwere other interactions.

Children’s preference for the actor puppet in different conditions and social contexts

Children’s preference for the actor puppet is shown in Fig. 3.A binary logistic regression was performed to examine the effect of age, social context, intention

condition, and outcome condition on children’s preference for the actor puppet. Dummy coding ofage, social context, intention condition, and outcome condition was done. Prosocial context, negativeintention, negative outcome, and 3 years old were coded as reference categories. The model (shown inTable 1) effectively explained children’s preference behavior, v2(12) = 35.16, p = .000. The interactionbetween age and intention was found to be significant. Further analyses showed that, compared with3-year-olds, 5-year-olds were more likely to prefer the actor puppet in the condition of positive inten-tion, Mann–Whitney U = 852.50, Z = 2.77, p = .006, and less likely to prefer the actor puppet in the con-dition of negative intention, Mann–Whitney U = 831.50, Z = �3.21, p = .001. The 5-year-olds weremore likely to prefer the actor puppet in the condition of positive intention, v2(2) = 6.28, p = .043,and less likely to prefer the actor puppet in the condition of negative intention, v2(2) = 11.41, p =.003, whereas the 3-year-olds demonstrated similar preferences for the actor puppet in both of theseconditions [positive intention: v2(2) = 5.43, p = .066; negative intention: v2(2) = 1.96, p = .375].

The interaction between age and social context was also found to be significant. Further analysesshowed that (a) in the antisocial context, 3-year-olds were less likely than 5-year-olds to prefer theactor puppet in the condition of positive intention, Mann–Whitney U = 200.00, Z = 2.31, p = .021,but not in the conditions of negative intention, positive outcome, and negative outcome; and (b) in

Fig. 3. Children’s preference for the actor puppet in different conditions and contexts. ‘‘Prosocial” refers to the social context ofthe prosocial target, and ‘‘Antisocial” refers to the social context of the antisocial target. The error bars refer to the standarderrors.

Table 1Regression effect of intention condition, outcome condition, and social context on children’s social preference.

Variable B Exp(B) v2 (Wald) p

Age �1.65 0.19 9.55 .002Social Context �0.55 0.58 1.20 .273Intention �0.66 0.52 2.58 .108Age * Social Context 1.47 4.34 3.88 .049Age * Intention 2.25 9.47 13.89 .000Social Context * Intention �0.06 0.94 0.01 .922Outcome �0.17 0.85 0.17 .683Age * Outcome 0.17 1.18 0.08 .782Social Context * Outcome 0.11 1.12 0.04 .848Age * Social Context * Intention �0.82 0.44 0.80 .372Age * Social Context * Outcome �1.29 0.28 1.86 .173Age * Social Context * Intention * Outcome 0.85 2.35 1.05 .305

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 85

the prosocial context, 5-year-olds were less likely than 3-year-olds to prefer the actor puppet in thecondition of negative intention, Mann–Whitney U = 162.00, Z = �2.96, p = .003, but not in the condi-tions of positive intention, positive outcome, and negative outcome. The 5-year-olds were less likelyto prefer the actor puppet in the condition of negative intention in the prosocial context, v2(2) = 12.25,p = .002, but demonstrated a similar preference for the actor puppet in the condition of positive inten-tion in the prosocial context, v2(2) = 3.25, p = .197, and in the antisocial context [positive intention:v2(2) = 3.31, p = .191; negative intention: v2(2) = 4.22, p = .121], whereas 3-year-olds demonstrateda similar preference for the actor puppet both in the prosocial context [positive intention: v2(2) =5.84, p = .054; negative intention: v2(2) = 2.96, p = .228] and in the antisocial context [positive inten-tion: v2(2) = 4.00, p = .135; negative intention: v2(2) = 0.08, p = .961].

86 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

Children’s distribution behavior toward the actor puppet in different conditions and social contexts

We examined whether young children distributed resources according to the moral valence of thesocial actions previously observed among the recipients. The difference between stickers for the actorpuppet and for the bystander puppet referred to the number of stickers for the actor puppet minus thenumber of stickers for the bystander puppet. Children’s distribution behavior is shown in Fig. 4.

The mixed designs of 2 (Age: 3 years old or 5 years old) ⁄ 2 (Intention: positive or negative) ⁄ 2(Social Context: prosocial or antisocial) and 2 (Age: 3 years old or 5 years old) ⁄ 2 (Outcome: positiveor negative) ⁄ 2 (Social Context: prosocial or antisocial) were analyzed by the repeated-measuresANOVA.

The main effect of intention was found to be significant, and post hoc multiple comparison showedthat children significantly distributed more resources to the actor puppet in the condition of positiveintention than in the condition of negative intention, F(1, 100) = 8.22, p = .005, gp2 = .08. The interactionbetween intention and age was significant, F(1, 100) = 12.22, p = .001, gp2 = .11, and post hoc simpleeffect analysis showed that 5-year-olds significantly distributed more resources to the actor puppetin the condition of positive intention than in the condition of negative intention, F(1, 100) = 20.41,p = .000, gp2 = .17, whereas 3-year-olds performed similarly in these two conditions, F(1, 100) = 0.20,p = .658, gp2 = .002. Compared with 3-year-olds, 5-year-olds distributed significantly less resourcesto the actor puppet in the condition of negative intention, F(1, 100) = 14.67, p = .000, gp2 = .13, whereas5-year-olds performed similarly to 3-year-olds in the condition of positive intention, F(1, 100) = 3.48,p = .065, gp2 = .04. The main effect of social context was not significant, neither was the main effect ofage, and nor were other interactions.

The main effect of outcome was also found to be significant, and post hoc multiple comparisonshowed that children significantly distributed more resources to the actor puppet in the conditionof negative outcome than in the condition of positive outcome, F(1, 100) = 5.89, p = .017, gp2 = .06. Fur-ther analysis found that the effect showed only in 5-year-olds (Mpositive outcome = �0.62, SE = 0.22;Mneg-

ative outcome = 0.01, SE = 0.22), F(1, 97) = 4.46, p = .037, gp2 = .04. This might be due to the fact that 5-year-olds allocated few stickers to the actor puppet in the condition of successful helper in the antisocialcontext. The main effect of social context was not significant, neither was the main effect of age,and nor were other interactions.

Fig. 4. Difference between stickers for the actor puppet and the bystander puppet. The mean refers to the number of stickers forthe actor puppet minus the number of stickers for the bystander puppet. ‘‘Prosocial” refers to the social context of the prosocialtarget, and ‘‘Antisocial” refers to the social context of the antisocial target. The error bars refer to the standard errors.

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 87

Discussion

Young children’s ability to identify harmful intentions and behaviors and to analyze agents’ previ-ous behavior toward third parties might be an early step on the way to full-fledged social and moralevaluation (Vaish et al., 2010). This study investigated the development of intention-based sociomoraljudgments and distribution behavior from a third-party stance in 3- and 5-year-olds. It was found that(a) preschoolers’ sociomoral judgments and distribution behavior were substantially influenced byintention, and 5-year-olds relied more heavily on intention than 3-year-olds when making sociomoraljudgments and distributing resources to the actor puppet in the context of indirect reciprocity; and (b)5-year-olds were sensitive to social context when they indicated intent-based social preferences in thecontext of indirect reciprocity.

Children’s development of intention-based sociomoral judgments from a third-party stance

In this study, it was found that both intention and outcome were considered when children mademoral judgments. The actor puppet was judged as most wrong in the successful hinderer condition,followed by the failed hinderer, failed helper, and successful helper conditions. An individual’s actionin the conditions of negative intention and negative outcome was less likely to be approved bypreschoolers. The ability to distinguish good guys from bad guys has been assumed as the foundationof human morality and as being beneficial to group living (De Waal, 2009; Henrich & Henrich, 2007;Price, Cosmides, & Tooby, 2002).

Children made significantly different intent-based sociomoral judgments in the context of indirectreciprocity between 3 and 5 years of age. The 5-year-olds, but not the 3-year-olds, showed signifi-cantly intent-based moral judgment and social preferences for the actor puppet. Specifically, com-pared with 3-year-olds, 5-year-olds demonstrated a greater preference for the actor puppet in thesuccessful helper condition and less preference for the actor puppet in the failed hinderer condition.It has been argued that young children focus on outcome more than older children, that they relymostly on outcome (Helwig, Hildebrandt, & Turiel, 1995), or that they focus equally on intentionand outcome (Cushman et al., 2013). Moreover, intention-based moral judgment continues to developbetween 4 and 8 years of age (Cushman et al., 2013), and as they get older children display greaterreliance on intent than on outcome when making moral judgments about others’ actions (Heiphetz& Young, 2014). In the cases of failed attempts to hinder and accidental hindrance, punishment ofattempted but failed hindrance increases with age (Helwig et al., 1995), whereas punishment of acci-dental hindrance decreases with age (Cushman et al., 2013; Killen et al., 2011).

Following life history theory, the elicited intent-based moral reasoning emerges roughly at 5 or 6years of age because children at this age are increasingly involved in social interactions with peers(Margoni & Surian, 2016). To avoid potentially dangerous conflicts, the tendency to attend to agents’intentions rather than outcomes becomes crucial precisely at this age, when the focus of children’ssocial life begins to shift from interaction with their parents at home to interaction with peers inschool (Marlowe, 2005). The ability to understand and evaluate others’ intentions might be systemat-ically used by children in a variety of social interactions during the later preschool years (Margoni &Surian, 2016). Another alternative explanation is that, children are able to adapt their understandingof theory of mind to interpret more complex moral situations as they grow older (Smetana et al.,2012). Children who pass false belief theory of mind tasks are less likely to attribute negative intentionto the accidental harm than children who do not pass false belief tasks (Killen et al., 2011). Moreover,the ETIC (emotional arousal, theory of mind, and inhibitory control) model of morality argues that the-ory of mind abilities involve the integration of the agent’s harmful or innocent intentions into moraljudgment (Buon, Seara-Cardoso, & Viding, 2016). In addition, based on greater awareness of others’experiences in the context of social interactions, children develop empathic sensitivity during theirsecond and third years of life (Decety, 2010). The empathy network is involved in encoding outcomes(especially harmful outcomes), which makes children more outcome based (Patil, Calò, Fornasier,Cushman, & Silani, 2017). However, this reliance on outcomes declines with development of executive

88 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

function that helps children to regulate their emotional reaction and, as a result, increases the influ-ence of intentions on moral decisions (Buon et al., 2016).

Social preference was affected by the target’s previous action (prosocial or antisocial behaviortoward others) only in 5-year-olds. Regression analysis found that the interaction between age andsocial context could significantly predict children’s social preference for the actor puppet. For example,5-year-olds were less likely to prefer the actor puppet in the condition of negative intention in theprosocial context, whereas 3-year-olds demonstrated a similar preference for the actor puppet bothin the prosocial context and in the antisocial context. This suggests that the capacity for intent-based indirect reciprocity in social preference increased with age. Previous evidence shows thatyounger children are able to act accordingly based on the target’s previous actions. It is argued thatchildren older than 3.5 years are able to perform prosocial behavior according to the agents’ previous(moral or immoral) behavior toward third parties (Kenward & Dahl, 2011; Olson & Spelke, 2008).However, more complicated context was set in our study; specifically, the agent showed either posi-tive or negative intention toward a target puppet that had previously performed a prosocial or anti-social action toward others (i.e., children witnessed various types of indirect reciprocity). It isargued that criteria, including the victim’s and children’s relationship with the transgressor(Wellman, Larkey, & Somerville, 1979) and the context of the transgression (Killen, 2007; Smetana,2006) are involved in individuals’ considerations when they make social and moral evaluations. Thus,it is reasonable to infer that a higher level of cognitive load (e.g., a higher level of theory of mind) isrequired in our study so that children’s social preference was influenced by the target’s previousaction until 5 years of age.

In addition, the effect of social context was mixed in young children. For example, in the conditionof positive intention, 5-year-olds showed significantly greater preference for the actor puppet than 3-year-olds in the antisocial context but not in the prosocial context. This might be because second-order theory of mind applied when children made intent-based sociomoral judgments from thethird-party perspective in this study. Indirect reciprocity seems to require the ability of theory of mind(Whiten & Byrne, 1997). Judging moral actions between third parties as ‘‘good” or ‘‘bad,” distinct fromjudging an organism’s own moral experiences as ‘‘good” or ‘‘bad,” requires a higher degree of empathyand developed abilities of cognition (refers to sophisticated theory of mind) (Martin & Sigmund, 2005).Theory of mind may facilitate individuals in making moral judgments in the context of indirectreciprocity. In addition, a close relationship is evident between intent-based moral judgment anddevelopmental changes in theory of mind (Chandler et al., 2001; Cushman et al., 2013; Killen et al.,2011). Although nearly half of 4.7-year-olds and all 6.0-year-olds are able to attribute second-orderfalse beliefs (Sullivan, Zaitchik, & Tager-Flusberg, 1994), successfully passing first- or second-orderfalse belief tasks does not mean that individuals are able to deploy the ability of theory of mind.Empirical evidence shows that many players, even adults, are restricted to adopting first-order abilityto reason others’ actions, and nearly all players are restricted to making second-order reasoning(Colman, 2003; Hedden & Zhang, 2002; Verbrugge & Mol, 2008). Keysar, Lin, and Barr (2003) alsoargued that there was an obvious discrepancy between the ability to thoughtfully distinguish one’sown beliefs from others’ beliefs and effectively using this ability to interpret others’ action, suggestingthat even adults are not able to fully deploy the sophisticated theory of mind to interpret the actions ofothers.

Children’s development of intention-based distribution behavior from the third-party perspective

It was found that 5-year-olds, but not 3-year-olds, distributed significantly different amounts ofstickers to the actor puppet and the bystander puppet in four conditions. Specifically, 5-year-olds dis-tributed fewer stickers to the actor puppet than 3-year-olds in the negative intention conditions. Thissuggested that 5-year-olds’ distribution behavior from a third-party stance was more likely to beaffected by intention than that of 3-year-olds. The results were similar to previous findings. It has beenargued that most 4.5-year-olds already justify their unequal distributions based on the recipients’moral behavior (Kenward & Dahl, 2011). In addition, children of this age recognize that moral trans-gressors deserved punishment (Smetana, Schlagman, & Adams, 1993). These results suggest that 4.5-year-olds are able to understand that morally valenced behaviors should be indirectly reciprocated.

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 89

In addition, it was found that children distributed more resources to the actor puppet in the con-dition of negative outcome than in the condition of positive outcome, and further analysis showed thatthe effect showed only in 5-year-olds. This might be due to the fact that 5-year-olds allocated fewstickers to the actor puppet in the condition of successful helper in the antisocial context. Accordingto the ETIC model of morality, children’s reliance on outcomes declines with age and older childrenrely more on intentions when they make moral judgments and perform moral behavior such as dis-tributing resources for others in our study (Buon et al., 2016).

However, the target’s previous action, here referred to as the prosocial/antisocial context, did notaffect young children’s distribution behavior in this study. It is difficult for children younger than 6years to distribute resources based on explicit reasoning concerning moral principles of indirectreciprocity (Arsenio & Kramer, 1992; Kenward & Dahl, 2011). In addition, the effect of sociomoraljudgments on distribution behavior might be mediated by a more complicated process as opposedto simply liking the positive action or intention. Young children may begin to understand the rule thatmorally valenced behavior should be indirectly reciprocated (Kenward & Dahl, 2011). It might be dif-ficult for young children to apply such reasoned moral principles.

The main effect of social context was not significant overall in all three tasks, whereas the interac-tion between age and social context was found only in the social preference task but not in the moraljudgment and distribution tasks. This showed that young children already perform indirect reciprocityof intent-based behaviors when making social preference, and 5-year-olds’ social preference might bemore sensitive to social context. This is probably due to the fact that moral judgment and resourcedistribution tasks are elicited response tasks. Apart from representing an actor’s intention and therecipient’s outcome, elicited response intention–outcome tasks also require individuals to hold com-peting pieces of information in mind simultaneously, select one piece of information to rely on, andinhibit a presumably dominant response based on real-world outcomes; thus, it may be difficult foryoung children to perform better in elicited response tasks (Hamlin, 2013). For example, in this study,children needed to represent the agent’s intention and the target’s previous behavior simultaneouslywhen they made moral judgments and distributed resources. Comparatively, puppet choice in thesocial preference task might be relatively insensitive to the factors that hamper children’s perfor-mance in elicited response tasks. Moreover, puppet choice might help to trigger an implicit form ofevaluation (Hamlin, 2013).

Another alternative explanation for the above result is offered by social intuitionist models, whichclaim that even adults’ moral judgment can be influenced by intuitive processes without reasoningand that these models are supported by a nonreasoned mechanism for indirect reciprocity of moralbehaviors (Cushman, Young, & Hauser, 2006; Haidt, 2007; Krebs & Denton, 2005). According to socialintuitionist models for moral cognition, adults’ moral judgments are affected by intuitive processes(Cushman et al., 2006; Haidt, 2001; Krebs & Denton, 2005). Our results raise the possibility that youngchildren’s mechanism for indirect reciprocity, sociomoral judgments, and distribution behavior mightbe developmentally continuous with similar intuitive processes in adults.

It was found that young children were more likely to focus on negative intention and negative out-come, and the tendency increased with age. First, in the condition of negative intention, 5-year-oldsshowed significantly less preference for the actor puppet than 3-year-olds in a prosocial contextbut not in an antisocial context. Second, 5-year-olds distributed significantly fewer stickers to theactor puppet than 3-year-olds in the condition of negative intention but distributed similar amountsin the helper conditions. It has been shown that negative intentions and outcomes were very salient to3- and 5-year-olds. This is in line with other findings suggesting that negative events are more salientto young children than positive ones (Vaish, Grossmann, & Woodward, 2008). Children as young as 3years already perform less helping behavior toward individuals who harm or intend to harm a thirdparty, suggesting that the ability to recognize harmful or intentionally harmful individuals and with-hold help from them emerges early on during childhood (Vaish et al., 2010). Kenward and Dahl (2011)also argued that 3-year-olds are afraid of the hinderer, meaning that the negative actions are so strongthat a child’s memory of a hinderer’s identity might be weakened.

The findings that children, or at least 5-year-olds, are able to adjust their sociomoral judgments andcomplex social behavior toward individuals according to the moral valence of the individuals’ actionstoward third parties with different moral valences suggest that similar behavior might be observed in

90 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

preschoolers’ everyday social interactions with others. In other words, indirect reciprocity based on athird party’s previous behavior has emerged in preschoolers. Indirect reciprocity has been shown to beimportant and meaningful in promoting cooperative social interactions (Nowak & Sigmund, 2005).Thus, the findings observed here might be effective in reducing moral transgressions in preschoolers’social interactions.

There is one limitation that should be addressed. The actor helped the target to open the boxtogether in conditions of positive intention. Collaboration is involved and mixed with help behaviorin this condition. Thus, children’s positive evaluation of the actor as the helper might be partly dueto the actor taking part in cooperative actions. Future work should deploy help behavior with onlyindividual effort involved in order to discriminate help from collaboration.

Conclusion

This study examined the development of intention-based sociomoral judgments and distributionbehavior from the third-party stance in 3- and 5-year-old children. It showed that preschoolers’ moraljudgment and distribution behavior was substantially influenced by intention; moreover, comparedwith 3-year-olds, 5-year-olds relied more heavily on intention when making moral judgments anddistributing resources. These results suggest that young children’s sociomoral judgments and distribu-tion behavior are increasingly influenced by intention from 3 to 5 years of age. Although social context(here referring to the target’s previous action) did not affect young children’s intent-based moral judg-ment and distribution behavior, 5-year-olds already performed indirect reciprocity of intent-basedbehaviors when making social preference. These results suggest that children’s judgments aboutintent-based indirect reciprocity develop from 3 to 5 years of age.

Acknowledgments

This research was supported by the National Natural Science Foundation of China (31300859). Wethank the children who participated in the study; Rahel Doehring-Jahn, Polina Lepikhina, and RonnyBarr for making videos; and Anne Tomm for help with data collection.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, in the online version, at https://doi.org/10.1016/j.jecp.2017.09.021.

References

Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. New York: Springer Science & Business Media.Armsby, R. E. (1971). A reexamination of the development of moral judgments in children. Child Development, 42, 1241–1248.Aronson, E., & Cope, V. (1968). My enemy’s enemy is my friend. Journal of Personality and Social Psychology, 8, 8–12.Arsenio, W. F., & Kramer, R. (1992). Victimizers and their victims: Children’s conceptions of the mixed emotional consequences

of moral transgressions. Child Development, 63, 915–927.Blake, P. R., McAuliffe, K., & Warneken, F. (2014). The developmental origins of fairness: The knowledge–behavior gap. Trends in

Cognitive Sciences, 18, 559–561.Bull, J., & Rice, W. (1991). Distinguishing mechanisms for the evolution of co-operation. Journal of Theoretical Biology, 149, 63–74.Buon, M., Seara-Cardoso, A., & Viding, E. (2016). Why (and how) should we study the interplay between emotional arousal,

theory of mind, and inhibitory control to understand moral cognition? Psychonomic Bulletin & Review, 23, 1660–1680.Catron, T. F., & Masters, J. C. (1993). Mothers’ and children’s conceptualizations of corporal punishment. Child Development, 64,

1815–1828.Chandler, M. J., Sokol, B. W., & Hallett, D. (2001). Moral responsibility and the interpretive turn: Children’s changing conceptions

of truth and rightness. In B. F. Malle, L. J. Moses, & D. A. Baldwin (Eds.), Intentions and intentionality: Foundations of socialcognition (pp. 345–365). Cambridge, MA: MIT Press.

Colman, A. M. (2003). Depth of strategic reasoning in games. Trends in Cognitive Sciences, 7, 2–4.Costanzo, P. R., Coie, J. D., Grumet, J. F., & Farnill, D. (1973). A reexamination of the effects of intent and consequence on

children’s moral judgments. Child Development, 44, 154–161.Cushman, F. (2008). Crime and punishment: Distinguishing the roles of causal and intentional analyses in moral judgment.

Cognition, 108, 353–380.Cushman, F. (2015). Deconstructing intent to reconstruct morality. Current Opinion in Psychology, 6, 97–103.

J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92 91

Cushman, F., Sheketoff, R., Wharton, S., & Carey, S. (2013). The development of intent-based moral judgment. Cognition, 127,6–21.

Cushman, F., Young, L., & Hauser, M. (2006). The role of conscious reasoning and intuition in moral judgment testing threeprinciples of harm. Psychological Science, 17, 1082–1089.

Dahl, A., Schuck, R. K., & Campos, J. J. (2013). Do young toddlers act on their social preferences? Developmental Psychology, 49,1964–1970.

Darley, J. M., & Shultz, T. R. (1990). Moral rules: Their content and acquisition. Annual Review of Psychology, 41, 525–556.De Waal, F. (2009). Primates and philosophers: How morality evolved. Princeton, NJ: Princeton University Press.Decety, J. (2010). The neurodevelopment of empathy in humans. Developmental Neuroscience, 32, 257–267.Dunfield, K. A., & Kuhlmeier, V. A. (2010). Intention-mediated selective helping in infancy. Psychological Science, 21, 523–527.Falk, A., & Fischbacher, U. (2006). A theory of reciprocity. Games and Economic Behavior, 54, 293–315.Fehr, E., & Fischbacher, U. (2004). Social norms and human cooperation. Trends in Cognitive Sciences, 8, 185–190.Fehr, E., & Gachter, S. (2002). Altruistic punishment in humans. Nature, 415, 137–140.Fry, D. P. (2006). Reciprocity: The foundation stone of morality. In M. Killen & J. G. Smetana (Eds.), Handbook of moral

development (pp. 399–422). Mahwah, NJ: Lawrence Erlbaum.Gintis, H., Henrich, J., Bowles, S., Boyd, R., & Fehr, E. (2008). Strong reciprocity and the roots of human morality. Social Justice

Research, 21, 241–253.Haidt, J. (2001). The emotional dog and its rational tail: A social intuitionist approach to moral judgment. Psychological Review,

108, 814–834.Haidt, J. (2007). The new synthesis in moral psychology. Science, 316, 998–1002.Hamilton, W. (1964). The genetical evolution of social behaviour. Journal of Theoretical Biology, 7, 17–52.Hamlin, J. K. (2013). Failed attempts to help and harm: Intention versus outcome in preverbal infants’ social evaluations.

Cognition, 128, 451–474.Hamlin, J. K., & Wynn, K. (2011). Young infants prefer prosocial to antisocial others. Cognitive Development, 26, 30–39.Hamlin, J. K., Wynn, K., Bloom, P., & Mahajan, N. (2011). How infants and toddlers react to antisocial others. Proceedings of the

National Academy of Sciences of the United States of America, 108, 19931–19936.Hedden, T., & Zhang, J. (2002). What do you think I think you think?: Strategic reasoning in matrix games. Cognition, 85, 1–36.Heider, F. (1958). The psychology of interpersonal relations. Hillsdale, NJ: Lawrence Erlbaum.Heiphetz, L., & Young, L. (2014). A social cognitive developmental perspective on moral judgment. Behaviour, 151, 179–199.Helwig, C. C., Hildebrandt, C., & Turiel, E. (1995). Children’s judgments about psychological harm in social context. Child

Development, 66, 1680–1693.Henrich, J. (2006). Cooperation, punishment, and the evolution of human institutions. Science, 312, 60–61.Henrich, J., Boyd, R., Bowles, S., Camerer, C., Fehr, E., Gintis, H., ... Ensminger, J. (2005). ‘‘Economic man” in cross-cultural

perspective: Behavioral experiments in 15 small-scale societies. Behavioral and Brain Sciences, 28, 795–815.Henrich, J., & Henrich, N. (2007). Why humans cooperate: A cultural and evolutionary explanation. Oxford, UK: Oxford University

Press.Ingram, G. P., & Bering, J. M. (2010). Children’s tattling: The reporting of everyday norm violations in preschool settings. Child

Development, 81, 945–957.Karniol, R. (1978). Children’s use of intention cues in evaluating behavior. Psychological Bulletin, 85, 76–85.Kenward, B., & Dahl, M. (2011). Preschoolers distribute scarce resources according to the moral valence of recipients’ previous

actions. Developmental Psychology, 47, 1054–1064.Kenward, B., & Östh, T. (2015). Five-year-olds punish antisocial adults. Aggressive Behavior, 41, 413–420.Keysar, B., Lin, S., & Barr, D. J. (2003). Limits on theory of mind use in adults. Cognition, 89, 25–41.Killen, M. (2007). Children’s social and moral reasoning about exclusion. Current Directions in Psychological Science, 16, 32–36.Killen, M., Mulvey, K. L., Richardson, C., Jampol, N., & Woodward, A. (2011). The accidental transgressor: Morally-relevant theory

of mind. Cognition, 119, 197–215.King, M. (1971). The development of some intention concepts in young children. Child Development, 42, 1145–1152.Krebs, D. L., & Denton, K. (2005). Toward a more pragmatic approach to morality: A critical evaluation of Kohlberg’s model.

Psychological Review, 112, 629–649.Kuhl, J. (1985). Volitional mediators of cognition–behavior consistency: Self-regulatory processes and action versus state

orientation. In J. Kuhl & J. Beckmann (Eds.), Action control: From cognition to behavior (pp. 101–128). New York: SpringerScience & Business Media.

Margoni, F., & Surian, L. (2016). Explaining the U-shaped development of intent-based moral judgments. Frontiers in Psychology,7. https://doi.org/10.3389/fpsyg.2016.00219.

Margoni, F., & Surian, L. (2017). Children’s intention-based moral judgments of helping agents. Cognitive Development, 41,46–64.

Marlowe, F. W. (2005). Hunter–gatherers and human evolution. Evolutionary Anthropology: Issues, News, and Reviews, 14(2),54–67.

Martin, N., & Sigmund, K. (2005). Evolution of indirect reciprocity. Nature, 437, 1291–1298.McAuliffe, K., Jordan, J. J., & Warneken, F. (2015). Costly third-party punishment in young children. Cognition, 134, 1–10.McCabe, K. A., Rigdon, M. L., & Smith, V. L. (2003). Positive reciprocity and intentions in trust games. Journal of Economic Behavior

& Organization, 52, 267–275.Meristo, M., & Surian, L. (2013). Do infants detect indirect reciprocity? Cognition, 129, 102–113.Meristo, M., & Surian, L. (2014). Infants distinguish antisocial actions directed towards fair and unfair agents. PLoS ONE, 9(10),

e110553.Moran, J. D., III, & O’brien, G. (1983). The development of intention-based moral judgments in three- and four-year-old children.

Journal of Genetic Psychology, 143, 175–179.Nelson, S. A. (1980). Factors influencing young children’s use of motives and outcomes as moral criteria. Child Development, 51,

823–829.

92 J. Li, M. Tomasello / Journal of Experimental Child Psychology 167 (2018) 78–92

Nobes, G., Panagiotaki, G., & Pawson, C. (2009). The influence of negligence, intention, and outcome on children’s moraljudgments. Journal of Experimental Child Psychology, 104, 382–397.

Noë, R. (1990). A veto game played by baboons: A challenge to the use of the Prisoner’s Dilemma as a paradigm for reciprocityand cooperation. Animal Behaviour, 39, 78–90.

Nowak, M. A., & Sigmund, K. (2005). Evolution of indirect reciprocity. Nature, 437, 1291–1298.Olson, K. R., & Spelke, E. S. (2008). Foundations of cooperation in young children. Cognition, 108, 222–231.Patil, I., Calò, M., Fornasier, F., Cushman, F., & Silani, G. (2017). The behavioral and neural basis of empathic blame. Scientific

Reports, 7(1), 5200.Piaget, J. (1997). The moral judgement of the child. New York: Simon & Schuster (Original work published 1932).Pouliot, L., & Cowen, P. S. (2007). Does perceived realism really matter in media effects? Media Psychology, 9, 241–259.Price, M. E., Cosmides, L., & Tooby, J. (2002). Punitive sentiment as an anti-free rider psychological device. Evolution and Human

Behavior, 23, 203–231.Sanderson, J. A., & Siegal, M. (1988). Conceptions of moral and social rules in rejected and nonrejected preschoolers. Journal of

Clinical Child Psychology, 17, 66–72.Siegal, M., & Peterson, C. C. (1998). Preschoolers’ understanding of lies and innocent and negligent mistakes. Developmental

Psychology, 34, 332–341.Smetana, J. G., Jambon, M., Conry-Murray, C., & Sturge-Apple, M. L. (2012). Reciprocal associations between young children’s

developing moral judgments and theory of mind. Developmental Psychology, 48, 1144–1155.Smetana, J. G. (2006). Social–cognitive domain theory: Consistencies and variations in children’s moral and social judgments. In

M. Killen & J. Smetana (Eds.), Handbook of moral development (pp. 119–153). Mahwah, NJ: Lawrence Erlbaum.Smetana, J. G., Schlagman, N., & Adams, P. W. (1993). Preschool children’s judgments about hypothetical and actual

transgressions. Child Development, 64, 202–214.Sullivan, K., Zaitchik, D., & Tager-Flusberg, H. (1994). Preschoolers can attribute second-order beliefs. Developmental Psychology,

30, 395–402.Trivers, R. L. (1971). The evolution of reciprocal altruism. Quarterly Review of Biology, 46(1), 35–57.Turiel, E. (2006). Thought, emotions, and social interactional processes in moral development. In M. Killen & J. Smetana (Eds.),

Handbook of moral development (pp. 7–35). Mahwah, NJ: Lawrence Erlbaum.Vaish, A., Carpenter, M., & Tomasello, M. (2009). Sympathy through affective perspective taking and its relation to prosocial

behavior in toddlers. Developmental Psychology, 45, 534–543.Vaish, A., Carpenter, M., & Tomasello, M. (2010). Young children selectively avoid helping people with harmful intentions. Child

Development, 81, 1661–1669.Vaish, A., Grossmann, T., & Woodward, A. (2008). Not all emotions are created equal: The negativity bias in social–emotional

development. Psychological Bulletin, 134, 383–403.Verbrugge, R., & Mol, L. (2008). Learning to apply theory of mind. Journal of Logic, Language, and Information, 17, 489–511.Wainryb, C., Brehl, B. A., Matwin, S., Sokol, B. W., & Hammond, S. (2005). Being hurt and hurting others: Children’s narrative

accounts and moral judgments of their own interpersonal conflicts. Monographs of the Society for Research in ChildDevelopment, 70(3, Serial No. 281).

Wellman, H. M., Larkey, C., & Somerville, S. C. (1979). The early development of moral criteria. Child Development, 50, 869–873.Whiten, A., & Byrne, R. W. (1997). Machiavellian intelligence II: Extensions and evaluations. Cambridge, UK: Cambridge University

Press.Yuill, N., & Perner, J. (1988). Intentionality and knowledge in children’s judgments of actor’s responsibility and recipient’s

emotional reaction. Developmental Psychology, 24, 358–365.