Over the last decade, research suggests that infants have a rich understanding of others’ social lives, potentially representing a heterogeneous set of core social domains. Thomas presents a bold thesis that could account for many of these findings from a unitary core capacity: Infants are predisposed to track relational cues and infer connections between others. We see this as a potentially transformative hypothesis with even broader implications than Thomas considers. Specifically, we suggest that findings interpreted as evidence that infants expect and prefer fair resource distributions might instead reflect a sensitivity to allocations as indicating a relationship status between the distributor and the recipients. On this view, unequal distributions might signal a relationship: unequal distributors favor one recipient over the other, suggesting closeness, whereas equal distributors treat both recipients impartiality, which doesn’t imply information about relationships. Research shows that unequal distributions lead children to think that the distributor is friends with the advantaged recipient (Liberman & Shaw, Reference Liberman and Shaw2017). Although infants haven’t been tested on a similar paradigm, we argue that data on infants’ responses to distributional outcomes may likewise reflect relationship inferences rather than moral evaluations. This possibility challenges prevailing interpretations and highlights the need for further research on infants’ sensitivity to equality matching (EM) relationships.

Research on infants’ reactions to distributional outcomes largely falls into two categories: infants look longer at unequal distribution outcomes – interpreted as a violation of expected equality (Meristo, Strid, & Surian, Reference Meristo, Strid and Surian2016; Schmidt & Sommerville, Reference Schmidt and Sommerville2011; Sloane, Baillargeon, & Premack, Reference Sloane, Baillargeon and Premack2012; Sommerville et al., Reference Sommerville, Schmidt, Yun and Burns2013; Ziv & Sommerville, Reference Ziv and Sommerville2017) – and they show preferences for equal distributors over unequal distributors (Lucca, Pospisil, & Sommerville, Reference Lucca, Pospisil and Sommerville2018; Surian & Franchin, Reference Surian and Franchin2017) – choosing to engage with the former over the latter. This research has almost exclusively been interpreted as evidence for babies’ sociomoral responses. However, longer looking times may reflect interest in relational cues, not violations of expected equality. And preferences for equal distributors may stem from seeing them as possessing an open alliance – open to forming new allegiances (including with the infant) – while unequal distributors may be inferred to already have preexisting ties reflected in their partiality.

Our proposal that infants are focused on relational information in resource distributions rather than violations of equality aligns with a body of research – much of which Thomas incisively summarizes – that infants are deeply attuned to social and affiliative cues like imitation, similarity, synchrony, and proximity (Kudrnova, Spelke, & Thomas, Reference Kudrnova, Spelke and Thomas2024; Liberman, Kinzler, & Woodward, Reference Liberman, Kinzler and Woodward2014, Reference Liberman, Kinzler and Woodward2018, Reference Liberman, Kinzler and Woodward2021; Liberman, Woodward, & Kinzler, Reference Liberman, Woodward and Kinzler2017; Powell & Spelke, Reference Powell and Spelke2013, Reference Powell and Spelke2018). It also challenges the idea that infants’ responses to inequality indicate moral evaluations. In reconciling her relational account with claims about an “innate moral core,” Thomas speculates that instead of inferring moral traits, infants might favor agents who initiate or maintain prosocial relationships, for instance, reaching for “those who distribute resources equally, which could be seen as establishing an equality matching relationship” (pp. 26–27). However, she admits difficulty in explaining why without invoking trait inferences. Our relational interpretation sidesteps this issue: If infants interpret allocations relationally rather than morally, their responses might reflect reasoning about social connections rather than unfairness. Ziv, Whiteman, and Sommerville (Reference Ziv, Whiteman and Sommerville2021) is, to our knowledge, the only study assessing whether infants directly reward and punish equal and unequal distributors, a hallmark of moral evaluation. While infants rewarded an equal distributor, they didn’t punish unequal distributors, which aligns with our proposal that infants aren’t viewing unequal distributors as immoral. And similarly, rewarding equal distributors doesn’t necessarily indicate moral evaluations but possibly a desire to affiliate. Rewarding, reaching, and moving closer are all different ways in which infants may seek to affiliate with the equal distributor.

Thomas’s argument that infants have some of the cognitive tools necessary to represent EM relationships rests largely on the work of Tatone and colleagues (Tatone & Csibra, Reference Tatone and Csibra2020; Tatone, Geraci, & Csibra, Reference Tatone, Geraci and Csibra2015; Tatone, Hernik, & Csibra, Reference Tatone, Hernik and Csibra2021), which shows that infants especially encode features of giving actions relative to taking. However, tracking exchanges is not the same as representing equality. Infants may simply be attuned to the especially social nature of giving over taking. While Tatone’s work emphasizes the relational importance of giving – supporting our relational account – open questions remain about whether infants represent parity and reciprocal interactions, which are key to representations of EM relationships.

Ultimately, a theoretical account is only as good as the empirical evidence supporting it. Many tests could be conducted to differentiate between accounts of infants’ social evaluations of distributional outcomes. For instance, a future study could test whether infants infer stronger relationships from unequal than equal distributions. Similarly, our account predicts that infants should expect unequal allocations after witnessing unequal affiliative information between a distributor and recipients, but not after witnessing equal or no affiliations. Additionally, research that probes whether infants generalize an agent’s distributive tendencies to other moral or immoral behaviors or vice versa would help evaluate our proposal. As Thomas notes, some studies have been done suggesting that infants infer stable moral traits of individuals (Gill & Sommerville, Reference Gill and Sommerville2023; Surian et al., Reference Surian, Ueno, Itakura and Meristo2018). But it’s also possible that infants are reasoning about relationships in these studies. Future experiments could explore this possibility.

Our relational account also complements recent proposals of infant social affiliation such as adopted utility calculus (Powell, Reference Powell2022) that similarly challenge trait-based views of infant sociomoral evaluations. Although a full comparison of these accounts is beyond the scope of this commentary, we note that both Thomas and Powell (Reference Powell2022) broadly accept the evidence that babies expect and prefer equal distributional outcomes. We reinterpret the evidence as reflecting infants’ attention to relational cues embedded in distributions and their outcomes.

In summary, we agree with Thomas’s broader claims and applaud her thought-provoking effort to reexamine disparate findings in infant social cognition under a relational umbrella. Building on her account, we argue that infants track resource distributions to infer affiliative bonds. This potentially explains why infants look longer at unequal distributional outcomes and why they show preferences for equal distributors without invoking moral considerations. In doing so, we aim to extend Thomas’s broader thesis toward accounting for sociomoral evidence on relational rather than moral grounds.