Humans are keenly attuned to the structure of the social networks around them and frequently apply knowledge of such networks in day-to-day life. Adults use knowledge of the distance (i.e., “degrees of separation”) between people and various measures of centrality (i.e., individuals’ importance) in social networks to predict how information will flow, then shape their behavior accordingly – e.g., when gossiping (Xia et al., Reference Xia, Teoh, Nassar, Bhandari and FeldmanHall2025), assessing the veracity of gossip (Hess & Hagen, Reference Hess and Hagen2006), choosing “scapegoats” (Ellwardt et al., Reference Ellwardt, Labianca and Wittek2012), and deciding how to spread information (Banerjee et al., Reference Banerjee, Chandrasekhar, Duflo and Jackson2019). The selective sharing of information in social networks – and its strategic deployment to manipulate alliances – is perhaps further influenced by the relative strengths of the relationships around oneself (DeScioli & Kurzban, Reference DeScioli and Kurzban2009; Krems et al., Reference Krems, Merrie, Rodriguez and Williams2024). Monitoring the patterns of relationships that comprise social networks also allows people to recognize potential third-party threats to their relationships (Krems et al., Reference Krems, Williams, Aktipis and Kenrick2021), strategically manipulate their social status (Faris, Felmlee, & McMillan, Reference Faris, Felmlee and McMillan2020; Krems et al., Reference Krems, Merrie, Rodriguez and Williams2024), and decide whom to trust (Schwyck et al., Reference Schwyck, Du, Li, Chang and Parkinson2024). People are so attuned to information about others’ social network positions (e.g., “degrees of separation” from oneself, how well-connected they are) that they spontaneously retrieve such knowledge when encountering familiar others (Parkinson, Kleinbaum, & Wheatley, Reference Parkinson, Kleinbaum and Wheatley2017; Zerubavel et al., Reference Zerubavel, Bearman, Weber and Ochsner2015).

Despite a surge of evidence for and interest in the impact of social network knowledge on cognition and behavior, little is known about the development of this knowledge, or of humans’ ability to efficiently learn or strategically apply it. Of particular relevance to the target article, one significant open question pertains to which aspects of humans’ understanding of social networks, if any, are rooted in “core knowledge” that emerges very early in development.

It is plausible that from early on in life, humans are prepared to perceive and reason about social networks. As Thomas points out, from birth, survival depends on such networks that extend well beyond one’s primary caregivers (Hrdy, Reference Hrdy2009); and as only some members of one’s parents’ social groups provide care, it would be beneficial to track specific relationships, not just group affiliations.

However, much remains to be discovered regarding how and when capacities to learn, represent, and use knowledge of social networks develop. For example, adults readily identify the people able to most effectively spread information in their networks due to the number and patterning of those people’s ties (Banerjee et al., Reference Banerjee, Chandrasekhar, Duflo and Jackson2019). The measures of centrality that tend to predict ability to effectively spread information depend on not just how many friends or allies someone has, but also the people to whom that person is connected, and how well-connected those individuals are. How early on do the capacities to acquire, represent, and use such information emerge? Do inferences about such characteristics require granular social network knowledge, or could they be supported by simpler heuristics (e.g., tracking who one hears about most often or associated traits)? Related questions pertain to information-seeking – e.g., since central individuals are exposed to more people, they often know more about others in the network and may be more representative of its norms. When do we start to understand and make use of such phenomena?

Elucidating the developmental origins of the ability to track and infer others’ centrality will also inform our understanding of how children reason about status and influence, as social network centrality facilitates influence and may be a particularly behaviorally relevant facet of status (Basyouni & Parkinson, Reference Basyouni and Parkinson2022). Importantly, tracking and reasoning about measures of centrality may require different cognitive underpinnings than those that support analogous operations on other, more widely studied, facets of social status (e.g., reasoning about people’s relative positions in linearly structured dominance hierarchies using transitive inference).

Research is also needed to elucidate the developmental trajectory of understanding and using information about distances in social networks. For example, when and how do our abilities to represent and apply knowledge of such distances develop, so that we can ascertain that people fewer degrees apart from one another in a network are more likely to know the same things, know things about one another, and be similar to one another?

Additionally, as Thomas notes, some results typically attributed to representations of social groups may instead reflect representations of social networks and their constituent relationships. Disentangling these possibilities will be informative in studies of both children and adults. This may be difficult, as some features of social networks and the schemas used to encode them can make representations of social networks and groups appear similar. For example, human social networks tend to contain many balanced triads and to have small-world structures (i.e., to consist of densely connected groups or “clusters” with few connections between them; Watts, Reference Watts1999). Adults tend to exaggerate these properties in their mental representations (Brands, Reference Brands2013; Kilduff et al., Reference Kilduff, Crossland, Tsai and Krackhardt2008). This tendency could arise from the need to compress representations of large, complex networks and/or the use of heuristics (Basyouni & Parkinson, Reference Basyouni and Parkinson2022). It could also stem from a tendency to represent coalitional groups rather than networks. Similarly, from infancy to adulthood, humans tend to infer affiliative ties between individuals who share certain properties (e.g., preferences, traits) even without ever witnessing them interact (Liberman, Kinzler, & Woodward, Reference Liberman, Kinzler and Woodward2021; Schwyck et al., Reference Schwyck, Du, Li, Chang and Parkinson2024). In some contexts (e.g., paradigms where groups are defined by individuals sharing particular properties), this tendency could render it difficult to discern whether participants’ behavior stems from representations of groups or from inferences regarding which individuals would be more likely to be connected by affiliative ties. Thus, more work is needed to elucidate interactions between mental representations of social networks and groups, and advancing understanding of social network knowledge and its development could reshape the way that certain existing findings are understood.

Understanding how and when humans begin to represent and reason about social networks will shed light on the development of skills that are central to everyday life. From deciding whom to trust to predicting how information will spread, network knowledge fundamentally shapes how people make sense of and act within their social worlds.