Thomas (in press) proposes that infants possess an intuitive core cognitive system for reasoning about social relationships. A Tinbergenian analysis (Bateson & Laland, Reference Bateson and Laland2013; Tinbergen, Reference Tinbergen2005) provides a valuable framework to deepen our understanding of this claim by examining how such relational reasoning evolved, develops, functions neurally, and serves adaptive purposes. Focusing specifically on Thomas et al. saliva-sharing experiments with infants, I evaluate the evidence and propose that integrating evolutionary, developmental, and mechanistic perspectives can strengthen and refine the proposed framework.
1. Phylogeny: how did it evolve?
Humans are unique among great apes in exhibiting cooperative breeding, wherein mothers and allomothers – grandmothers, siblings, and unrelated group members – actively contribute to infant care and feeding (Hrdy, Reference Hrdy2009; Hrdy & Burkart, Reference Hrdy and Burkart2020). This cooperative system likely emerged around 1.8 million years ago with Homo erectus, profoundly reshaping human infant psychology, emotion, and cognition. According to Hrdy (Reference Hrdy2009), this socio-ecological shift selected for infants equipped with enhanced social cognitive skills, enabling them to infer others’ intentions, preferences, and commitments – skills crucial for engaging multiple caregivers.
Thomas et al.’s finding that infants infer emotional closeness from food/saliva-sharing behaviors resonates with this evolutionary background. In ancestral environments, behaviors such as kiss-feeding – mouth-to-mouth transfer of food – served as salient cues of nurturance and caregiving commitment. Infants’ sensitivity to food (and saliva) sharing thus plausibly reflects deep phylogenetic adaptations shaped by cooperative breeding. The human-specific evolutionary context of communal caregiving and provisioning must be considered foundational when theorizing about infants’ social cognitive abilities and their sensitivity to feeding-related relational cues.
2. Ontogeny: how does it develop?
Developmentally, it is important to consider the timeline of exposure to caregiving and feeding behaviors. Kiss-feeding practices, involving the transfer of food or honey-flavored saliva, have been observed among hunter-gatherers, beginning around 6 months, coinciding with infants’ eruption of teeth (Hrdy, Reference Hrdy2009; Hrdy & Burkart, Reference Hrdy and Burkart2020). Even in WEIRD societies, solid food introduction is encouraged around 6 months, often involving close physical feeding interactions (American Academy of Pediatrics, 2012).
Notably, the youngest infants tested in Thomas et al. were 8.5 months old – an age by which most would have already experienced feeding interactions. Thus, infants’ expectations about relational closeness based on food/saliva-sharing cues may reflect accumulated social experience rather than exclusively core relational reasoning. To rigorously test Thomas’s (in press) proposal that infants possess core relational reasoning abilities, studies should investigate younger infants, ideally before 6 months, prior to substantial feeding experience. Such research would clarify the extent to which sensitivity to food/saliva-sharing cues is experience-dependent versus reflecting early, possibly evolved, conceptual structures.
Moreover, Ciftci and colleagues (Reference Ciftci, Kominsky and Csibra2025), while replicating Thomas et al.’s findings with infants, noted that the original study did not test whether infants’ responses were driven by food sharing in general or saliva sharing specifically. They pointed out that alternative conditions – such as feeding without saliva sharing or saliva sharing without direct feeding – were not tested. Since feeding alone may signal a strong social relationship and is easier for infants to detect, it could independently explain infants’ responses.
3. Brain mechanisms: how does it work?
Thomas’s (in press) proposal pays limited attention to the underlying brain mechanisms supporting infants’ relational reasoning. Recent neuroimaging work offers valuable insights. Farris et al. (Reference Farris, Kelsey, Krol, Thiele, Hepach, Haun and Grossmann2022) demonstrated that infants between 6 and 13 months engage the dorsomedial prefrontal cortex (dmPFC) when observing third-party social interactions characterized by physical contact and synchrony – cues relevant for communal sharing. These findings align with evidence from adult neuroscience, where the medial prefrontal cortex is implicated in mentalizing and social evaluation (Amodio & Frith, Reference Amodio and Frith2006).
The recruitment of dmPFC in infants suggests that brain specialization for processing social relational information emerges early. However, an important open question is whether the infant brain differentiates between relational categories beyond communal sharing, such as authority ranking and equality matching, as proposed by Fiske’s (Reference Fiske1992) Relational Models Theory. To date, no neuroimaging studies, even in adults, have systematically tested these relational distinctions within the same experimental paradigm. Future work should explore whether brain processes are differentially engaged by distinct relational models and how the neural bases of social relationship reasoning develop during infancy.
4. Adaptive value: what is it for?
From an adaptive perspective, early relational reasoning likely provided significant survival benefits in human evolutionary contexts. Infants who could recognize supportive, nurturing alloparents – beyond the primary caregiver – would have been more likely to survive and thrive (Hrdy, Reference Hrdy2009; Hrdy & Burkart, Reference Hrdy and Burkart2020). Sensitivity to behavioral cues such as food sharing, affectionate touch, and synchrony would allow infants to identify individuals willing to invest caregiving resources.
Relational reasoning thus likely evolved as an ontogenetic adaptation enabling infants to actively engage and recruit alloparents. Given the extraordinary costs of prolonged human infancy and childhood, the capacity to navigate complex social networks from an early age would have been crucial. Infants’ ability to interpret food/saliva-sharing behaviors as indicators of emotional closeness fits within this broader adaptive framework.
More generally, early relational reasoning may form part of the broader suite of human-specific social cognitive adaptations supporting cooperation (Tomasello, Reference Tomasello2020). Sensitivity to social relational information, including communal caregiving structures, may be an essential component of the evolutionary foundations of human sociability.
Conclusion
Applying Tinbergen’s framework to Thomas’s (in press) proposal enriches our understanding of infants’ social relationship reasoning. Human infants’ sensitivity to relational cues such as food sharing is likely shaped by both evolutionary pressures favoring cooperative caregiving and developmental experiences within socially rich environments. Brain specialization for processing relational information appears early but requires further empirical investigation to fully map onto relational models.
While Thomas makes a strong case for early-developing relational knowledge, future work should rigorously disentangle core cognitive architecture from experiential learning and probe the neural mechanisms supporting relational distinctions. A Tinbergenian approach reveals that infants’ relational reasoning is likely not only an early-emerging skill but also an adaptive product of human cooperative social life.
Ultimately, understanding how infants reason about social relationships requires integrating evolutionary history, developmental processes, brain function, and adaptive significance. Doing so promises a more comprehensive account of the origins of the human social mind.
You have
Access
Thomas (in press) proposes that infants possess an intuitive core cognitive system for reasoning about social relationships. A Tinbergenian analysis (Bateson & Laland, Reference Bateson and Laland2013; Tinbergen, Reference Tinbergen2005) provides a valuable framework to deepen our understanding of this claim by examining how such relational reasoning evolved, develops, functions neurally, and serves adaptive purposes. Focusing specifically on Thomas et al. saliva-sharing experiments with infants, I evaluate the evidence and propose that integrating evolutionary, developmental, and mechanistic perspectives can strengthen and refine the proposed framework.
1. Phylogeny: how did it evolve?
Humans are unique among great apes in exhibiting cooperative breeding, wherein mothers and allomothers – grandmothers, siblings, and unrelated group members – actively contribute to infant care and feeding (Hrdy, Reference Hrdy2009; Hrdy & Burkart, Reference Hrdy and Burkart2020). This cooperative system likely emerged around 1.8 million years ago with Homo erectus, profoundly reshaping human infant psychology, emotion, and cognition. According to Hrdy (Reference Hrdy2009), this socio-ecological shift selected for infants equipped with enhanced social cognitive skills, enabling them to infer others’ intentions, preferences, and commitments – skills crucial for engaging multiple caregivers.
Thomas et al.’s finding that infants infer emotional closeness from food/saliva-sharing behaviors resonates with this evolutionary background. In ancestral environments, behaviors such as kiss-feeding – mouth-to-mouth transfer of food – served as salient cues of nurturance and caregiving commitment. Infants’ sensitivity to food (and saliva) sharing thus plausibly reflects deep phylogenetic adaptations shaped by cooperative breeding. The human-specific evolutionary context of communal caregiving and provisioning must be considered foundational when theorizing about infants’ social cognitive abilities and their sensitivity to feeding-related relational cues.
2. Ontogeny: how does it develop?
Developmentally, it is important to consider the timeline of exposure to caregiving and feeding behaviors. Kiss-feeding practices, involving the transfer of food or honey-flavored saliva, have been observed among hunter-gatherers, beginning around 6 months, coinciding with infants’ eruption of teeth (Hrdy, Reference Hrdy2009; Hrdy & Burkart, Reference Hrdy and Burkart2020). Even in WEIRD societies, solid food introduction is encouraged around 6 months, often involving close physical feeding interactions (American Academy of Pediatrics, 2012).
Notably, the youngest infants tested in Thomas et al. were 8.5 months old – an age by which most would have already experienced feeding interactions. Thus, infants’ expectations about relational closeness based on food/saliva-sharing cues may reflect accumulated social experience rather than exclusively core relational reasoning. To rigorously test Thomas’s (in press) proposal that infants possess core relational reasoning abilities, studies should investigate younger infants, ideally before 6 months, prior to substantial feeding experience. Such research would clarify the extent to which sensitivity to food/saliva-sharing cues is experience-dependent versus reflecting early, possibly evolved, conceptual structures.
Moreover, Ciftci and colleagues (Reference Ciftci, Kominsky and Csibra2025), while replicating Thomas et al.’s findings with infants, noted that the original study did not test whether infants’ responses were driven by food sharing in general or saliva sharing specifically. They pointed out that alternative conditions – such as feeding without saliva sharing or saliva sharing without direct feeding – were not tested. Since feeding alone may signal a strong social relationship and is easier for infants to detect, it could independently explain infants’ responses.
3. Brain mechanisms: how does it work?
Thomas’s (in press) proposal pays limited attention to the underlying brain mechanisms supporting infants’ relational reasoning. Recent neuroimaging work offers valuable insights. Farris et al. (Reference Farris, Kelsey, Krol, Thiele, Hepach, Haun and Grossmann2022) demonstrated that infants between 6 and 13 months engage the dorsomedial prefrontal cortex (dmPFC) when observing third-party social interactions characterized by physical contact and synchrony – cues relevant for communal sharing. These findings align with evidence from adult neuroscience, where the medial prefrontal cortex is implicated in mentalizing and social evaluation (Amodio & Frith, Reference Amodio and Frith2006).
The recruitment of dmPFC in infants suggests that brain specialization for processing social relational information emerges early. However, an important open question is whether the infant brain differentiates between relational categories beyond communal sharing, such as authority ranking and equality matching, as proposed by Fiske’s (Reference Fiske1992) Relational Models Theory. To date, no neuroimaging studies, even in adults, have systematically tested these relational distinctions within the same experimental paradigm. Future work should explore whether brain processes are differentially engaged by distinct relational models and how the neural bases of social relationship reasoning develop during infancy.
4. Adaptive value: what is it for?
From an adaptive perspective, early relational reasoning likely provided significant survival benefits in human evolutionary contexts. Infants who could recognize supportive, nurturing alloparents – beyond the primary caregiver – would have been more likely to survive and thrive (Hrdy, Reference Hrdy2009; Hrdy & Burkart, Reference Hrdy and Burkart2020). Sensitivity to behavioral cues such as food sharing, affectionate touch, and synchrony would allow infants to identify individuals willing to invest caregiving resources.
Relational reasoning thus likely evolved as an ontogenetic adaptation enabling infants to actively engage and recruit alloparents. Given the extraordinary costs of prolonged human infancy and childhood, the capacity to navigate complex social networks from an early age would have been crucial. Infants’ ability to interpret food/saliva-sharing behaviors as indicators of emotional closeness fits within this broader adaptive framework.
More generally, early relational reasoning may form part of the broader suite of human-specific social cognitive adaptations supporting cooperation (Tomasello, Reference Tomasello2020). Sensitivity to social relational information, including communal caregiving structures, may be an essential component of the evolutionary foundations of human sociability.
Conclusion
Applying Tinbergen’s framework to Thomas’s (in press) proposal enriches our understanding of infants’ social relationship reasoning. Human infants’ sensitivity to relational cues such as food sharing is likely shaped by both evolutionary pressures favoring cooperative caregiving and developmental experiences within socially rich environments. Brain specialization for processing relational information appears early but requires further empirical investigation to fully map onto relational models.
While Thomas makes a strong case for early-developing relational knowledge, future work should rigorously disentangle core cognitive architecture from experiential learning and probe the neural mechanisms supporting relational distinctions. A Tinbergenian approach reveals that infants’ relational reasoning is likely not only an early-emerging skill but also an adaptive product of human cooperative social life.
Ultimately, understanding how infants reason about social relationships requires integrating evolutionary history, developmental processes, brain function, and adaptive significance. Doing so promises a more comprehensive account of the origins of the human social mind.
Acknowledgments
None.
Financial support
T.G. was partially supported by the National Science Foundation #2017229.
Competing interests
No competing interests.