Reassessing the two-stage theory of social learning development: a discussion

Yaming Shang, Ying Xia, Yifang Yin, Da DongFrontiers in Psychology 2025 · DOI

TL;DR

This paper reevaluates the "two-stage theory" of social learning, which describes how children transition from being sensitive to pedagogical cues (like eye contact and direct instruction) to becoming "selective learners" who vet the reliability of their information sources. The authors argue that understanding this developmental shift is essential for designing modern educational practices and intelligent learning technologies that foster a child's "epistemic agency."

What Problem Does This Paper Address?

In developmental psychology, social learning isn't just about absorbing information; it’s about who you learn from and how they signal their intent to teach. Traditionally, research focused on pedagogical learning—how infants respond to communicative signals (gestures, "parentese," eye contact). However, as children grow, they become more discerning.

The authors address the challenge of the "two-stage" transition: how does a child move from being a relatively passive recipient of "teaching" to an active, autonomous agent who evaluates a source's expertise? In an era of digital misinformation and AI-driven education, understanding how children develop the "filters" to prioritize reliable information over mere pedagogical flashiness is a critical gap in both psychology and educational theory.

Methods

As a Perspective article, this paper does not introduce a new laboratory experiment or algorithm. Instead, it employs a theoretical synthesis and literature review. The authors:

  • Reevaluate the Two-Stage Theory: They synthesize existing longitudinal studies (e.g., Carpenter et al., 1998) and meta-analyses to map the trajectory of social learning.
  • Cross-Disciplinary Integration: They bridge developmental psychology (Piaget’s formal operations, Vygotsky’s Zone of Proximal Development) with educational technology frameworks like TPACK (Technological Pedagogical Content Knowledge) and ILE (Intelligent Learning Environments).
  • Conceptual Modeling: They propose a shift in the educator's role from "knowledge transmitter" to "architect of the learning environment" based on the child's maturing selective abilities.

Datasets

Not discussed in the paper. The article relies on previously published data, including a cited meta-analysis of 63 studies involving 6,525 participants regarding selective vocabulary learning in children aged 2 to 5.

Key Insights and Ideas

  • The "Waning" of Pedagogical Cues: Around 18–24 months, the "magic" of pedagogical cues begins to fade. While a 12-month-old might be captivated by direct eye contact, a 20-month-old is equally likely to follow a gaze regardless of whether the adult makes eye contact first. They start looking past the signal to the substance.
  • The Rise of the Expert: By ages 4–7, children show a clear preference for learning from individuals they perceive as "experts." They aren't just imitating; they are performing a "cognitive assessment" of the source's reliability.
  • Epistemic Agency: The authors emphasize that selective learning is a form of autonomy. Children actively filter information based on personal goals and perceived relevance, moving beyond the "passive recipient" model.
  • The Educator’s Pivot: The paper argues that as children’s selective learning matures, the teacher’s role must shift from direct instruction to facilitation. This involves providing "supportive signals" rather than "direct commands," helping children navigate a wider range of resources.
  • Algorithmic Bias in Education: A surprising insight is the warning regarding "selective categorization" in machine learning. The authors note that if educational AI models reject certain predictions or pathways based on biased data, they can exacerbate inequalities, mirroring the very risks selective learning aims to mitigate in humans.

Key Results

  • Developmental Milestones: The paper highlights that 14-month-olds already show selective imitation based on perceived competence (e.g., imitating an adult who puts shoes on correctly vs. incorrectly), while 18-month-olds will actively decline to learn labels from "inaccurate" informants.
  • Independence of Social Learning: The authors cite research showing no significant correlation between domain-general learning abilities and infant selective social learning, suggesting that the "social filter" operates as a distinct cognitive mechanism.
  • Framework Evolution: The authors successfully link the psychological transition to the ILE (Intelligent Learning Environment) framework, arguing that adaptive technology must mirror the child's developmental shift toward autonomy to be effective.

Limitations and Future Work

  • Cultural Variability: The authors acknowledge that most current social learning trajectories are based on Western models. They call for cross-cultural longitudinal studies to see how different value systems shape information selection.
  • Internal Mechanisms: While we know that children become selective, we don't fully understand how emotional states or specific social relationships (like in-group vs. out-group bias) modulate this selectivity.
  • Neuroimaging: The authors suggest that future research should integrate experimental paradigms with neuroimaging to identify the neural correlates of the transition from pedagogical to selective learning.
  • Impact of EdTech: There is a need for empirical studies on how emerging technologies (VR, online platforms) influence the way children observe and emulate behaviors compared to traditional face-to-face interactions.

Summary updated

Abstract

This article reevaluates the two-stage development theory of social learning by examining the distinct roles of pedagogical and selective learning. Pedagogical learning involves infants' sensitivity to adults' communicative intentions, while selective learning highlights how infants prioritize information from reliable sources. Over time, the influence of pedagogical learning diminishes, whereas selective learning continues to play a central role as children grow. We argue for the distinctiveness and plurality of human social learning and emphasize the foundational role of multiple social interactions in cognitive development. Special attention is given to the transition from reliance on pedagogical cues to selective learning, underscoring the critical role of educators in facilitating this shift. We conclude by highlighting the increasing sophistication of cognitive and social selection processes in observation, imitation, and learning, alongside the implications for educational practices. Future research should investigate differences between social and other forms of learning, the mechanisms underlying the diminishing effects of pedagogical cues with age, the cultural variability in social learning trajectories, and the impact of educational technology on the development of social learning.
BibTeX
@inproceedings{shang2025reassessing,
  author = {Yaming Shang and Ying Xia and Yifang Yin and Da Dong},
  title = {{Reassessing the two-stage theory of social learning development: a discussion}},
  booktitle = {Frontiers in Psychology},
  year = {2025},
  doi = {10.3389/fpsyg.2025.1576714},
}
Figure