Cognitive Connie
Tronick's still face experiment explained
In 1978, Edward Tronick and colleagues at Harvard Medical School published what became one of the most emotionally arresting demonstrations in developmental psychology: the still face paradigm. In this simple but powerful procedure, a mother and her 3–6-month-old infant engage in normal, animated face-to-face interaction — the mother smiling, talking, mirroring the baby's expressions. Then, without warning, the mother assumes a neutral, unresponsive "still face" — no smiling, no vocalising, no response to the infant's bids for contact. Within seconds, the infant notices. They try increasingly to re-engage the mother — pointing, reaching, smiling, vocalising. When this fails, many infants withdraw: they look away, their posture slumps, their affect flattens. Some become distressed.
Defining features
Key figures
T. Berry Brazelton
1918–2018American paediatrician and co-author of the original still face study. Brazelton's broader programme of research on infant competence and caregiver-infant interaction created the institutional context in which the still face paradigm was developed. His Neonatal Behavioral Assessment Scale (NBAS) was a parallel attempt to reveal the richness of neonatal social competence.
Edward Tronick
1942–presentAmerican developmental psychologist and infant researcher, formerly at Harvard Medical School. Designed the still face paradigm and developed the mutual regulation model — the theoretical framework holding that early mother–infant interaction is a continuously co-regulated process in which attunement, misattunement, and repair shape the infant's developing emotional regulation capacities.
Katharine Adamson
fl. 1970sCo-author of the original 1978 still face study with Tronick, Als, Wise, and Brazelton — one of the first members of the team at Boston Children's Hospital to systematically observe and code the paradigm.
Methods
How the still face paradigm works
Participants
Typically infants aged 2–6 months and their primary caregivers (usually mothers in original studies; later studies included fathers and strangers). The paradigm is best suited to this age range because face-to-face interaction is the primary social modality for young infants who cannot yet reach or approach.
The three-phase structure
Phase 1 (Normal interaction, ~2 minutes): caregiver engages naturally with the infant — face-to-face, talking, smiling, following the infant's cues. Phase 2 (Still face, ~2 minutes): caregiver adopts a neutral, unresponsive expression — no talking, no smiling, no response to infant bids. Phase 3 (Reunion, ~2 minutes): caregiver resumes normal interaction.
What is coded
Infant facial expressions (positive, neutral, negative), vocalisations, gaze direction (toward mother, away), body posture, and self-comforting behaviours are coded from video recordings. The primary measures are proportion of time in positive affect, gaze to mother, and self-regulatory behaviours across the three phases.
Variations
Subsequent research tested the effect with fathers instead of mothers, with strangers (to control for familiarity), during phone calls (mother speaks on phone, face present but socially unavailable), with simulated still faces from televised images, and with depressed versus non-depressed mothers. Each variation probes what specifically drives the infant's response.
Findings
What Tronick found
The still-face effect is robust
Infants at 2–6 months consistently show significantly more positive affect, more gaze to the caregiver, and more social engagement during normal interaction than during the still face — and this difference is large and reliable across hundreds of replications. A 2001 meta-analysis by Mesman, van IJzendoorn, and Bakermans-Kranenburg confirmed the effect across 65+ published studies.
Infants first escalate, then disengage
A characteristic sequence during the still face: infants initially increase positive social bids (smiling, vocalising, pointing), apparently trying to re-engage the unresponsive caregiver. When these fail, behaviour shifts: gaze aversion increases, positive affect drops, some infants cry or show self-soothing behaviours (thumb-sucking, looking away, touching their own body).
Carry-over effect in reunion
A key finding: even when the mother resumes normal interaction in Phase 3, infants show less positive affect and engagement than at baseline. This 'still-face carry-over effect' — decreased positive affect persisting into reunion — suggests that even brief social disruption has a lasting impact on infant emotional state within the session.
Maternal depression affects the interaction quality
Infants of depressed mothers show a pattern resembling the still face even during normal interaction — because depressed mothers' face-to-face behaviour is less contingent, less animated, and less responsive. Repeated exposure to a depressed caregiver's low-contingency interaction may produce a chronically downregulated social engagement style in the infant.
Criticism
Challenges and limitations
Ecological validity of the still face
No actual caregiver maintains a perfectly neutral face for two minutes. The still face is a laboratory extreme — it produces a stronger disruption of the interaction than naturally occurring inattention or distraction. Whether the paradigm reflects real-world misattunement is debated.
Sample diversity
The majority of still face studies have used Western, middle-class, predominantly white samples. The centrality of intense face-to-face interaction in the paradigm may reflect a particular caregiving style (Western proximal care) not universal across cultures. In cultures where caregiving involves more body contact and less face-to-face engagement, infant responses to the still face paradigm may differ.
Interpretation of 'disengagement'
When infants look away during the still face, this is typically interpreted as withdrawal or distress. But looking away is also a known strategy for self-regulation — infants may avert gaze to reduce arousal rather than as a distress signal. Distinguishing regulatory looking-away from disengaged withdrawal requires careful coding and physiological measures.
Long-term effects not well established
The still face effect is highly consistent within a single session, but whether a single exposure to the paradigm has lasting effects on infant development is less well established. The clinical relevance is strongest for chronic exposure to unresponsive caregiving (e.g., maternal depression) rather than brief experimental disruption.
Relevance for today
Why it still matters
Infant mental health and maternal depression
The still face has been widely used in research on maternal depression and infant social-emotional development. Infants of depressed mothers show chronically reduced positive affect and contingent social engagement — a pattern associated with insecure attachment, emotional dysregulation, and later internalising problems. Interventions targeting maternal sensitivity and responsiveness (e.g., interaction coaching, video feedback) can improve both maternal behaviour and infant outcomes.
Screen time and digital distraction
Contemporary researchers have used still-face logic to study the impact of parental smartphone use on infant interaction — posing the smartphone as a modern analogue of the still face, in which the parent is physically present but socially unavailable. Studies show that parental phone use during face-to-face interaction reduces infant-directed speech and contingent responsiveness, with effects on infant attention and engagement.
Co-regulation and self-regulation development
The mutual regulation model has become influential in understanding how self-regulation develops: through repeated cycles of attunement, misattunement, and repair, infants gradually internalise regulatory capacities. Tronick's framework influenced the field of infant mental health and is foundational to models of how early caregiving shapes the developing emotion regulation system.
Early intervention and attachment
The still face paradigm is used as an assessment tool in clinical and research settings to evaluate the quality of parent–infant interaction and identify dyads at risk. Observing how a parent and infant co-regulate during and after the still face provides information about the interactive quality that cannot be obtained from questionnaires alone.
Frequently asked questions
What is the still face experiment?+
Edward Tronick's still face paradigm (1978) has three phases: normal face-to-face interaction between a caregiver and a 2–6-month-old infant; a still face phase in which the caregiver adopts a neutral, unresponsive expression for ~2 minutes; and a reunion phase in which normal interaction resumes. During the still face, infants initially intensify their social bids to re-engage the caregiver, and when these fail, disengage — showing reduced positive affect, gaze aversion, and self-soothing behaviour. Even in reunion, positive affect remains below baseline — a 'carry-over effect.'
Why is the still face experiment important?+
It demonstrated that even 2–3-month-old infants are exquisitely sensitive to the quality of social contingency — not just the presence of a caregiver, but whether the caregiver is actively and responsively engaged. It showed that very young infants already try to repair disrupted interactions, and that repeated exposure to unresponsive caregiving (as in maternal depression) can produce chronic downregulation of infant social engagement — with implications for early emotional development, attachment, and infant mental health.
Sources
Last reviewed August 2026- 1.
Tronick E., Als H., Adamson L., Wise S., & Brazelton T.B. (1978). The infant's response to entrapment between contradictory messages in face-to-face interaction. Journal of the American Academy of Child Psychiatry, 17(1), 1–13. https://doi.org/10.1097/00004583-197800170-00001
+About this source
Original paper introducing the still face paradigm and the mutual regulation model.
- 2.
Mesman J., van IJzendoorn M.H., & Bakermans-Kranenburg M.J. (2009). The many faces of the Still-Face Paradigm: A review and meta-analysis. Developmental Review, 29(2), 120–162. https://doi.org/10.1016/j.dr.2009.02.001
+About this source
Comprehensive meta-analysis of 65+ published still face studies confirming the robustness of the still-face effect.