A naturalistic study comparing children at two schools with sharply different disciplinary cultures found that 3 and 4 year-olds at the school using corporal punishment lied more often and sustained their lies more convincingly than children at the school using non-punitive discipline.
Why This Question Was Worth Studying
Research on children’s lying has focused mainly on the cognitive side of the question: when children begin to lie, and how that ability develops with age. Far less attention had been paid to whether the social environment specifically, how adults respond to misbehavior shapes how much children lie and how well they do it.
Two competing ideas exist in the literature. One holds that strict punishment deters transgression, including dishonesty. The other, which researchers had not previously tested directly, is that the fear of severe punishment may motivate children to become more dishonest. The ethical impossibility of randomly assigning children to punitive versus non-punitive environments makes direct experimental testing infeasible. The existence of two comparable schools with markedly different disciplinary policies provided a rare opportunity to examine the question under naturalistic conditions.
What Was Already Known
Children’s ability to tell convincing lies develops gradually through early and middle childhood. Children aged 3 to 4 will often peek at a forbidden object when left alone and then deny it but they tend to give themselves away when asked follow-up questions, naming the object they claimed not to have seen. The ability to maintain a consistent deception across follow-up questioning called semantic leakage control typically becomes reliable around ages 6 to 7. Prior research had linked this skill to improvements in executive functioning, particularly working memory and inhibitory control, and to theory of mind, the understanding that another person may hold beliefs different from one’s own.
Research had also established that authoritarian parenting characterized by high control, low warmth, and punitive discipline is associated with delays in executive functioning and theory of mind, the same cognitive capacities that normally underpin effective lying. No study prior to this one had examined whether a punitive disciplinary environment specifically affects lying behavior.
How the Study Was Conducted
Victoria Talwar of McGill University and Kang Lee of the University of Toronto recruited 84 children aged 3 and 4 from two private schools in the same West African neighborhood. The schools were selected because they represented opposite ends of the disciplinary spectrum while drawing from the same local population, which controlled for socioeconomic background.
The punitive school used a traditional authoritarian discipline model. Corporal punishment including striking children with a stick, slapping, and pinching was administered publicly and routinely for offenses ranging from forgetting a pencil to classroom disruption. The non-punitive school used time-outs and verbal reprimand for minor infractions; more serious offenses resulted in a visit to the principal’s office. Physical punishment was not used.
Each child was brought individually to a test room and seated facing the researcher. A toy was placed on a table behind the child, out of direct sight. The researcher told the child not to peek at the toy, then left the room for 60 seconds. A hidden camera recorded whether the child peeked. Upon returning, the researcher asked three questions in sequence: whether the child had peeked; what the child thought the toy was (a probe for semantic leakage — if the child named the toy, they revealed having looked); and a final consistency probe. This procedure, called the Temptation Resistance Paradigm, is a validated method for studying lying in young children. It creates a naturalistic situation in which children are motivated to lie to conceal a transgression, without any instruction to do so.
Responses were coded for whether the child peeked, whether the child lied about peeking, and whether the child successfully concealed knowledge of the toy’s identity in follow-up responses.
What the Results Showed
Peeking rates did not differ significantly between the two schools. Both groups were comparably tempted.
Among children who peeked, those at the punitive school were significantly more likely to deny having done so. Approximately 82% of peekers at the punitive school lied about it, compared with approximately 38% at the non-punitive school.
On the follow-up questions, children at the punitive school were significantly better at maintaining their lie either naming an incorrect toy or feigning ignorance of the toy’s identity. Children at the non-punitive school were more likely to reveal knowledge of the toy inadvertently, which is consistent with the typical pattern for their age. The researchers note that the 3- and 4-year-old lie-tellers at the punitive school showed semantic leakage control at a level comparable to what prior research has found in 6- to 7-year-olds from non-punitive settings a finding they describe as surprising, given that punitive environments are generally associated with delays, not advances, in the executive functioning that normally supports this skill.
What the Researchers Concluded
Talwar and Lee concluded that a punitive environment not only increases the frequency of dishonest behavior in young children but also accelerates the specific competence required to maintain deception under questioning.
They proposed two non-exclusive explanations. First, children at the punitive school may have perceived the disciplinary policies as unjust, making lying a justifiable strategy to protect themselves from disproportionate consequences. Second, consistent with social learning theory, children may have improved their lying through repeated practice, refining deceptive strategies because the incentive to succeed avoiding punishment was consistently high.
As the authors write: “The present evidence thus suggests that a punitive environment not only fosters dishonesty but also children’s ability to lie to conceal their transgressions.”
They note that this finding sits in apparent tension with the standard cognitive-developmental account, which ties deceptive skill to executive function and theory of mind capacities that punitive environments tend to suppress rather than accelerate. Their interpretation is that social-motivational pressures, under sufficient intensity, can drive the development of specific deceptive competencies independently of general cognitive advancement.
This is a naturalistic comparison, not a randomized experiment. Children were not randomly assigned to schools. Although both schools drew from the same neighborhood and socioeconomic population, the researchers cannot rule out pre-existing differences between families who chose each school including differences in parenting attitudes or child temperament that might partially account for the observed differences in lying behavior.
The study involved 84 children from a single West African country. Whether findings generalize to other cultural contexts, to parenting settings outside of school, or to older children is not tested.
The study measured lying in a single controlled scenario. Whether this reflects children’s lying behavior across everyday contexts is an open question. The punitive/non-punitive distinction concerns the school environment; home parenting practices were not measured.
The study demonstrates an association between disciplinary environment and lying behavior. It does not isolate the mechanism. The two explanations the authors propose fear-motivated learning and social practice remain post hoc.
The accelerated semantic leakage control observed in children from the punitive school is, as the authors acknowledge, unexpected given known associations between harsh environments and cognitive delays, and requires further investigation.
References
Talwar, V., & Lee, K. (2011). A punitive environment fosters children’s dishonesty: A natural experiment. Child Development, 82(6), 1751–1758.
Victoria Talwar, Department of Educational and Counselling Psychology, McGill University.
Kang Lee, Institute of Child Study, University of Toronto.
Funded by the Social Sciences and Humanities Research Council of Canada and the National Institute of Child Health and Human Development (R01 HD047290; R01 HD048962).
Supporting citations: Talwar, Arruda & Yachison (JECP, 2015); Setoh et al. (JECP, 2020); Gershoff (Psychological Bulletin, 2002); Darling & Steinberg (Psychological Bulletin, 1993); Talwar & Lee (Child Development, 2008); Talwar, Gordon & Lee (Developmental Psychology, 2007).
