From Humour to Curiosity – How Surprise, Arousal, and Attention Shape Learning.

During my PhD, we examined how humour supports infant learning and through which mechanisms. Building on Esseily et al. (2015), Study 1 (n = 119; 14–22 months) tested a rake-use task in which an experimenter demonstrated how to retrieve an out-of-reach toy with a tool, either humorously or neutrally. Infants exposed to the humorous demonstration – whether or not they laughed – performed better than those exposed to the neutral one. Combining fine-grained behavioural coding with physiological measures from a connected, ankle-worn wristband, we found that better performers showed higher HRV and lower HR (a marker of infant attention; Richards & Casey, 1991) and gazed more at the experimenter during the demonstration, suggesting that humour’s benefit is partly attention-driven. Because humour entails expectation violation, we hypothesized a role for the emotion of surprise. Before testing this, we needed to measure surprise in a non-standardized, social tool-use context. A methodological study (n = 132; 14–22 months) compared candidate markers: infants’ social gazes (Mireault et al., 2014), facial expressions of surprise via three algorithms I designed within BabyFaceReader® (automated facial analysis), physiological indices (HR, electrodermal activity), and movement quantified by accelerometery. Social gazes emerged as the most accurate surprise marker, whereas facial expression of surprise and freezing behaviours were more closely linked to sustained attention. Finally, Study 3 (n = 80; 17–22 months) tested whether surprise alone could partly explain the effect of humour on learning in infants. In this context, infants were exposed to three different demonstration types: neutral, surprising (non-humorous), and humorous. Our results showed that infants exposed to the humorous demonstration outperformed the two other conditions. We interpreted this result as: while surprise can have a positive influence on learning in infants, it is not sufficient to explain the positive effect of humour on infants’ learning. We suggested that humour could involve other factors explaining why it is particularly efficient at improving infants’ learning. We hypothesised that positive affect, affiliative bonding, and focused attention likely act together, giving infants a particularly strong opportunity to learn.

Since 1 February 2025, I have joined Adélaïde de Heering’s team to investigate how uncertainty shapes infants’ curiosity, building on Kidd et al.'s (2012, 2014) finding that 8-month-old infants attend most to stimuli that are neither too uncertain nor too certain – the Goldilocks effect. In the same vein as my PhD, I bring a multidimensional approach that I intend to make my specialty. As an ethologist and evolutionist by master’s training, I also hold that to understand a construct we must understand its evolutionary history (Dobzhansky, 1973); accordingly, we will adapt this protocol across several species to probe the evolutionary function/history of the Goldilocks effect and ask: is it an adaptation that optimises the cost–benefit trade-off in information processing, and is it shared with our closest cousins, non-human primates? Our project includes three independent work packages (WPs) to address the current lack of a multidimensional understanding of how uncertainty relates to curiosity, with the goal of (1) identifying how uncertainty, surprise, arousal, and attention shape curiosity within a multidimensional framework combining behavioural and physiological measures (WP1); (2) investigating the evolutionary continuity of the Goldilocks effect as an adaptation that maximizes the cost-to-benefit ratio in information acquisition (Kidd & Hayden, 2015), by testing whether this adaptation is a common feature shared with our closest relatives, the non-human primates (WP2), and (3) formalizing the relationships among uncertainty, surprise, arousal, attention, and curiosity in a computational model (WP3).