Axis 1
In this axis we examine whether high-level visual categorization of food is present already in the first year of life. An approach particularly suited to capture the development of visual categorization processes is the fast periodic visual stimulation paradigm (FPVS) coupled with electroencephalography (EEG). This FPVS-EEG approach relies on the property of the brain to synchronize with stimuli presented periodically, eliciting a periodic change of voltage amplitude in the EEG signal at the same frequency. In this paradigm, streams of visual stimuli (i.e. base stimuli) are presented at a fixed presentation rate (usually 6Hz, i.e. six images per second), eliciting an EEG response at the same frequency. Stimuli from a target category are periodically inserted in the sequence of images at a slower rate (e.g. every fifth images, that is at a 1.2Hz frequency, AAAABAAAAB…), eliciting an additional EEG response at this slower frequency if the brain discriminates between the base and target stimuli and generalized across target items
In axis 1, two groups of adults and two groups of infants (12-month-old and 6-months infants) were tested in an FPVS paradigm, with contrasts spanning the food / non-food distinction. Given the novelty of using food stimuli in FPVS paradigms, adults were tested to provide a reference for infants’ electrophysiological responses. Twelve-month-old infants were selected because they have some experience with solid foods: The typical age of food diversification in France is 4- to 6-months and by 12 months they already eat solid foods in pieces such as raw and cooked fruits and vegetables, starchy foods etc. (Somaraki et al., 2024; Nicolas, 2019). Six-month-old infants were chosen because it is around that age that infants are first introduced to solid food, allowing us to investigate if infants start to be able to categorize food as soon as they transition to solid foods.
In summary, our results from Axis 1 revealed that both adults and 12-month-old infants show neural responses indicative of visual categorization of food from natural kinds and artefacts, although infant categorization was less pronounced compared to adults. These results confirm the sensitivity of the FPVS paradigm in measuring adults’ and infants’ neural responses to food. Analysis with 6-month-old infants is ongoing. Overall, the study contributes to the development of methods such as FPVS to study cognitive development in infants and provide new insights into when visual category representations emerge in infants’ brains.
Axis 2
Food learning in early life relies on social learning and infants are selective in the way they use social information to learn about food: Infants learn that a particular entity is edible after observing an adult eating it, but not when they handle it in other ways. However, it is still unclear how infants react to negative information applied to food.
To examine this question, we conducted a series of looking time experiments with 14-month-olds because they have had experience with solid foods and often engage in social referencing, i.e. looking at an adult and adjusting their behavior based on the adult's behavior. We employed a violation of expectation (VoE) paradigm to test whether infants recognize that disgust, but not other negative emotions, carries information about (non) edibility. In VoE paradigms, infants watch a familiarization event. Then, they watch two test events, one coherent with the familiarization event and one incoherent and infants’ looking time during these two test events are recorded. A longer looking time during the incoherent event indicates that infants’ expectations have been violated. VoE paradigms are robust and extremely well-validated paradigms in developmental psychology.
In a first experiment, at familiarization I presented infants with an actor being disgusted by the sight of food A and displaying neutral emotion towards food B. At test phase, infants saw the same actor eating food A (incoherent event) and B (coherent event). I found longer looking times during the incoherent event, revealing that infants expect an adult to avoid eating food for which they have shown disgust previously. In a second experiment, I presented infants with an actor being angry at the sight of food A and displaying neutral emotion towards food B. At test phase, infants saw the same actor eating food A (incoherent event) and B (coherent event). In that second experiment, I did not find longer looking times during the incoherent event, suggesting that infants did not expect an adult to avoid eating a food for which they have shown anger previously. These are promising results, showing selectivity in how infants use negative information about food. In a third experiment, infants saw the same familiarization as in Experiment 1 but during the test phase they saw a second actor (E2) eating both foods. It remains to be tested whether infants interpret disgust expressions as conveying information that generalizes across individuals instead of a person-specific subjective feeling. In this experiment 3 infants looked longer during the incoherent event, revealing that they interpreted disgust expressions as conveying information that generalizes across individuals instead of a person-specific subjective feeling.