An intrairial proactive interference design was used to examine the nature of pigeons' memory for duration in a delayed matching task. Short (2 s) or long (10 s) target samples were preceded on test trials by a short or long presample. The durations were consistent on some trials (shortshort or long-long) and inconsistent on others (short-long or long-short). Contrary to predictions based on prospective or categorical coding, accuracy was not related to duration consistency. Instead, accuracy was reduced on short-short and long-short trials and somewhat enhanced on short-long and long-long trials, suggesting that the pigeons "summed across" the durations. This occurred even with a 10-s interstimulus interval (Experiment 1) and even when the presample and target sample were physically distinct (Experiment 2). These results suggest that pigeons remember event durations in an analogical and retrospective fashion.

The early and preconscious processing of stimuli that are meaningful in everyday life includes systematical activations of many semantic, emotional and motor representations. Inhibitions should then occur in order to select, among these primed representations, those that are consistent with the context. Even in a lab this context is social, as it typically consists of the experimenter and of the instructions and stimuli (s)he provides. Three recent N400 studies confirm this social view of experimental settings by showing that socially driven processes affect what was primed by prior stimuli. The small amplitudes of the N400 event-related potentials (ERPs) elicited by stimuli preceded by semantical primes were found to be enhanced by the mere presence of a person next to participants when they know this person did not have the semantic primes. It thus seems that N400 processes inhibit what these primes have activated so that participants can also have the perspective of the uninformed person. This inhibition interpretation implies that N400s should be notably reduced when nothing allows to determine what should be inhibited, that is, when the social context is not defined and when task’s instructions require minimal inhibition. We tested this prediction by having a stranger next to participants (n=29) and by presenting meaningful unpredictable images in a simple memorization task. As foreseen, N400s were small. They were enhanced by definable social contexts, that is, in participants alone with the experimenter (n=30) and in those with a friend (n = 36). The amplitudes of the N300s were also enhanced. A second experiment revealed that these N300 and N400 enhancements were larger for friends who felt in the presence of their partner during most of the experiment. As to the late posterior positivities (LPPs) immediately succeeding the N400s, they were found to be larger in the unknown social context of the first experiment, suggesting that more information ended up being placed into the working memory when inhibitions could not occur. These results are compatible with a serial 3-stages framework of the processing of stimuli meaningful in everyday life. Early and broad systematic activations (priming) would be followed by automatic late selections done according to the social-context and then, by the participant’s consciousness of the meanings of the stimulus in this context. As inadequate late selections would cause impairments of social and cognitive behaviors, the present results could have implications for psychiatric disorders, such as autism and schizophrenia.

This research focuses on exploring personalized timing strategies to optimize learning in robot-child tutoring. Non-task breaks are commonly used in education to address children's limited attention spans and promote cognitive rejuvenation. Robots provide a valuable opportunity to deliver personalized breaks tailored to the specific needs of individual students, enhancing their learning experiences. We develop an autonomous robot tutoring system that assesses students' performance and administers breaks based on personalized schedules aligned with their individual progress. Through a field study, we compare the effectiveness of different break strategies, including a fixed timing approach, a reward strategy that aligns break timing with performance improvements, and a refocus strategy that aligns break timing with performance declines. Our results demonstrate that personalized strategies significantly enhance children's learning outcomes compared to the fixed strategy. Furthermore, we observe immediate benefits in terms of improved efficiency and accuracy in completing educational tasks following personalized breaks, underscoring the restorative effects of breaks when provided at optimal moments. These findings contribute to the understanding of how personalized timing strategies can optimize learning in the context of robot-child tutoring.