Abstract/Summary

While curiosity – the intrinsic desire to know – is a concept central to the human mind and knowledge acquisition, scientific research targeting the understanding of curiosity is still in its infancy and has only recently begun to unravel it. Studies on information-seeking, a popular way to manipulate and measure curiosity in the lab, found that information shows similar rewarding properties as other, extrinsic rewards/incentives like food or money. Indeed, both can motivate behaviour and elicit a response in the dopaminergic structures of the neural reward circuits. The dopaminergic response further enhances encoding of information that is presented around its release by influencing dopamine-dependent cellular mechanisms of learning in the hippocampus. As such, extrinsic rewards/incentives and curiosity motivate and facilitate learning, illustrating their importance in educational contexts and knowledge acquisition. Taken together, their large overlap in neural response and behavioural effects suggests that both may be supported by common neural processes. However, this implies that their combined use would be associated with sub-additive effects. On the other hand, if both were supported by differential neural effects, they could be used in an additive manner. Importantly, the question of how extrinsic rewards/incentives and curiosity interact in their effects on behaviour and cognition overall and memory in particular can only be answered if both effects are studied in conjunction rather than individually as often done in previous research. Another limitation stems from the way how studies thus far have investigated the effects of curiosity on memory, and in some cases, its interaction with extrinsic rewards/incentives, not only because they nearly exclusively all use the same paradigm, but more so because the paradigm itself has some inherent limitations that might affect how curiosity is conceptualised. The present work tries to address these gaps in the literature. In doing so, a new paradigm – the magic trick paradigm – was developed, in which curiosity and the availability of extrinsic incentives were manipulated to measure their effects on encoding. In the magic trick paradigm, curiosity was elicited using short videos of magic tricks. Participants engaged in an orientation task combined with ratings of the “subjective feelings of curiosity” and performance therein was incentivised using a between-subject design. Unbeknown to the participants, their memory for the magic tricks was tested a week later. Crucially, after behavioural pilots, the paradigm was adopted for usage with functional magnetic resonance imaging (fMRI) to be able to investigate the neural underpinnings of incentive- and/or curiosity-motivated incidental learning during encoding as well as early consolidation. To the best of our knowledge, the associated fMRI dataset – the Magic, Memory, and Curiosity (MMC) Dataset – is the first of its kind, making it highly valuable to the nascent field investigating the effects of curiosity on memory because (1) fMRI data was acquired during the magic trick paradigm, but also before and after, allowing to study neural mechanisms underlying encoding as well as early consolidation, and (2) videos of magic tricks as dynamic stimuli allow for a plethora of analysis approaches to answer myriads of research questions. Chapter 2 describes the methods and procedures used to generate the MMC Dataset (N = 50), presented in a way that allows independent researchers to re-use it according to their needs. Additionally, high data quality comparable to other openly available datasets in the field has been demonstrated by performing data quality assessments and basic validation analysis. This further lays the groundwork for Chapters 3 and 4 where the fMRI data acquired during encoding and consolidation, respectively, will be used. In Chapter 3, a meta-analytical approach was used to analyse the behavioural data from three studies (two behavioural studies and one fMRI study) using the magic trick paradigm to investigate the effects of curiosity, the availability of extrinsic incentives, and their interaction on memory. The main memory outcome was high-confidence recognition, a recollection-based memory measurement, but other indices were also examined to derive a more detailed picture. This revealed positive effects of curiosity and monetary incentives on encoding, in the absence of interaction effects. Exploratory analyses further showed that curiosity and monetary incentives might impact encoding differently, overall suggesting that the effects might be at least partially non-overlapping. Analysing the fMRI data acquired during the presentation of magic tricks using the intersubject synchronisation framework to account for the dynamic nature of the stimuli, we found that while the effects of curiosity on memory were located in the hippocampus and dopaminergic brain areas, neither the effects of curiosity nor incentives themselves were found in the often-implicated reward network, but instead were associated with regions involved in processing uncertainly and attention. Likewise, the effects of curiosity on memory spread further across broad cortical and subcortical networks. Overall, this suggests that the subjective feeling of curiosity and its effects on memory recruits broad brain networks when investigated with dynamic stimuli, caveating a too narrow focus on a small list of regions-of-interest while there is yet so much more to be learned about the effects of curiosity on memory. In Chapter 4, resting-state data acquired before and after learning was used to investigate changes in brain activity at rest following learning. The pre-learning rest data can be used as a baseline, allowing any changes from pre- to post-learning to be attributed to the learning experience itself. Because previous research has repeatedly pointed to similarities between extrinsic rewards/incentives and curiosity, our analysis focused on the change in resting-state functional connectivity between the dopaminergic midbrain and the anterior hippocampus, a dopaminergic consolidation mechanism previously reported in the context of extrinsically motivated learning. Contrary to our hypothesis, we did not find an overall change nor that individual differences therein predicted behavioural measures of learning. However, brain-behaviour correlations differed significantly depending on the availability of extrinsic incentives. In sum, this suggests that curiosity-motivated learning might be supported by different consolidation mechanisms compared to extrinsically motivated learning and that extrinsic motivation could re-configure resting-state networks supporting early consolidation. Overall, this work adds to the literature by replicating the effects of curiosity on encoding. More importantly, however, this work suggests that the systems supporting extrinsically and curiosity-motivated learning might differ more than previously assumed, especially when investigating activity across the whole brain rather than focusing on a priori candidate regions implicated in dopaminergic effects. Indeed, our results allow for the possibility that other neurotransmitter play a role as well in extrinsically and curiosity-motivated learning, further highlighting the need for more research in the area.