Nature, Stability and Integration of Associations Acquired by Means of Fast Mapping

• Applicants: Dr. Regine Bader; Professor Dr. Patric Meyer
• Subject Area: General, Cognitive and Mathematical Psychology ; Human Cognitive and Systems Neuroscience
• Term: since 2015
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 266464799

Contrary to traditional theories, it has been shown that novel associations can be rapidly integrated into cortical networks through a learning paradigm called fast mapping (FM), possibly bypassing hippocampal consolidation processes. In the FM paradigm, an unknown item is presented next to a known item together with a question referring to an unfamiliar label. By recognizing the known item, one implicitly infers that the label must belong to the unknown item. Thereby, the unknown item and the label are bound to an association, which may be ultimately incorporated into cortical networks. However, there is still a gap in current knowledge on the factors driving rapid cortical integration through FM. Most of the current approaches are contradictory or difficult to interpret. Difficulties in obtaining reliable data have even led to the assumption that “the FM phenomenon does not exist”. A new approach that takes into account the nature and complexity of the associations once they have been integrated into cortical networks could provide groundbreaking incremental knowledge in this research area. In our previous studies, we found that FM allows for the rapid integration of novel associations into cortical networks and that this can be promoted by a high feature overlap and incidental learning. Also, fMRI data showed that the perirhinal cortex (PrC) supports the binding of novel picture-label associations when they are encoded through FM. This was predicted based on the functional characteristics of the PrC, which is not only involved in binding and storing single items or unitized representations but is also especially qualified to discriminate between highly similar objects. Consequently, we suggested that triggering PrC processing by increasing the demands on object discrimination during FM should foster the binding of the unknown item to the label and their rapid cortical integration. Importantly, PrC binding during learning should lead to very specific representational characteristics of the newly-built association. Hence, insights on the quality of these memory representations would help to find one central missing piece in the debate on FM learning. The nature of information processed in the PrC is significantly different from representations processed in the hippocampus in terms of their complexity, flexibility, stability, and their way of integration into existing networks. In particular, PrC representations should be characterized and distinguished from hippocampal representations by being acontextual, recognizable by familiarity, inflexible, prone to a specific pattern of forgetting due to decay and interference, and cortically integrated. Therefore, the objective of this project is to elucidate the nature of associations integrated by means of FM in a series of complementary experiments using EEG and behavioral indices of cortical integration and memory processes.

• DFG Programme: Research Grants