Presentation
Search Abstracts | Symposia | Slide Sessions | Poster Sessions | Lightning Talks
Cortical and subcortical mechanisms of orthographic learning
Poster A126 in Poster Session A, Tuesday, October 24, 10:15 am - 12:00 pm CEST, Espace Vieux-Port
Also presenting in Lightning Talks A, Tuesday, October 24, 10:00 - 10:15 am CEST, Auditorium
Brenda Rapp1, Yuan Tao1, Teresa Schubert2, Robert Wiley3, Craig Stark4; 1Johns Hopkins University, 2Springer Nature, 3University of North Carolina, Greensboro, 4University of California, Irvine
Introduction. Research on language learning has been carried out largely independently from basic research on learning and memory. However, their greater integration may be valuable. The learning/memory literature offers increased understanding of mechanisms of encoding, memory formation, consolidation and retrieval, while language provides a learning domain that is far richer, more complex and more highly integrated within existing knowledge structures than those typically considered in learning/memory research. The Complementary Learning Systems framework (McClelland, et al., 1995) has been highly influential in the learning/memory literature. CLS theory proposes that hippocampal and neocortical regions instantiate different types of learning systems performing different computations, with contributions that vary over the time-course of learning. While there has been some research on spoken word learning within the CLS framework, research on orthographic learning (in reading and dyslexia) has largely considered cortical mechanisms. In this study we examined the neural substrates of orthographic learning in real time, focusing on neocortical and hippocampal contributions. To do so, we adapted the Law et al. (2005) learning paradigm requiring participants to learn the relationship between the pronunciations and spellings of pseudo-words, while undergoing fMRI scanning. Thirteen healthy volunteers participated in a forty-minute fMRI scanning session during which they learned the spellings of pseudowords that were repeatedly presented to track the trajectory of in-scanner orthographic learning. Spelling accuracy of the items was evaluated immediately post scanning and one week later. In the hippocampus and the left ventral occipital cortex (LVOT; associated with orthographic processing), we analyzed BOLD response for learning trials that differed in their memory strength (weak, medium, strong; Smith et al., 2004). We found: (1) Both left and right hippocampi showed that activation increased with memory strength (LH: p=0.0021, RH: p=0.0002), whereas the LVOT showed decreased activation with memory strength (p=0.03); (2) Higher post-scan accuracy was correlated with greater activation changes of the hippocampus (r=0.61, p=0.01), but not LVOT (r=0.38, p=0.11); (3) Whole brain analyses identified two bilateral sets of brain areas that exhibited either the increasing or decreasing patterns exhibited by the hippocampus and the LVOTC, with independent functional connectivity analyses indicating that these two sets of brain areas functioned as two internally synchronized, integrated networks, at least during initial stages of learning; 4) The degree of internal synchronization of these two networks was associated with accuracy of recall up to one week after training. The results are consistent with the hypothesis that two distinct neurocognitive mechanisms are involved in orthographic learning. Generally consistent with the CLS framework, the hippocampus specifically contributes to better learning outcomes, while the decreased activation observed in LVOT orthographic processing regions may reflect BOLD adaption/familiarization that, at least in early in learning, does not support long-term retention.
Topic Areas: Reading, Writing and Spelling