Search Abstracts | Symposia | Slide Sessions | Poster Sessions
The N400: A comprehensive neurocognitive synthesis and new explanatory framework unifying processing and learning within neocortical-medial temporal-lobe circuitry
Poster Session A, Friday, September 12, 11:00 am - 12:30 pm, Field House
Michael T. Ullman1, João Veríssimo2, Joshua Buffington1, Robert V. Reichle3; 1Georgetown University, 2University of Lisbon, 3University of Texas at Austin
The N400 is a widely-studied brain response elicited by both language and non-language stimuli. N400s have generally been taken to reflect aspects of lexical and/or semantic processing. Building on previous work, we propose that N400s reflect both the processing and learning of lexical/semantic knowledge. Specifically, we suggest that the N400 is rooted in a processing-and-learning “what” circuit that links the temporal-lobe ventral stream with anterior portions of medial-temporal-lobe-based declarative memory, in particular in (peri)rhinal cortex and the anterior hippocampus. In this circuit, the ventral stream hierarchically processes (visual and/or auditory) inputs, posteriorly to anteriorly. The resulting representation is then passed to perirhinal cortex, which evaluates to what extent it matches already-existing representations, by retrieving that information from (mainly temporal) cortex. A mismatch (in content or strength) warrants updating/strengthening the existing representations, in which case perirhinal cortex passes the incoming representation downstream via (anterior) entorhinal cortex to the (anterior) hippocampus. The hippocampus then binds together novel and existing aspects of the relevant representations upstream in (mainly temporal) neocortex, particularly in the ventral stream(s) and middle/anterior temporal cortex. We posited that N400s are generated by various portions of this circuit. To test this framework and more generally to comprehensively examine the neurocognitive bases of N400s, we examined multiple lines of evidence. First of all, as expected by the framework, N400s are indeed found largely for stimuli that involve the “what” circuit and are potentially meaningful (related to already-learned representations, thus warranting their retrieval), especially when such stimuli are unfamiliar or unexpected (and thus they are likely to engage perirhinal cortex and lead to hippocampal learning). Conversely, N400s are typically not found when the input does not involve the “what” circuit, for example when the stimuli engage procedural memory (e.g., in Serial Reaction Time tasks, or for grammatical anomalies in native languages), or when they involve the “what” circuit but are not potentially meaningful (e.g., consonant strings like ‘qxmz’) or are more expected (e.g., semantically congruous) or familiar (e.g., high-frequency/real items). To directly examine the neuroanatomical substrates of the N400, we performed two meta-analyses, each based on systematic searches of the relevant literatures. First, to help identify which brain structures are involved in generating N400s, we performed a quantitative synthesis of structures (throughout the brain) implicated in generating N400s in direct (intracranial) brain recording. Second, to help reveal which brain structures are necessary for generating N400s, we performed a quantitative synthesis examining lesion-symptom (structure-N400) mappings; that is, we examined the degree to which damage in neuroanatomical structures (throughout the brain) across a range of disorders predicts N400 amplitude reductions in those disorders in comparisons with healthy controls. Both of these meta-analyses particularly implicated both middle/anterior temporal neocortex and the anterior medial temporal lobe, especially (peri)rhinal cortex. Overall, converging evidence supports the proposed framework positing that N400s are generated by the neocortical-medial temporal-lobe “what” circuit. Further, the evidence identifies exactly which structures within the circuit play crucial roles in generating N400s. We conclude by discussing basic research and translational implications.
Topic Areas: Meaning: Lexical Semantics,