Though activation of Broca's region in the combinatorial processing of symbols (language, music) has been revealed by neurometabolic studies, most previous neurophysiological research found the earliest grammar indices in the temporal cortex, with inferior-frontal generators becoming active at relatively late stages. We use the attention- and task-free syntactic mismatch negativity (sMMN) event-related potential (ERP) to measure rapid and automatic sensitivity of the human brain to grammatical information in participants' native language (French). Further, sources underlying the MMN were estimated by applying the Parametrical Empirical Bayesian (PEB) approach, with the Multiple Sparse Priors (MSP) technique. Results showed reliable grammar-related activation focused on Broca's region already in the 150-190 ms time window, providing robust documentation of its involvement in the first stages of syntactic processing.
Fast real-time processing of external information by the brain is vital to survival in a highly dynamic environment. A ubiquitous information medium used by humans is spoken language, but the neural dynamics of its comprehension is still poorly understood. Here, we scrutinized the earliest electrophysiological activity elicited in the human brain by spoken words and matched meaningless word-like stimuli using a lexical auditory oddball paradigm, an established technique for investigating cortical activation patterns underlying early automatic stages of language processing. We show that the earliest cortical reflection of word comprehension takes place during the electrophysiological P1 evoked response, at about 30 ms following the word disambiguation point, and takes the form of an enhanced brain activation for familiar meaningful words, even when they are presented outside the focus of attention. This previously unknown ultra-early lexicality effect is underpinned by left temporo-frontal cortical circuits and likely reflects a first-pass automatic lexical access that precedes later stages of lexical and semantic processing described in previous literature. The results suggest that the brain operates with maximum speed and efficiency to extract meaningful (including linguistic) information from the sensory input, which is a neurobiological capacity essential for timely and appropriate reactions to external events.
Despite extensive research on face recognition, only a few studies have examined the integration of perceptual features with semantic, biographical, and episodic information. In order to address this issue, we used repetitive transcranial magnetic stimulation (rTMS) to target the left inferior frontal gyrus (IFG) and the left occipital face area (OFA) during a face recognition task. rTMS was delivered during the encoding of "context" faces (i.e., linked to an occupation, e.g., "lawyer") and "no-context" faces (i.e., linked to a nonword pattern, e.g., "xxxx"). Subjects were then asked to perform a recognition memory task. Accuracy at retrieval showed a mild decrease after left OFA stimulation, whereas rTMS over the left IFG drastically compromised memory performance selectively for no-context faces. On the other hand, absence of rTMS interference on context faces might be due either to the fact that pairing an occupation to a face makes the memory trace stronger, therefore less susceptible to rTMS interference, or to a different functional specificity of the left IFG subregions.