Pilar Casado

Semantics prevalence over syntax during sentence processing: A brain potential study of noun–adjective agreement in Spanish

A review of the literature about the interplay of syntax and semantics, using event-related brain potentials (ERPs), revealed that the results are highly heterogeneous, owing to several possible variables. An experiment was conducted with Spanish sentences that factorially combined syntactic and semantic violations in the same sentence-intermediate adjective and controlled for working memory demands, variables that in previous studies have rarely been taken into consideration. Violations consisted in noun-adjective number or gender disagreements (syntactic violation), noun-adjective semantic incompatibility (semantic violation), or both (combined violation). The N400 to semantic violations was unaffected by additional syntactic violations. The P600/SPS component, considered to reflect syntactic processes, was elicited by both single syntactic and semantic violations but seemed to be diminished in combined violations relative to single syntactic violations. These results suggest that under the conditions of the present experiment semantic information may have a prevailing role over syntactic information.

Brain activation in discourse comprehension: a 3t fMRI study.

To date a very small number of functional neuroimaging studies have specifically examined the effects of story coherence on brain activation using long narratives, a procedure fundamental to the study of global coherence. These studies, however, not only yielded notably divergent results, but also featured a number of caveats. It is the purpose of the present study to try to overcome some of these limitations. A left precuneus/posterior cingulate activation related to global coherence comprehension was in consonance with a part of previous literature. However, our most important results corresponded to left parietal regions (angular gyrus, BA 39), this diverging from the previous studies. Recent developments of the situational models of narrative comprehension could explain all these apparently inconsistent results. According to these, different situation models would be created as a function of the content of the narratives, which would yield in turn different patterns of brain activity. Our data also suggest that the same content might also give place to different situation models as a function of the degree of global coherence achieved by the reader or listener.

Electrophysiological Evidence of Automatic Early Semantic Processing.

This study investigates the automatic-controlled nature of early semantic processing by means of the Recognition Potential (RP), an event-related potential response that reflects lexical selection processes. For this purpose tasks differing in their processing requirements were used. Half of the participants performed a physical task involving a lower-upper case discrimination judgement (shallow processing requirements), whereas the other half carried out a semantic task, consisting in detecting animal names (deep processing requirements). Stimuli were identical in the two tasks. Reaction time measures revealed that the physical task was easier to perform than the semantic task. However, RP effects elicited by the physical and semantic tasks did not differ in either latency, amplitude, or topographic distribution. Thus, the results from the present study suggest that early semantic processing is automatically triggered whenever a linguistic stimulus enters the language processor.

Similarities and differences between phrase structure and morphosyntactic violations in Spanish: An event-related potentials study

Event-related potentials (ERPs) were employed to compare word category and verb inflection violations in Spanish. A similar frontal negativity was found between 250-400 ms for both violation types, suggesting that they equally disrupt initial syntactic analyses. Also, word category violations elicited a negativity at posterior electrodes larger than verb inflection anomalies and correct sentences within this time window, probably reflecting difficulties to semantically analyse phrase structure violations. Finally, a centroparietal positivity in the 500-700 ms interval was found for both error types. Both violations did not differ along the first half of this interval (500-600 ms), but larger effects for verb inflection violations were reported along the second half of this period (600-700 ms). These findings suggest that whereas processes dealing with the reanalyses of sentence structure are equally triggered by both anomaly types, further attempts of repairing the structure of sentences occur only with verb inflection violations.

Electrophysiological evidence of an early effect of sentence context in reading.

Recognition Potential is an electrophysiological response of the brain that is sensitive to semantic aspects of stimuli. According to its peak values (about 250ms), Recognition Potential appears as a good candidate to reflect lexical selection processes. Consequently, Recognition Potential might be sensitive to contextual information during reading a sentence. In present study, the standard procedures to improve the visibility of Recognition Potential (Rapid Stream Stimulation paradigm) were used in a task in which sentence context was crucial. A parieto-occipital Recognition Potential was observed to peak about 264ms after stimulus onset, followed by a centro-parietal N400 peaking at about 450ms. Recognition Potential was affected by contextual information though, contrary to N400, presenting larger amplitude to contextually congruous words. These results support the assumption that Recognition Potential may reflect lexical selection processes, representing also evidence of context effects on ERP around 250ms after stimulus onset during sentence reading.

The influence of emotional words on sentence processing: electrophysiological and behavioral evidence.

Whereas most previous studies on emotion in language have focussed on single words, we investigated the influence of the emotional valence of a word on the syntactic and semantic processes unfolding during sentence comprehension, by means of event-related brain potentials (ERP). Experiment 1 assessed how positive, negative, and neutral adjectives that could be either syntactically correct or incorrect (violation of number agreement) modulate syntax-sensitive ERP components. The amplitude of the left anterior negativity (LAN) to morphosyntactic violations increased in negative and decreased in positive words in comparison to neutral words. In Experiment 2, the same sentences were presented but positive, negative, and neutral adjectives could be either semantically correct or anomalous given the sentence context. The N400 to semantic anomalies was not significantly affected by the valence of the violating word. However, positive words in a sentence seemed to influence semantic correctness decisions, also triggering an apparent N400 reduction irrespective of the correctness value of the word. Later linguistic processes, as reflected in the P600 component, were unaffected in either experiment. Overall, our results indicate that emotional valence in a word impacts the syntactic and semantic processing of sentences, with differential effects as a function of valence and domain.

Dorsolateral prefrontal cortex contribution to abnormalities of the P300 component of the event-related potential in schizophrenia

Abnormalities of the P300 component of the event-related potential are a common finding in schizophrenia. It seems possible that the dysfunction in the dorsolateral prefrontal (DLPF) region that has been reported in schizophrenia contributes to this finding. To explore this possibility, we calculated the relationship between, on the one hand, P300 latency and amplitude and, on the other hand, the degree of DLPF atrophy (as measured by magnetic resonance imaging) and metabolic activity during an attentional task (as measured by positron emission tomography). Seventeen schizophrenia patients with a brief duration of illness and minimal exposition to treatment and 25 healthy controls were studied. Patients exhibited significantly lower metabolic activity in the DLPF region, but they did not show cortical atrophy. P300 amplitude was also significantly reduced in the schizophrenia patients compared with the controls. Right DLPF region metabolic activity correlated significantly with P300 amplitude. This pattern remained after partialling out the influence of activity in the hippocampus, superior temporal gyrus and parietal lobe. It is therefore suggested that the prefrontal cortex could be implicated in the P300 amplitude reduction in schizophrenia.

Semantic processing of open- and closed-class words: an event-related potentials study.

Previous research on open- and closed-class words has revealed the existence of several differences in the processing of these types of vocabulary. In this paper the processing of open- and closed-class words was compared by means of an early electrical brain response, recognition potential (RP), which indexes semantic processing and originates from basal extrastriate areas. The effects of word frequency on closed-class words were also investigated. For these purposes, open- and closed-class words, among other stimuli, were presented by means of the rapid stream stimulation procedure. Results showed that there were no significant differences when comparing the RP evoked by open- and closed-class words in the left hemisphere. However, in the right hemisphere this situation changed: the RP evoked by open- and closed-class words did differ. Moreover, there were no differences between the RP evoked by closed-class words and pseudowords. These patterns of results suggest that the semantic processing of closed-class words shares some aspects with the processing of open-class words, despite the existence of some differences. Thus, whereas the semantic processing of open-class words recruits brain areas of both hemispheres, the semantic processing of closed-class words is left-lateralized. A second purpose of this work is to study word-frequency effects on closed-class words. Our results show the insensitivity of closed-class words to word-frequency effects.

Spatial distribution and cognitive correlates of gamma noise power in schizophrenia.

BACKGROUND Brain activity is less organized in patients with schizophrenia than in healthy controls (HC). Noise power (scalp-recorded electroencephalographic activity unlocked to stimuli) may be of use for studying this disorganization. Method Fifty-four patients with schizophrenia (29 minimally treated and 25 stable treated), 23 first-degree relatives and 27 HC underwent clinical and cognitive assessments and an electroencephalographic recording during an oddball P300 paradigm to calculate noise power magnitude in the gamma band. We used a principal component analysis (PCA) to determine the factor structure of gamma noise power values across electrodes and the clinical and cognitive correlates of the resulting factors. RESULTS The PCA revealed three noise power factors, roughly corresponding to the default mode network (DMN), frontal and occipital regions respectively. Patients showed higher gamma noise power loadings in the first factor when compared to HC and first-degree relatives. In the patients, frontal gamma noise factor scores related significantly and inversely to working memory and problem-solving performance. There were no associations with symptoms. CONCLUSIONS There is an elevated gamma activity unrelated to task processing over regions coherent with the DMN topography in patients with schizophrenia. The same type of gamma activity over frontal regions is inversely related to performance in tasks with high involvement in these frontal areas. The idea of gamma noise as a possible biological marker for schizophrenia seems promising. Gamma noise might be of use in the study of underlying neurophysiological mechanisms involved in this disease.

Long-Term Olanzapine Treatment and P300 Parameters in Schizophrenia

The well-known amplitude reduction of the P300 appears to be unaffected by the treatment with classical antipsychotics in schizophrenia, whereas the effects of atypical neuroleptics on this event-related potential are less understood. The study of these changes could help in deciding whether the P300 amplitude reduction in schizophrenia is a trait or state marker of that illness and in better describing the effect of atypical antipsychotics on altered cognitive functions. We present a prospective longitudinal study of P300 amplitude and latency before and after 6 months’ treatment with olanzapine in 11 patients with schizophrenia. A healthy control group (n = 30) was also studied. Overall, no significant changes, either in amplitude or in latency as measured at Pz and Fz electrodes, were found when comparing the pre- and postolanzapine conditions, despite the overall improvement in positive and negative symptoms. Nevertheless a direct specific association was observed between a P300 amplitude increase with olanzapine and the improvement in negative symptoms. These data would suggest that P300 amplitude reduction in schizophrenia may be relatively independent from clinical state and treatment, thus constituting a trait marker of schizophrenia. Our data also suggest that, in addition to this, some further changes in P300 amplitude might depend on the clinical state of the patients.