Cognitive and Neural Computations of Semantics

Descripción del proyecto

La paradoja de la congruencia semántica en la cognición

En nuestro día a día, dependemos de forma constante de las relaciones entre los elementos de nuestro entorno (lo que se denomina «información semántica») para interactuar de forma adecuada con el mundo. Esa información puede simplificar la comprensión reforzando patrones familiares o captar la atención destacando elementos inesperados. El reto radica en la característica dual de esa información, que plantea interrogantes sobre cómo y cuándo el cerebro recurre al conocimiento semántico almacenado. En el proyecto CONNECTS, financiado por el Consejo Europeo de Investigación, se examinará esta paradoja cognitiva con un marco unificado basado en el principio de transferencia de procesamiento adecuado. A través del estudio de representaciones neuronales, se investigará cómo los cálculos cognitivos se alinean con la congruencia semántica. En último término, el proyecto tiene como objetivo conciliar las pruebas contradictorias en todos los ámbitos cognitivos con una solución integradora.

Objetivo

In our everyday lives, we rely on existing relations among elements in our environment (i.e. semantic information) to interact efficiently with the world. This information can either be used to facilitate understanding by exploiting redundant (congruent) evidence or to signal out salient stimuli by highlighting unexpected (incongruent) elements. This duality raises fundamental questions about how and when our brains utilize stored semantic knowledge as its influence seems to vary depending on the specific cognitive domain. This seemingly paradoxical state represents a cognitive puzzle that questions whether the presence of (in)congruent contextual information in a given situation has a positive or negative impact on how we perceive, process, and remember information.

CONNECTS seeks to address the paradoxical effects of semantic congruity across various cognitive domains by providing a unified framework. The proposed framework builds on the Transfer Appropriate Processing principle and brings it to a neural representational level. By examining the transformations of neural representations, it is possible to quantify the degree of overlap in cognitive computations as a measure of appropriate transfer. Thus, CONNECTS dissolves the paradox by proposing that performance would be optimal when the required cognitive computation is oriented towards the same stimulus properties emphasized by the semantic (in)congruity of the stimulus. This proposal not only reconciles conflicting evidence on specific domains but also provides a domain-agnostic framing of the conundrum that ensures its integrative aim.

CONNECTS combines a solid theoretical foundation with cutting-edge neuroscientific techniques. The project's multi-method approach, including behavioural, neural (fMRI and EEG), and computational data from artificial neural networks. This approach offers a comprehensive exploration of the phenomenon which is a core requirement for a unifying framework.