Saltar para: Posts [1], Pesquisa e Arquivos [2]

"The role of prediction in social neuroscience"


It is clear that perception is not just a merely reactive process in which sensory input is received and registered, but is more likely a construction of predicted representations of the environment. The concept of generative internal models in perception was proposed many years ago, and represents another common principle that is at the basis of some predictive coding frameworks. 


During motor control, the matching process between the generative forward model of the planned action and the sensory consequence of the action is thought to be the neural basis for distinguishing between self-generated and externally generated motor actions . Predictive coding in visual processing has been implemented in various hierarchical models that generally propose an integration of top-down expectancies and bottom-up sensory input reflecting stimulus information. Feedforward connections carry error signals, but in addition, cortical feedback connections transmit expectancy biases or predictions. A similar matching process is thought to occur during the execution of motor actions, and has been argued to be the underlying mechanism that maintains a sense of agency or ownership over our own motor actions. A mismatch, i.e., a discrepancy between the predictive forward model and the actual sensory feedback, can have pathological consequences, particularly relevant to schizophrenia and pathological abnormalities in ownership of action. This discrepancy between the expected and actual sensory outcome of a visual percept or motor action can also be referred to as a prediction error.


Social interaction and social functioning involves a multitude of socially relevant cognitive processes including, to name a few, social perception, understanding others' actions, observational social learning, social decision-making, and empathy. Top-down influences of social information can directly drive how we process visual information. More evidence is emerging which suggests that a similar mechanism, or “shared neural representation” is used for understanding others' actions, whereby an internal model of others' actions allows us to make predictions about the consequence and outcome of an observed action, and consequently understand and interpret the goals and intentions of the action. Many authors have suggested how conceptualizations of fundamental cognitions such as learning, could be extended to explain mental processes required for social understanding, social interaction, and social learning. There is also substantial work to indicate that there is top-down influence of social information and social interaction on fundamental error processing, learning, and decision-making processes.


Autoria e outros dados (tags, etc)


Pesquisar no Blog

subscrever feeds