The review provides an analysis of modern cognitive psychological theories of emotions and neural basis of mechanisms standing behind retention of emotional information in working memory. The tendency seen in modern neurocognitive theories is characterized by the rejection of localization for any psychological functions in favor of network models, where higher-order processes are supported by distributed neural networks connecting functionally specific cortical and subcortical structure formations. Such networks are supposed to be able to reorganize themselves in accordance with goal-oriented behavior and task requirements. We analyze existing approaches to determine the cognition and emotion interactions, within which they can be considered both as competing systems and the components of a single mechanism of goal-oriented behavior, for example in case of affective working memory. We also provide a hypothesis for the potential neural basis of affective working memory based on large-scale distributed networks including saliency network, default mode network, and the frontoparietal network, considering the possibility of dynamic changes in the system and obtaining additional nodes, for example, associated with the processing of social information.
E. V. Kochetkova (Wed,) studied this question.