Открытые лекции нейропсихолога Тимоти Шаллиса

Профессор Шаллис будет рассказывать о латерализации функций в префронтальной коре головного мозга человека, о когнитивном контроле и мозговых механизмах решения инсайтных задач. Ниже приводятся аннотации лекций.

 Lecture 1: Neuropsychology and the lateralisation of prefrontal functions

 The model I developed many years ago with Don Norman was essentially designed as an information-processing realisation of Alexander Luria’s approach to the operation of prefrontal cortex as the system for the programming, regulation and verification of activity. Within that framework, can one unpack the innards of a prefrontal Supervisory System? The lecture will examine the contrasting functions of the left and right lateral frontal cortices, and to a lesser extent those of the medial and anterior (Brodmann area 10) frontal cortices, from the perspective of the effects of lesions to frontal cortices. Imaging investigations of the hierarchical internal structure of action control in prefrontal cortex (Koechlin) and critical so-called multiple demand systems (Duncan) tend to suggest bilateral involvement. By contrast, a variety of neuropsychological investigations provide evidence of some lateralisation of function. There is an overall tendency for verbal -related processes to be more left lateralised visuospatial -related processes to be somewhat more right lateralised. In addition, I will argue that abstraction, problem solving in small tight well-formed problem spaces, and task or criterion setting are relatively left lateralised. By contrast, active monitoring and construction of a strategy in ill-formed problem spaces tend to be right lateralised. Examples will be drawn from both reaction time studies and problem-solving ones.

 Lecture 2: Is there an underlying mechanistic account for the differential lateralisation of prefrontal neuropsychological processes?

 A brief account will be given of the different types of models that have been used in cognitive neuropsychology. Two main types have been employed. Initially classical information-processing ones were used. However, they have tended to be replaced by standard connectionist or neural net ones. This will be illustrated from disorders in the domains of memory, reading and spelling. It will be argued, however, that connectionist models of particular domains need to be tweaked by the addition of specific types of components or properties so as to fulfil the functions explicitly assumed in the first type of model. The example will be given of competitive queueing models in the domains of short-term memory and spelling. Returning to the issue of the lateralisation of lateral frontal functions, discussed in lecture 1, what could be the “tweak” in that domain? It will be argued that the degree of lateral inhibition between alternative schemas or links may be a key difference between the two lateral prefrontal cortices. Functions that tend to be lateralised to the left are ones that require strong selection between alternatives, for instance in the creation of a hierarchical structure and so must be built on winner-take-all neural nets. By contrast, ones that tend to be lateralised to the right require relatively parallel processing across alternatives. An example of the first type will be given from investigations of dynamic aphasia arising from left inferior frontal lesions. An example of the second type will be given from an imaging study of difficult problem-solving that requires insight following multiple unsuccessful attempts.