Showing papers by "Tim Shallice published in 1994"

Brain regions associated with acquisition and retrieval of verbal episodic memory

Abstract: It is widely held that conscious recall of past experiences involves a specific system--episodic memory. Patients with amnesia have gross impairments of episodic memory while other kinds of memory remain intact, suggesting that a separable brain system underlies episodic memory. We have used positron emission tomography (PET) to identify components of this system in normal volunteers. A dual-task interference paradigm was used to isolate brain areas associated with acquisition, and a cueing paradigm to isolate the areas concerned with retrieval from verbal episodic memory. Acquisition was associated with activity in the left prefrontal cortex and the retrosplenial area, whereas retrieval was associated with activity in right prefrontal cortex and the precuneus. Our results provide clear evidence that episodic memory involves a network of specific prefrontal and posterior structures which can be fractionated into different component processes.

Multiple levels of control processes

Abstract: That a useful processing distinction can be made between the operation of the relatively permanent structural features of the cognitive system and processes briefly set up by subjects to attain a particular goal has been generally accepted since it was proposed by Atkinson and Shiffrin (1968). There has, however, been far less agreement on how control processes should be characterized. Early theorists saw them as the province of a single system or processing mode. Three later theories, those of Schneider, Newell. and Norman and Shallice, take more complex positions. These theories are reviewed in the light of two main sources of evidence: standard human experimental (in particular, dual-task performance) and neuropsychological (particularly studies of Parkinson's-disease and frontal-lobe lesion patients).

Connectionist Modelling in Cognitive Neuropsychology: A Case Study

Abstract: Computational models offer tools for exploring the nature of human cognitive processes. In connectionist, neural network, or parallel distributed processing models, information processing takes the form of cooperative and competitive interactions among many simple, neuron-like processing units. These models provide new ways of thinking about the neural basis of cognitive processes, and how disorders of brain function lead to disorders of cognition. This monograph is an expanded version of a recent issue of the journal Cognitive Neuropsychology. It presents the most comprehensive existing "case study" of how the effects of damage in connectionist models can replicate the detailed and diverse patterns of cognitive impairments that can arise in humans as a result of brain damage. It begins with a review of the basic methodology of cognitive neuropsychology and of other attempts at modeling neuropsychological phenomena. It then focuses on a particular form of acquired reading disorder, "deep dyslexia," in which previously literate adults with brain damage exhibit a wide range of symptoms in pronouncing written words, the most striking of which is the production of semantic errors (e.g. reading RIVER as "ocean"). A series of simulations investigate the effects of damage in connectionist models that pronounce written words via their meaning. The work systematically explores each main aspect of the design of the models, identifying the basic computational properties that are responsible for the occurrence of deep dyslexia when the models are damaged. Although the investigation concerns a specific form of reading impairment, the computational principles that emerge as critical are very general ones: representation of concepts as distributed patterns of activity, encoding of knowledge in terms of weights on connections between units, interactivity between units to form stable attractors for familiar activity patterns, and greater richness of concrete vs. abstract semantics. The fact that damage to models embodying these principles and damage to the brain can produce strikingly similar behaviour supports the view that the human cognitive system operates according to similar principles.

Fractionation of the frontal-lobe syndrome

Abstract: It is argued that the severity of the cognitive problems seen following frontal damage may reflect the number of different individual processes affected rather than just overall severity of impairment in one core process. A number of recent studies are described which seem to suggest that a fractionation of the frontal lobe syndrome may be possible. However the considerable methodological problems involved are discussed, and it is argued that approaches not commonly used in neuropsychology may be required.

Word Reading in Damaged Connectionist Networks: Computational and Neuropsychological Implications

Abstract: Connectionist networks are also called neural networks because of their abstract structural similarity to groups of neurons. Based on this similarity, many researchers believe that computation in these networks reflects important properties of neural computation. One piece of evidence often put forward in support of this claim is that, like brains, connectionist networks tend to degrade gracefully with damage. That is, if some proportion of units and/or connections are removed from a network, performance on a task is typically only partially impaired rather than completely abolished. Most demonstrations of graceful degradation in networks have used only very general measures of performance, such as total error on a task. However, the argument that connectionist computation is fundamentally similar to neural computation would be far more compelling if the way in which connectionist networks degraded under damage—their patterns of impaired performance—mirrored the patterns of impaired behavior observed in patients with neurological damage. To the extent that this held, a detailed investigation of the behavior of damaged connectionist networks would provide insight into both normal and impaired human cognition. A complementary motivation for studying the effects of damage in networks is to extend our understanding of the nature of computation in the networks themselves. Here again, our concern is not just with the development of a network that accomplishes a task, but with understanding how the network accomplishes the task—the nature of its representations and processes. In most connectionist research, the adequacy of a network is evaluated by testing how well its performance generalizes to novel external input drawn from the same distribution as the training examples. In a