Perseverative and Semantic Influences on Visual Object Naming Errors in Optic Aphasia: A Connectionist Account

Abstract

Although perseveration-the inappropriate repetition of previous responses-is quite common among patients with neurological damage, relatively few detailed computational accounts of its various forms have been put forth. A particularly well-documented variety involves the pattern of errors made by "optic aphasic" patients, who have a selective deficit in naming visually presented objects. Based on our previous work in modeling impaired reading via meaning in deep dyslexia, we develop a connectionist simulation of visual object naming. The major extension in the present work is the incorporation of short-term correlational weights that bias the network towards reproducing patterns of activity that have occurred on recently preceding trials. Under damage, the network replicates the complex semantic and perseverative effects found in the optic aphasic error pattern. Further analysis reveals that the perseverative effects are strongest when the lesions are near or within semantics, and are relatively mild when the preceding object evokes no response. Like optic aphasics, the network produces predominantly semantic rather than visual errors because, in contrast to reading, there is some structure in the mapping from visual to semantic representations for objects. Viewed together with the dyslexia simulations, the replication of complex empirical phenomena concerning impaired visual comprehension based on a small set of general connectionist principles strongly suggests that these principles provide important insights into the nature of semantic processing of visual information and its breakdown following brain damage.

Relearning after Damage in Connectionist Networks: Toward a Theory of Rehabilitation

“…Much of this work has been focused on deficits in word reading-the acquired dyslexias (Mozer & Behrmann, 1990;Patterson, Seidenberg, & McClelland, 1990;Plaut & Shallice, 1993a)-but there has also been considerable work in other domains, including spelling (Shallice, Glasspool, & Houghton, 1995;Olson & Caramazza, 1994), speech production (Harley & MacAndrew, 1992;Martin, Dell, & Schwartz, 1994), face recognition (Burton, Young, Bruce, Johnston, & Ellis, 1991;Farah, O'Reilly, & Vecera, 1993), visual object naming (Gordon, 1982;Plaut & Shallice, 1993b), spatial attention (Cohen, Romero, Servan-Schreiber, & Farah, 1994;Humphreys, Freeman, & Müller, 1992), learning and memory McClelland, McNaughton, & O'Reilly, 1995), semantic memory (Farah & McClelland, 1991;Horn, Ruppin, Usher, & Hermann, 1993), and control of action and responding (Bapi & Levine, 1990;Cohen & Servan-Schreiber, 1992;Levine & Prueitt, 1989). Although still in its infancy, the relative success of this work suggests that connectionist modeling may provide an appropriate formalism within which to explore how disorders of brain function give rise to disorders of cognition.…”

Spelling-to-sound correspondences affect acronym recognition processes

“…However, low frequency 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 words take longer to be recognized and therefore they are able to benefit recognition from semantic activation. Plaut and Shallice (1993) argued that highly imageable words have richer semantic representations than low imageable words. Thus low frequency but high imageable words receive more semantic input than low imageable words, and recognition times are faster as a result.…”

Mechanisms underlying perseveration in aphasia: evidence from a single case study

“…Recent theoretical accounts developed from psychological and neuropsychological perspectives have suggested that recurrent perseverations result from the same mechanisms that give rise to priming effects in normal subjects ( [12,37,73]; see [56] for a similar view). For example, Cohen and Dehaene [12] have proposed a general information processing account that is capable of addressing a broad range of empirical effects associated with priming and perseveration.…”

“…Although these errors did not always change along with perseverations, it is possible that the mechanisms that underlie the different error types are partially shared. For example, one might view an omission as reflecting an elicited pattern of neural activity that is too different from any known state to support a response (see [56] for a similar view). Under a cholinergic deficit, sensory input might partially overcome persistent activity, yielding a novel pattern of activity that either produces no response (an omission) or a response blend (a neologism or other error).…”

“…Recent theories developed from psychological and neuropsychological perspectives have attempted to account for the existence of recurrent perseverations by appealing to the same mechanisms proposed to underlie priming effects in normal subjects ( [12,37,73]; see also [56]). One such theory, proposed by Cohen and Dehaene [12], holds that cognitive representations at many processing levels commonly persist for some period of time following initial activation.…”