The human ability to look at and label objects, though taken for granted even in very young children, has long puzzled and fascinated scientists investigating the organization of knowledge systems in the brain. A central question in this investigation is the way knowledge of objects are categorized, encoded, and stored within the neural network. This study of the "semantic organization" of the cortex has relied in large part on patients who, through the damaging effects of stroke, disease, or physical injury, have cerebral injuries which impair their faculties of recognition and object identification. After extensive testing with these patients, researchers have arrived at varying conclusions. Some hold that the deficits point to a category-specific semantic organization, others suggest a modality-specific structure, and still others claim the deficits are simply generalized cognitive impairment. Yet, many of these are at best, partial explanations and at worst, misinterpretations of observations.
[...] Yet, this scheme of the brain's semantic structure offers no explicit reason why this should be so. These logical problems are why category-specificity is not the only neural framework offered by scientists. In contrast to the idea of category specificity in the brain's semantic organization is the theory of modality-specificity. This theory posits that the agnosia of these patients is purely a function of the visual properties of the stimulus and not the categorical or semantic properties. For example, in an experiment where both auditory and visual identification were observed, “there was an insignificant degree of consistency for comparison between modalities In contrast, there was a significant degree of response consistency for all responses within a modality (Warrington & Shallice, 1984). [...]
[...] As important as this faculty is to an organism's survival, it is surprising that the brain does not recruit even more cortex for the task. The more cortex devoted to the lexical nodes and the systems they subserve, the more function is preserved in the case of a lesion. Finding a parsimonious, economical model for the brain's semantic organization seems almost paradoxical in this sense. In contrast to the category-specific, modality-specific, and visual- difficulty hypotheses, the “neural nodes” theory of semantic organization as stated by Damasio and others reconciles many problems quite well. [...]
[...] Considering that the three theories are all only partially useful at addressing the issue, we must synthesize information from these and other studies to arrive at the best possible model of neural semantic organization. To keep the useful parts of the category-specific theory, while eliminating the problem of arbitrarity of categories, the theory must include the findings from neuroimaging studies. PET scanning has shown that particular regions of the cortex respond to tasks in which subjects are asked to identify either animals or tools (Martin et al., 1996). [...]
[...] By incorporating data from neural scanning, explaining puzzling phenomena about the various types of agnosia, and offering a compelling model of neural mechanisms responsible for linking conceptual information with visual information, it proves to be the most thorough and insightful of the theories. All these studies, however, have built upon each other and have illustrated to us that the brain's systems are neither compartmentalized nor generalized. Almost all complex functions require an orchestrated collaboration of associated areas. This is a promising insight because in this case even dramatic lesions cannot utterly eliminate a person's capacity to interact with their environment. [...]
[...] Instead of considering particular cortical areas as wholly responsible for the comprehension of visual stimuli of a certain category, the paradigm is restructured so that associated cortical areas, whose purposes correspond to the functions or intrinsic properties of the stimulus, aid basic visual and verbal neural pathways. In the study, the ventral temporal lobes and Broca's area were always involved in the task, regardless of whether the stimulus was an animal or tool. This indicates that these two areas perform a function basic to visual-verbal identification, while the other areas provide extra information necessary for accurate and reliable identification. [...]
Online readingwith our online reader
Content validatedby our reading committee