Главная Случайная страница


Категории:

ДомЗдоровьеЗоологияИнформатикаИскусствоИскусствоКомпьютерыКулинарияМаркетингМатематикаМедицинаМенеджментОбразованиеПедагогикаПитомцыПрограммированиеПроизводствоПромышленностьПсихологияРазноеРелигияСоциологияСпортСтатистикаТранспортФизикаФилософияФинансыХимияХоббиЭкологияЭкономикаЭлектроника






Perceptual properties and functional properties: Experimental and neurolinguistic data

It is interesting to compare the theoretical analysis conducted by structural seman­tics and Greimasian semiotics with experimental data from psycholinguistics and neurolinguistics, which seem to converge in indicating different levels of activa­tion, in both normal and pathological linguistic functioning, in relation to what in structural terms we could call different semic categories. As far as psycholin-guistic data are concerned, a number of experimental studies10 have demonstrated the existence of different activation procedures for different semantic properties of referential terms, that is, common nouns used to denote real objects. In particular, two types of component have been identified in the semantic information relating to a noun: perceptual components and conceptual or abstract components. Per­ceptual information consists of properties based on attributes directly present

in our perceptual experience of the referents of a word, such as form or color, while conceptual information includes non-perceptual properties of various kinds, above all functional properties, but also relations with other objects or member­ship of a superordinate category.

If, for example, we consider a term like coffee, we find that its semantic rep­resentation includes a wide range of encyclopedic properties. We know that coffee is a drink, it is dark, it has a particular bitter flavor, it has stimulating effects because it contains caffeine, etc. Some of these attributes, like being black or liq­uid, are perceptual, while other properties, like the fact that it is derived from the toasted beans of a particular plant, contains caffeine, and such like, are part of a set of encyclopedic knowledge (conceptual in the terms of these authors) that is not derivable from perceptual experience. Note that the difference between the two types of properties cuts across the distinction between natural kinds and artifacts, and does not imply predominance of perceptual properties in one class rather than another; for a term like tiger, the properties based on sensory-perceptual information are the form and the color of the coat, while all the other information making up our semantic competence on tigers belongs to non-perceptual sources—they are fierce, carnivorous mammals, live in India, etc.11

Various experiments have shown not only that the two types of properties have separate activation paths and are at least in part independent, but also that the perceptual properties are the first ones activated, and are recognized before the functional and abstract ones. This priority of perceptual properties could be based, according to the hypothesis of Flores d'Arcais, Schreuder, and Glazenborg (1985), on different stages of word meaning acquisition. Properties based on the physical characteristics of the object may be acquired before the more abstract ones and consequently may be more deeply rooted in our semantic system. This also finds confirmation in studies on language acquisition from which it has emerged that perceptual properties like form, dimension, movement, and sound are the first to be codified in a child's learning of the meaning of concrete terms. '2 According to Flores d'Arcais and his co-researchers, "this different acquisitional history of perceptual and functional attributes of the meaning of a word might bear some consequences on the process of semantic activation, and therefore be related to the readiness with which the semantic information of a word becomes available" (ibid.: 41).

Psycholinguistic data seem, therefore, to support a distinction between prop­erties, at least in relation to referential terms, and a psychological priority of per­ceptual features (even though what is defined as perceptual in the different disci­plines is not exactly the same). These data fit with linguistic data that we have already considered, such as that relating, for example, to the use of modifiers tike fake, which when applied to artifacts block the functional properties but not the perceptual ones. What may need to be asked is how the different activation ol properties interacts within other nominal classes in which perceptual data and functional conceptual (lata have different roles.

Mote generally, it seems to me that the psychologically founded distinction

between perceptual and functional properties requires more detailed examination, particularly the relation between form and function, which may be connected in different ways. Consider the perceptual/functional relation of an object like a shoe or a glove: it is difficult, if not impossible, to describe the form of a glove without referring to its function, because its function wholly determines its form. Clearly, we can see or perceive the form of a glove even if we are not in a position to recognize the function,13 but we do not have a specific linguistic term to denomi­nate that form.

The case of ball is different. Here we have a formal category for describing its property of being round. Roundness, although deriving from perceptual data, appears to be a more abstract and general category, and probably for this reason is lexicalized in many languages. There are, then, forms for which there are cate­gories and corresponding terms of description, and forms for which there are not, as with almost all the complex forms of artifacts. One might hypothesize that there is a hierarchy of importance within descriptive categories as far as form is concerned, which may relate to natural saliencies: "natural" forms, those which relate to a morphology of the natural world, may be more salient and more gen­erally lexicalized. This hypothesis is linguistically supported by the fact that in languages that possess class systems for nouns, the principal classes are in general connected to natural forms such as round objects, long objects, flat objects, and so on. This suggests the possibility of rethinking the relation between the elabo­ration of perceptual data and their codification in the linguistic system. On this point, a number of other interesting considerations are suggested by neuropsy­chological data on linguistic pathologies. Many studies of aphasic patients15 have demonstrated selective damage of abstract concepts and not of concrete concepts, and the reverse, proof of a certain autonomy in the functioning of the different properties. There may be an alteration in the capacity to recognize and attribute names from an image without any impairment of the capacity to identify ency­clopedic functional properties such as categorial relations and membership of superordinate categories, or other features which we might broadly define as en­cyclopedia. For example, in the case of terms denoting animals, form and color would relate to visual-perceptual information, while knowledge of the geographi­cal location of the ideal habitat, nutritional habits, degree of ferocity or edibility, and even the kind of sounds they produce would be classified as encyclopedic-conceptual information.

The fact that the damage suffered by the patient does not affect the two types of information to the same extent strongly suggests that encyclopedic knowledge and visual information are not represented in the same way and at the same level in our brains. This hypothesis is particularly interesting because it seems to pro­vide neuropsychological verification of the distinction, suggested by Marconi and discussed in section 7.5.4, between referential and inferential competence. The two competences would thus in principle seem to be separable and based on autonomous modules; the capacity to recognize referents, which is largely visual, may be damaged without any alteration of our connected encyclopedic knowledge

and inferential capacities. Moreover, these data suggest the existence of diversified representations for concrete terms and abstract terms: in the first case we will, in fact, have a double representation, both visual and propositional, while in the second there will be just one non-perceptual, conceptual level.

It is difficult, given the current state of psycholinguistic and neurolinguistic research, to predict how these initial results will influence a general semantic model of representation; certainly, however, even in their current fragmented form, they provide us with interesting points for reflection on the role that visual-perceptual information plays in the construction of meaning, and raise once again the problem of finding a unitary level capable of giving a simultaneous account of linguistic and perceptual representation.

The relationship between perceptual data and their inscription into the lin­guistic system points to the wider problem of the shift from the continuous to the discrete, and how a flux of unsegmented perceptions can be rendered discrete by linguistic and conceptual categories. Obviously, not all our perceptual experience is codified, or codifiable, in language; only a small part of these phenomenological riches passes into language. This raises at least two crucial theoretical issues. Above all, how much and in what way is perception, which in itself is indepen­dent of language and culture, selectively reinforced by linguistic-cultural struc­tures (a problem which underlies all semiotic discussion of iconism).1 Secondly, what actual margin of variation is there between languages: can languages differ from each other without limits, as a radical relativist would argue, or are there regularities, "points of resistance in being," as Eco (1996) has defined them, which open up possible lines of segmentation of the perceptual continuum, making some categories more probable than others? Without going into these questions here, I would like to emphasize that the problem of the move from the continuous to the discrete does not only concern perceptual experience, but also extends to other spheres, in particular that of the psycho-affective states that emerge linguis­tically in the axiological dimension of the semantic system.

Последнее изменение этой страницы: 2016-08-11

lectmania.ru. Все права принадлежат авторам данных материалов. В случае нарушения авторского права напишите нам сюда...