The nature of semantic properties

The observations made thus far have tended to connect differences found on the surface level of linguistic-lexical manifestation 10 a deeper semantic structuring directly motivated by the differing nature of underlying experiences. Such a per­spective raises the prospect of a typology of the features that enter into the seman­tic configuration of individual terms. The most productive leads in this respect arc to be found in the semiotics of Greimas and in the classification he proposes between various classes ol semes, where seme is understood to be the minimal unit of signification.⁴ The distinction first proposed in Cireimas 1966 and sub

sequently developed⁵ is between figurative, abstract, and proprioceptive semes, which taken together articulate the entire semantic universe, that is, "they are co-extensive with a given culture or human person."

Figurative semes are those which refer to "sensible qualities of the world," to an articulation of sensorial orders that capture the exteriority of the world (hence the alternative term exteroceptive). To quote from the Dictionary:

figurative semes are dimensions of the content plane of natural languages, corre­sponding to elements on the plane of expression of the semiotics of the natural world, that is, to articulations of sensorial orders, to sensible qualities of the world. (Greimas and Courtes 1979: 333; my translation)

Abstract (or interoceptive) semes are

dimensions of content that do not refer to any exteriority, but which serve rather to categorize the world and to establish it as meaning: these are, for example, the categories of relation/term, object/process. (Ibid.: 334)

Finally, proprioceptive semes

connote the semic microsystems according to the category of euphoria/dys-phoria, constituting them in effect into axiological systems. (Ibid.)

Leaving to one side for a moment the analysis of the third semic level, the pro­prioceptive one ditectly connected to the problem of value, I would like to mm my attention to the figurative and abstract planes (exteroceptive and interocep­tive).

The distinction between these two semic categories seems to boil down to a difference between what is directly drawn from the external world (in fact, he talks about figures in the natural world) and thus linked to perception, and what does not have a correspondence in the external world and the perceptual experi­ence of it, but which lather represents its presupposition. Once again, then, it is perception which articulates the difference between the two levels: if figurative semes- derive from the natural world on the basis of our perceptual experience (in all its forms), abstract semes are universale of cognition and perception itself, and as such precede it and structure its forms and modes of organization. In other words, the abstract level is that of the interpretative categories or schemata of the sensible world, and is effectively situated at the level of the universale of language; according to some, it is situated at the level of the imaginaty and the metapsy-chological, a much deeper level than that of the linguistic sign.⁶ In this sense, abstract semes have undeniable analogies with the semantic primitives of cognitive research which were discussed in chapters 1 and 2.

Greimas' proposal is of evident interest because it raises once again the ques­tion of a basic ontology of deep semantic properties. However, the distinction between the abstract and the figurative is far from being completely satisfactory, and the two terms effectively constitute two purely theoretical paradigms which cannot in reality be separated/ Moreover, closer analysis reveals that the exrero-

ceptive level is also much more complex than a straightforward reference to "sen­sible qualities of the world." A definition like this seems too generic to fully cap­ture the different modalities by which the world appears to our senses. The prob­lem is not only the manifold nature of the sensible qualities of the world, but above all the different degree of density and structural stability of the exterocep­tive semes.

Recognition of and reflection on this is clearly present in the traditional philosophical distinction between primary and secondary qualities which, starting with Democritus, works its way, via Galileo, all the way through to the English empiricists. In this tradition, primary qualities correspond to determinations of the object as such, which possess an objective measurability as regards solidity and spatiality: number, extension, figure, movement, shape, size, etc. Secondary quali­ties, on the other hand, relate to sensible determinations of the object, such as colors, sounds, flavors, etc., which are thus variables in our experience.

The difference between the two orders lies, therefore, principally in the ob­jective character of the primary qualities, that are inherent to the physical nature of the object. However, this objective character was questioned by Berkeley and Hume, who contested the possibility of a clear separation of the two orders of properties. The impossibility of isolating an objectivity of the natural world that is separate from its subjectively experienced aspects means that primary and secondary qualities are inextricable; as a result, primary properties lose their ob­jectivity and they all end up with the same subjective character of "ideas."

From the point of view of a phenomenology of perception, both kinds of properties have a similar complexity in terms of their genesis and internal me­chanics, and their sensitivity to the influences of the perceived environment and to strategies of observation.⁸ At the same time, however, some qualities, features, and properties continue to appear more primary than others: shape and dimen­sion, for example, seem more "resistant" than reflections of light on the surface of an object.

This different degree of resistance may be thought of, as Bozzi does, in terms of different degrees of stability, whereby certain groups of properties ensure the stability of the world, while others are more pliable and relative.

I am able to think that the true primary qualities are those structures which can be observed with greater stability and which in each moment form the scenario of the wotld, the state of movement and rest of objects, and so on. Drawing on what is now an unaccustomed use of language, we can readily say that position in space, the solidity and movement of objects, are lived as primary qualities of our experiential surroundings. (Bozzi 1990: 99; my translation)

The interesting thing about observations like these is that they have a precise correspondence on the plane of linguistic structure: some perceptual properties, or exteroceptive semes, are constituent components of the semantics ol the respec­tive terms, 'lb take just one example, think about how our knowledge ol the mot phology ol loiius determines different interpretations in the case ol certain noun

phrases formed by a noun plus an adjective. With color adjectives, we interpret very differently which part or component of a given entity the chromatic deter­mination applies to, and we do so solely on the basis of knowledge relating to the forms of objects. A red watermelon and a red apple are not red in the same way, because a watermelon is only red on the inside, while an apple is only red on the outside. A white house has white external walls, while a white room has white internal walls. A blue suit is uniformly blue, while only the bodywork of a blue car is blue. The application of the color terms in the interpretation of these noun phrases depends exclusively on the form and configuration of the entities which the color terms modify. This demonstrates the linguistic importance of the prop­erties in relation to the form and thus the necessity of their inclusion in the se­mantic representation.

It is hard to abstractly define how many and which perceptual properties should be included in the semantic representation: this varies by semantic field. Form and dimensions, as has been said, certainly seem to possess a high degree of stability in a typical configuration, and in some cases color also has to be in­cluded.⁹ Exteroceptive semes may therefore be of varying density and combine in various ways in the semantics of individual lexemes, which will in each case have to specify the degree of determination. As I have already observed, perceptual properties generally contribute to the characterization of typical features rather than constitutive ones, which can often be identified with functional components. The distinction between form and function does not translate easily into Greimas' distinction between interoceptive and exteroceptive, but seems rather to assume an intermediate position on the continuum, because functional properties are not entirely reducible to perceptual aspects, although in a broad sense they depend on a phenomenology of corporeal movement. The functional component seems, however, to play an important role for the semantics of certain lexical classes, as is demonstrated by the results of experimental data from other disciplines.