3. CAPÍTULO 3
3.5. Análisis de resultados
In discussing the question of gradability, it is desirable to incorporate some descriptions of the class of adjectives in the literature. Quirk et al.
(1985, 403) describe the class of adjectives in terms of typicality. They propose four main ‘‘criteria for adjectives’’:
(a) Occurrence in attributive function (b) Occurrence in predicative function (c) Premodification by the intensifier very
(d) Occurrence in comparative and superlative forms (inflectional or periphrastic)
Most adjectives, the so-called ‘‘central’’ ones, will satisfy both criteria (a) and (b). But there are, of course, ‘‘peripheral’’ adjectives that only occur attributively; compare utter in utter fool or medical in medical student. Others only occur in predicative position, such as afraid, or call for complementation, e.g. glad(that) or subject (to) (Bolinger 1967a). For the reasons that both types clearly convey adjectival qualities and behave almost the same way with respect to degree modification, ‘‘central’’ and
‘‘peripheral’’ adjectives are of the same interest for this study. Criteria (c) and (d) have to do with gradability. In this case, it is only gradable adjectives which can be premodified by very, and can occur in comparative and superlative form. These two characteristics are both evidence of the fact that there is a scalar feature in the adjective.
As gradability is a constitutive principle of (c) and (d), one can argue that the latter is a special case of the former, to put it in another way, they represent two sides of the same coin. If an adjective can be modified by very, it can also undergo comparison, and vice versa. Lorenz (1999) states that comparatives and superlatives are grammatical ways of scaling upwards or downwards from a given point of reference, namely, the basis of comparison. To say that something is very important amounts to saying that it is more important than just ‘important’. In Lorenz’s view, criterion (d) is technically redundant.
According to Collins Cobuild English Grammar (CCEG) (2005, 65) the most important things to notice about an English adjective are:
1. what structure it is in, that is, before a noun or after a linking verb 2. what type of adjective it is, that is, ‘‘qualitative’’ or ‘‘classifying’’
While the first point obviously corresponds to attributive and predicative positions of adjectives discussed earlier by Quirk et al., the second point merits attention. The division of adjectives into ‘‘qualitative’’ and
‘‘classifying’’ is closely related to the question of gradability. In their function to identify qualities which someone or something has,
‘‘qualitative adjectives’’ are gradable, which means that the person or thing referred to can have more or less of the quality mentioned. Such as sad in a sad story, pretty in a pretty girl, small in a small child. Like Quirk et al., (CCEG) identifies a gradable adjective by the possibility of
‘‘submodification’’ by very and rather, and by the possibility of these adjectives to occur in the comparative and the superlative. ‘‘Classifying adjectives’’, on the other hand, are said to identify the class that something belongs to. Financial in the case of financial help is used as an example of classifying adjective. Financial has the function of classifying help, which is to say that, there are various kinds of help and financial help is one of them. As they place something in a class, classifying adjectives are not gradable in the way that qualitative adjectives are. Things are either in a particular class or not. Therefore, classifying adjectives do not have comparatives and superlatives and are not normally used with degree modifiers such as very or rather.
It is also pointed out in (CCEG) that there are many adjectives which are polysemous between being qualitative and classifying, and accordingly, they vary with respect to gradability. For instance, in the phrase the emotional needs of children, emotional is classifying and nongradable. It can neither undergo comparison, nor be modified by a degree modifier.
However, in the phrase an emotional person, emotional is qualitative and
gradable. It has a comparative and superlative form, and it can be used with very. A person can be very emotional or more emotional than somebody else. This example suggests that what makes the reading of the adjective gradable or nongradable depends on the noun it modifies.
At this point it has something of value to refer to a model which is capable of explaining what constitutes ‘‘polysemy’’ of such adjectives. In her study on adjectives, Warren (1984) presents a model of the relationship between the adjective and the noun to which it applies. She states that to uncover the nature of the adjective, one must be familiar with both the denotation of the adjective and with its relation to the noun it qualifies. Her semantic analysis involves two components, ‘‘referential content’’ and ‘‘relator’’ (See Figure 3-2).
a sad girl
referential content relator ‘‘sadness’’ ‘‘experiencing’’
Figure 3-2 Warren’s semantic model of the meaning of an adjective
The meaning of sad in a sad girl is analyzed in two components and spelled out as ‘‘X experiences sadness’’. In the following examples, it is the relator which seems to be responsible for constituting the polysemy of the adjectives: nervous breakdown versus nervous man, criminal court versus criminal assault and musical instrument versus musical child. Consider the elements of nervous in nervous breakdown and nervous man in Figure 3-3
a nervous breakdown a nervous man
‘‘nerves/nervousness’’ ‘‘caused by’’ ‘‘nervousness’’ ‘‘experiencing’’
Figure 3-3 The semantic elements of nervous
A nervous breakdown can be spelled out as ‘‘X is caused by nerve’’ and a nervous man as ‘‘X experiences nervousness’’. Warren states that nervous in nervous breakdown is ‘‘classifying’’ in that it restricts the application of breakdown. A nervous breakdown is a breakdown among other types of breakdown. While nervous in a nervous man is ‘‘qualitative’’ in that it qualifies the person in question. She points out that when nervous combines with man, it has a ‘‘relator’’ which allows grading, but when nervous combines with breakdown, it has a ‘‘relator’’ which is not compatible with gradability and so is resistant to modification of degree. Thus, polysemy in an adjective can be said to be due to the differences in at least one of the two main semantic components of adjectives, i.e. in the ‘‘referential content’’
and/or in the ‘‘relator’’.
According to Paradis, even in the case of ‘‘monosemy’’, some adjectives which are typically classifying in their right contexts can acquire a gradable reading. For instance, the adjective Russian in the man is Russian is classifying and nongradable, in other words, it has a bias towards nongradability. However, it can be coerced into a gradable reading, and the presence of a degree modifier may serve to confirm this, e.g. the man is very Russian, which is to be interpreted as ‘‘Russian style’’. The reason for promoting such a reading is that Russian can take a relator involving
resemblance, which is compatible with grading (1997, 46). At this point it has been shown that nervous is polysemous between a gradable/qualitative and nongradable/classifying meaning. While Russian is monosemous and basically nongradable but can undergo ‘‘contextual modulation’’. Problems of this nature will be discussed later in this chapter.
Returning now to the question of establishing criteria for gradability in adjectives. Admittedly, the traditional criteria for gradability, namely, (comparability) and (possibility to be modified by very), are of a limited value since there are many adjectives which do not occur in the comparative or the superlative but nevertheless occur with degree modifiers, although not with the degree modifiers of the type very. Paradis (1997, 47) gives the example of identical to illustrate the case in point. Identical can be modified with respect to ‘‘totality’’ by means of maximizing and approximating modifiers such as absolutely and almost. Identical, is thus not gradable in the sense that good is, since it can not be compared (*more identical, most identical) and since it is restricted to certain degree modifiers. Writing along the same line, Lorenz suggests a rephrasing to Quirk et al’s third criterion, namely, premodification by very, and he reasonably recommends
‘‘premodification by very or absolutely’’ (1999, 40).