a grammatical topic ~ a purely grammatical topic — the most grammatical of the suggested topics.

In order to overcome the demonstrated lack of rigour in the definitions in question, we may introduce an additional linguistic distinction which is more adaptable to the chances of usage. The suggested distinction is based on the evalua­tive function of adjectives. According as they actually give some qualitative evaluation to the substance referent or only point out its corresponding native property, all the adjective functions may be grammatically divided into "evaluative" and "specificative". In particular, one and the same adjec­tive, irrespective of its being basically (i.e. in the sense of the fundamental semantic property of its root constituent) "relative" or "qualitative", can be used either in the evalua­tive function or in the specificative function.

For instance, the adjective good is basically qualitative. On the other hand, when employed as a grading term in teach­ing, i.e. a term forming part of the marking scale together with the grading terms bad, satisfactory, excellent, it acquires the said specificative value; in other words, it becomes a spe­cificative, not an evaluative unit in the grammatical sense

(though, dialectically, it does signify in this case a lexical evaluation of the pupil's progress). Conversely, the adjective wooden is basically relative, but when used in the broader meaning "expressionless" or "awkward" it acquires an eval­uative force and, consequently, can presuppose a greater or lesser degree ("amount") of the denoted properly in the corresponding referent. E.g.:

Bundle found herself looking into the expressionless, wooden face of Superintendent Battle (A. Christie). The su­perintendent was sitting behind a table and looking more wooden than ever.

The degrees of comparison are essentially evaluative for­mulas, therefore any adjective used in a higher comparison degree (comparative, superlative) is thereby made into an evaluative adjective, if only for the nonce (see the examples above).

Thus, the introduced distinction between the evaluative and specificative uses of adjectives, in the long run, empha­sizes the fact that the morphological category of comparison (comparison degrees) is potentially represented in the whole class of adjectives and is constitutive for it.

Among the words signifying properties of a nounal referent there is a lexemic set which claims to be recognized as a separate part of speech, i.e. as a class of words different from the adjectives in its class-forming features. These are words built up by the prefix a- and denoting different states, mostly of temporary duration. Here belong lexemes like afraid, agog, adrift, ablaze. In traditional grammar these words were generally considered under the heading of "pre­dicative adjectives" (some of them also under the heading of adverbs), since their most typical position in the sentence is that of a predicative and they are but occasionally used as pre-positional attributes to nouns.

Notional words signifying states and specifically used as predicatives were first identified as a separate part of speech in the Russian language by L. V. Shcherba and V. V. Vinogradov. The two scholars called the newly iden­tified part of speech the "category of state" (and, correspond­ingly, separate words making up this category, "words of the category of state"). Here belong the Russian words mostly ending in -o, but also having other suffixes: тепло, зябко, одиноко, радостно, жаль, лень, etc. Traditionally the Russian words of the category of state were considered as constituents of (he class of adverbs, and they are still considered as such by many Russian schiolars.

On the analogy of the Russian "category of state", the English qualifying a-words of the corresponding meanings were subjected to a lexico-grammatical analysis and given the part-of-speech heading "category of slate". This analysis was first conducted by B. A. llyish and later continued by other linguists. The term "words of the category of state", being rather cumbersome from the technical point of view, was later changed into "stative words", or "statives".

The part-of-speech interpretation of the statives is not shared by all linguists working in the domain of English, and has found both its proponents and opponents.

Probably the most consistent and explicit exposition of the part-of-speech interpretation of statives has been given by B. S. Khaimovich and B. I. Rogovskaya. Their theses supporting the view in question can be summarized as follows.

First, the statives, called by the quoted authors "adlinks" (by virtue of their connection with link-verbs and on the analogy of the term "adverbs"), are allegedly opposed to adjectives on a purely semantic basis, since adjectives denote "qualities", and statives-adlinks denote "states". Second, as different from adjectives, statives-adlinks are characterized by the specific prefix a-. Third, they allegedly do not possess the category of the degrees of comparison. Fourth, the combinability of statives-adlinks is different from that of adjectives in so far as they are not used in the pre-positional attributive function, i.e. are characterized by the absence of the right-hand combinability with nouns.

The advanced reasons, presupposing many-sided categorial estimation of statives, are undoubtedly serious and worthy of note. Still, a closer consideration of the properties of the analysed lexemic set cannot but show that, on the whole, the said reasons are hardly instrumental in proving the main idea, i.e. in establishing the English stative as a separate part of speech. The re-consideration of the stative on the basis of comparison with the classical adjective inevitably discloses (lie fundamental relationship between the two, — such relationship as should be interpreted in no other terms than identity on the part-of-speech level, though, naturally, providing for their distinct differentiation on the subclass level.

The first scholar who undertook this kind of re-consider­ation of the lexemic status of English statives was L. S. Barkhudarov, and in our estimation of them we essentially fol­low his principles, pointing out some additional criteria of argument.

First, considering the basic meaning expressed by the stative, we formulate it as "stative property", i.e. a kind of property of a nounal referent. As we already know, the adjective as a whole signifies not "quality" in the narrow sense, but "property", which is categorially divided into "substantive quality as such" and "substantive relation". In this respect, statives do not fundamentally differ from classical adjectives. Moreover, common adjectives and partici­ples in adjective-type functions can express the same, or, more specifically, typologically the same properties (or "qual­ities" in a broader sense) as are expressed by statives.