We summarize our results as follows: 1. 1. Style is a function of pragmatic variables. 2. 2. The main pragnatic components are SPEAKER and ADDRESSEE. 3. 3. Each of these components is characterized by a certain expectancy norm regarding features of stylistic expression. It is true that these features of stylistic characterization often interfere with style features dictated by certain roles which one of the communication partners adopts; nevertheless, they are fundamental for assessing the phenomenon ‘role’. 4. 4. In recent linguistic approaches to poetic style, it has been the common tendency (implicit or explicit) to regard the general feature of ‘innovation’, among others, as underlying a structural description with generative power. In describing the deviation of poetic forms of expression from everyday language we have not tried to investigate strategies which make this general feature of ‘innovation’ in poetics predictable in any functional sense. Therefore, our question as to those regularities which remain unviolated in modern poetry has necessarily remained unanswered. However, we think that we have devised rules to explain metaphorical phenomena which are more plausible than earlier attempts. 5. 5. The fact that in explaining style features we speak of the deviant extension of the meaning of lexical items in syntactic sequences, on the one hand, and presupposition, on the other, indicates that we master different levels of competence. The transfers of imagery (psychologically speaking, the obtaining, by one or more semantic features of a lexical item, of priority over the rest of the features and its (their) relation to a different object, which under this transfer receives the phonological representation of the original object name: for example, [function: stand] for the human leg as well as for the leg of a table), furthermore that intersecting of semantic features which deviates from everyday language (for example, lila rede cf. p. 136f.) are principle strategies within stylistic competence indicative first of our commanding over feature complexes which are NORMAL relative to the respective idiosyncratic competence, and second that, beyond that, we also utilize a great number of accepted associations of stylistic features. 6. 6. Acceptability is a fundamental factor in the interpretation of dynamic competence, which, beyond the basic grammatical competence, is responsible for the strategies of stylistic classification. In other words, beyond a certain (more or less invariant) basic competence, we master a number of strategies which, being functions of pragmatic variables, make possible the identification of different styles. Such strategies are not to be explained on the basis of those formal and substantive predispositions assumed by generative transformational grammar (Katz, 1966). They are only to be explained as processes of structuring constantly being revised in accordance with the ever changing and widening human level of experience.