Abstract
The terms analytic and synthetic are extensively used in linguistic literature, but as yet there exist no generally accepted definitions of these terms. The concepts of analysis and synthesis stand in direct relation to the concept of word. The common notion is that a one-word expression, like Estonian majasse ‘into the house’, palusin ‘I requested’, is synthetic, whereas a two-word expression, like maja sisse ‘into the house’, olen palunud ‘I have requested’, is analytic. But the concept of word is as vague and as difficult to define as are the concepts of analysis and synthesis (cf. Velten IF 53 21). Even in such languages and in such cases in which a word is easily identified as a distinct entity, no conclusive evidence concerning its analytic or synthetic nature can be derived from the fact that the expression consists of a separate word. It is not sufficient to classify all expressions as analytic or synthetic, because in practice there are more degrees. First and foremost, the relativity of the concepts of analysis and synthesis is to be taken into consideration. An expression may be more analytic in comparison with another and more synthetic when compared with a third. Thus, one-word expressions are not all of them synthetic to an equal extent. In Estonian, for instance, the illative majja ‘into the house’ is more synthetic than the synonymous ill. majasse. The latter again is more analytic than majja, but more synthetic than the equivalent alternative maja sisse. Therefore, expressions may be divided into two classes, analytic and synthetic, in very rough outline only. Instead of such a classification it would be more important to indicate the criteria by means of which we may decide whether a given expression is more analytic or synthetic than another. In discussing the types of morphological expressions the concepts of analysis and synthesis cannot, however, be dispensed with.