The purpose of the thesis is to compare the traditional and the modern structuralist approaches to the problem of classifying the Russian verbs. In the Introduction a brief historical outline of the treatment of the problem is given. Within the traditional school two main tendencies developed: classification by the present- and by the infinitive-stem. Within the framework of modern structural linguistics new approaches to the problem of classifying the Russian verb were attempted. These attempts culminated in the descriptive system of Roman Jakobson. Besides its main purpose: a comparison of the results of Jakobson with those of two of the newer representatives of the traditional school - Berneker and Unbegaun -, the thesis has as secondary purpose to explain and to a certain extent to criticize the work of Jakobson. In Chapter I the classification of Berneker which starts from the infinitive is presented and discussed; in spite of its doubtless pedagogical merit, the classification is found to contain flaws in its methodology: it is based on mixed criteria, it is not strictly synchronic, it is not exhaustive and it separates groups of verbs which linguistically belong together (as a result of preoccupancy with script). In Chapter II the classification of Unbegaun, which is based on the present tense form, is discussed. Unbegaun’s classificatory technique is found to be stricter than Berneker’s, but this very strictness accentuates the shortcomings of the system. Like Berneker's, Unbegaun's system is largely based on script, and in his case the consequences are more serious. Chapter III is devoted to a discussion of Jakobson’s approach. Jakobson has solved the problem on which all traditional classifications stranded - the matching of present- and infinitive-stems. His solution consists of the setting up of a (sometimes artificial) underlying stem-form, from which the alternations of the stem can be predicted on the basis of the simplest possible set of rules. The alternations of the stem are, in the main, described in terms of truncation (loss of a final stem phoneme). In view of this feature the basic stem-forms are subdivided into stem in vowel, stem in j, v, m, n, and stem in other consonants. The subdivision proves useful in the statement of the rules for softening and stress, for which Jakobson has been the first to state general rules. In the Conclusion it is demonstrated that, as opposed to the confusion of varied criteria of classification characteristic of the traditional school, the basis of Jakobson's system is simply the phonemic structure of the basic stem-form. Furthermore, Jakobson's systematizing technique differs basically from that of his predecessors. Whereas the latter carry out consecutive subdivisions of the material thus obtaining separate classes of verbs - set up on the basis of separate criteria, Jakobson’s descriptive system forms one closely-knit whole, where a minimum of distinctions is employed to the describe the behavior of a maximum of the total number of Russian verbs. / Arts, Faculty of / Linguistics, Department of / Graduate

This thesis is concerned with one aspect of the problem of Russian verbal morphology: the basic patterns of different arrangements of the derivational affixes - prefixes and suffixes - of the stems of the Russian verb and the one or more stems that may be assigned to a particular arrangement. This thesis does not consider the desinential morphology of the verb, which may be neatly distinguished from the derivational. The arrangements of the patterns of the derivational affixes is determined by isolating all verb-stems with the same derivational affixes in the same relative position. To quote our examples in section 4.20, page 14 we have two verbs topta / potopta ņiza / naņiza vtaptivaj / vtopta naņizivaj / naņiza vitaptivaj / vitopta proņizivaj / proņiza with the following pattern of recurring forms (morphemes) -a / po- -a v- -ivaj / v- -a vi- -ivaj / vi- -a -a / na- -a pro- -ivaj / pro- -a na- -ivaj / na- -a and by subsuming the prefixes vi- v- pro- and na- as P- we may produce the following schema -a / P- -a P- -ivaj / P- -a for which the second line represents all prefixed forms that occur with the particular stem-base which has this derivational arrangement or matrix . The above schema has the following form in this thesis. -a / P- p-( / -a) All the stem-bases of a matrix are called the form-class of the stem-bases of that matrix. The form-classes are inserted into the matrix , morphological rules are applied, and the original stems are formed, plus those that may in theory exist according to the matrix and morphological rules. In this way we wish to demonstrate the different basic patterns of derivational morphology, and the differences of these patterns as shown by the different matrices. This second objective of this thesis is done by means of a matrix array. A matrix array is an arrangement of matrices according to certain regularities of structure. Form-Class Matrix 1 to Form-Class Matrix 47 is summarized on page 158. The remainder , Form-Glass Matrix 48 to Form-Class Matrix 52 are the complex matrices of the indeterminate | determinate stems and are found on pages 139 to 150. Besides the numbered matrices there are some special cases considered throughout the text. From an examination of the matrix array we may see that row FCM - 10 to FCM - 34 shows stems with varied aspectival imperfective structure , prefixed perfective or perfective in -nu or both. The lexical perfective is attested, but not the lexical imperfective. Row FCM - 1 to FCM - 37 shows the addition of iterative | non-iterative stems formed by -áj or -váj and the presence of lexical imperfective stems with the same structure as the iterative stems. Row FCM - 5 -to FCM - 36 (plus the unnumbered matrices of the row) show iteratives in -ivaj and lexical imperfectives with the same structure. The lexical perfective stems are identical in structure to the aspectival imperfectives and the stems in -nu also occur as lexical perfectives. Row FCM - 13 to FCM - 38 (and the unnumbered matrices) show iteratives in -ìvàj (and in one case -aj) and lexical imperf ectives in - ìvàj or one of –àj ̴ -vàj ̴ eváj. The Extended Matrices are excluded from the above mentioned for three main reasons. (1) FCM - 6, 7, 8, 14, and 22 have aspectival perfectives without an overt formant. (2) FCM - 27 and 32 have an additional lexical perfective in -i and to included these matrices with their above ones would intrude into the general pattern of the matrix array. (3) FCM - 33 , with formants in -a(j) shows characteristics of two matrices and is entered beneath them. Forms is -a(j) are rathered unsettled in usage in modern Russian, to judge by conflicting remarks in the literature. FCM - 39, 40, and 41 examines stems with the formants -aj / -i and FCM - 42, 43, 44, 45, and 46 are concerned with verbs of dual aspect , both imperfective and perfective, and with verbs formed from -ova- ̴ -uj. The form-class matrices for the indeterminate | determinate defy easy characterization, but their complications are in 'in terms of’ the other matrices. If we excluded certain specialized cases we may say that the derivational morphology of the Russian verb may be considered under five very general classes. (1) Stems with various aspectival imperfective structure, with prefixed perfectives, or perfectives in -nu, or in both. The iterative and lexical aspectival stems are formed by -áj ~ -váj ~ -V-váj or - ìvàj or both - áj ~ -váj ~ -V-váj and – ìvàj. (2) Stems with aspectival perfectives without formant and with prefixed imperfectives in - ìvàj, - áj or váj as stated for the particular case. (3) Stems with lexial perfective forms in -i in addition to -V and -nu or - áj and -nu. (4) Stems with the formant -a(j) , i.e. both -a and -aj appear as stem formants. (5) Stems with the formant pair -a / -i (6) Stems with dual aspect and stems in -ova- ~ -uj- (7) Indeterminate l determinate stems / Arts, Faculty of / Central Eastern Northern European Studies, Department of / Graduate

In his article "Russian Conjugation" (Word IV, 1948) Prof. Roman Jakobson presented an alternative analysis of the Russian conjugation based on the principle of the 'full-stem' and the "pivotal feature" of all Slavic conjugation, truncation, i.e. the loss of the final phoneme of a (verbal) stem or of the initial phoneme of the desinence. By confining this strictly synchronic formal analysis of present-day Standard Russian to simple verbs (with unprefixed one-root stems) and to the purely-verbal categories (the finite forms and the infinitive), adequate coverage of all principles and problems, including many illustrations, was possible in just a little more than 9 pages and the final section, listing only 41 exceptions, though somewhat oversimplified, gave striking proof of the superiority of the principles tested and a hint of the use that might be made of them in teaching Russian. With these two impressions in mind (i.e. the system's apparent superiority and its possible usefulness in the class), a further examination has been made here which involves the extension of the analysis to include those verb forms purposely excluded by Jakobson (i.e. the compound and semiverbal forms) in an attempt to demonstrate the extent of its validity and, at the same time, to investigate the practicality of its adoption in teaching Russian. It would, after all, be to the credit of the full-stem system of analyzing the Russian conjugation if it were possible, without too much complication, to predict the behavior of a verb's gerunds, participles and prefixed finite forms on the basis of its full-stem. In any case, the teacher of Russian requires some demonstration of the system's performance when it is applied to these additional forms as students are apt to want to do. A consistent effort has been made to preserve the essence of the full-stem, as conceived by Jakobson, as a sort of formula or 'key' to the grammatical forms. No new types of full-stem have been introduced, but two basic additions to the general principles do appear. One of these, the principle of the two-full-stem or multi-full-stem verb, is instrumental in reducing the number of exceptions admitted by Jakobson and is the result of the logical extension of Jakobson's basic principle to the position that any verbal form must realise a full-stem, even if it means proposing two or more full-stems to account for all the forms of certain verbs. A second major addition involves one of the most important 'significant features’ of the full-stem, namely the accent type. This is a direct result of the inclusion of semiverbal and prefixed forms. In order to explain the different stress preferences in these forms in otherwise identical full-stems, e.g. the nonsyllabic unaccented V'J_ (twist) and L'J__ (pour) with Masc. Pret. and short Masc. Past Pass. Part. na+v'í_l-#, na+v'í_t-# and ná+l'i__l-#, na+l'i_t-# (preferred to na+l'í_l-#, na+1'i_t-#) respectively, the principle of 'degrees of unaccentedness' within full-stem types has been introduced (i.e. V' J__ unacc., L' ˋJ_ maximally unacc.). In an attempt to improve the already established principles of describing the accent types of various full-stems, more emphasis has been placed on the feature of columnarity of stress pattern as the primary criterion for accentedness. This topic is of particular concern among the monosyllabic broadly-closed full-stems, which Jakobson divides into accented and unaccented varieties, but which may alternatively be considered two different varieties of accented full-stems whose "columnar" stress patterns contrast with those of the unaccented items among the other types of non-polysyllabics. With only a few changes in the detail of the system as conceived by Jakobson, the additions named above have been successfully adapted to the full-stem idea. This has however been achieved only at the expense of the clarity, precision and brevity of the exposition. Though the validity of the system has only become ever more apparent as the more complex data regularly submitted to formulization, the full complexity of the problem as it exists without any artificial limitations has made the preservation of the system's original virtues impossible even though, as will be seen, the number of verbs that need be strictly termed exceptional can be effectively reduced, even from Jakobson's low total of 41. The system's attractiveness from the pedagogical point of view is, however, seriously affected "by this increased complexity". / Arts, Faculty of / Central Eastern Northern European Studies, Department of / Graduate