REVIEWS
Hans-Jochen Pasenow, Die konjunktionale Hypotaxe in der Nikonchronik, Berlin 1987 (=Osteuropa-lnstitut an der Freien Universitdt Berlin. Slavistische Ver6ffentlichungen 62), X + 194 pp. It is interesting to note that historical linguistics seem to stage a comeback these days. And it pleases me to see that Russian linguistics do not lag behind (as they often do) in this respect. There are two reasons for the renewed debate on historical topics: one is what might be called "ideological", the other factual. Both are closely intertwined. "Ideologically" the salient point seems to be that a scholar has a preconceived theoretical framework and tries to make the facts fit the theory. And that is where the factual aspect of the problem comes in: we do not have a clear picture of the development of the Russian language. And it is precisely this lack of factual evidence that allows for (or even calls for) bold theories and sweeping generalisations. Such theories, unless they are tested and verified by applying them to the Russian language, not only contribute little to our knowledge: they are actually detrimental to historical linguistics since they tend to be accepted as proven facts whereas they are little more than speculations. What we need, then, is a sound material basis first; a theory to fit the facts must come later. The only way to arrive at a sound material basis is by describing the linguistic evidence. And the evidence is to be found in manuscripts. Each ms. of a given period ideally represents the product of a linguistic system that can be reconstructed by analysing the language of the ms. ~ Similarly all the ross. of a given period and a given area are representatives of the language that can be reconstructed by analysing and generalising the linguistic systems of all the mss. What we need first and foremost, therefore, are descriptions of the language of individual ross., descriptions that try to explain all the linguistic facts of a given ms. And we need such descriptions for a sufficient number of mss. It is for this reason that publications analysing the language of individual mss. must be welcomed. Unfortunately they are not very frequent. Somewhat more frequent are analyses of one aspect of the language. They are not ideal since they isolate one part of the linguistic system but they are still useful provided that they subject this part to an exhaustive analysis. An example of this approach is Pasenow's book. 2 Following his teacher (H. Brfiuer) he concentrates on a particular aspect of syntax: the dependent
Russian Linguistics 13 (1989), 55.
56
REVIEWS
clause introduced by a conjunction. 3 The ms. chosen for examination is the Obolenski copy (cnncoK O6oJIeHcKoro, O=I_IFA~A qb. 201, No. 163) of the Nikon chronicle dated to the end of the 1620's. P. did not work with the ms. itself: he used the edition of the chronicle in PSRL 9-13 where, however, the codex electus is the Patriarch copy (Flaxpr~aptuecrr~ cnnco~, P = B A H 32.14.8). The texts he quotes abundantly thus acquire a peculiar surrogate quality: wherever the text of O coincides with P the latter is quoted (13)4; if O deviates it has to be assumed that for the length of the varia lectio 0 is quoted only to give way to P afterwards. This is, of course, a most unsatisfactory procedure for a linguist but since it is syntax the author is interested in (and not graphematics or morphology) the results are not noticeably influenced by this. Still I cannot help feeling uneasy with this kind of CBO,~IHbI~TerCX. P. chose a very bulky text for his analysis, partly because this text has not yet received the attention it deserves (1), but also because his method requires a lot of material. P. classifies more than 5000 subordinate clauses: quite an impressive material basis. Since he applies statistical methods to this population and since there are many parameters according to which the subordinate clauses are distributed (40 conjunctions, 11 types of clauses, and 7 time periods) it turns out that the material basis is still too small: certain statistical tests cannot be performed on account of this. Nevertheless the amount of text analysed surpasses that of most comparable linguistic analyses. Quantity, however, does not guarantee quality. In order to gauge the usefulness of P.'s book I shall first indicate the contents, then discuss.major points of methodology, and finally mention a few minor points in the order in which they are treated in the book. The book opens with an introduction that contains general informations about the text and the ms. P. then explains how the subordinate clauses are classified (viz. according to the function they fulfill in the main clause) and states that infinitive and participial constructions are also included, provided they meet certain requirements. In some detail he describes the quantitative aspects of the investigation. The text is divided into seven periods (A.I-A.7) comprising a century each but being of unequal length (A.I has less than three percent of text, A.7 almost thirty). The author distinguishes several types of tables. Since the data are trivariate he theoretically obtains three sets of tables: conjunction - type of clause, conjunction - period, type of clause period (the latter is not used but can be reconstructed from the others). The two types of tables used may contain absolute or relative frequencies. Relative frequencies are also represented in the form of curves and bar charts (41 and 45). Finally the author introduces the concept of theoretical frequencies.
REVIEWS
57
In the main body of the book P. analyses the use of the different conjunctions for each type of clause: subject, predicative, object, attributive, causal, conditional, concessive, consecutive, final, temporal clauses, and clauses of manner. Each chapter starts with a general definition and informations relevant to the particular type of clause, information on the quantitative distribution of conjunctions in the text as a whole and in individual periods, and then the conjunctions are analysed in alphabetic order. Everywhere the findings are illustrated by examples. An appendix containing cumulative tables (for A.I-A.7 separately and for the text as a whole) and a bibliography are added at the end. The new approach in P.'s book as compared to other analyses is undoubtedly the use of statistical procedures. Unfortunately, this is also its weak spot. The application of statistical methods is very useful and indeed necessary since a ms. must be viewed as a random sample of the parole of a given langue. And parole (in contradistinction to langue) has the characteristics of a statistical population. 5 But this is only a very general statement that does not take into account all the problems to be encountered in historical linguistics. Mss. containing a traditional text and even more so mss. containing chronicle texts are but very imperfect representatives of parole of a given period. It is therefore advisable to test the assumption that a ms. (in this case O) is a random sample of a well-defined statistical population (viz. the Russian "literary" language of the seventeenth century) that will yield linguistically useful informations before going to the trouble of applying statistical procedures. Regarding O I entertain serious doubts as to the validity of the assumption and it seems that P. is not so sure about it, either. The main problem in P . ' s eyes is the fact that the population is too small and that the results may be influenced by a chance high frequency of clauses attested infrequently elsewhere. This is so evident that P. cannot circumnavigate the problem. When he discusses the subdivision of the text into periods he admits that the number of subordinate clauses introduced by a conjunction is too small for periods of fifty years and that he had to extend the time span to 100 years (4, n. 17). Even this is not enough, however: quite often he warns us that certain frequencies are not to be considered statistically valid: "Der Gipfel im vierten Abschnitt ist wegen der geringen Zahl der Belege nicht fiberzubewerten." (48) "Der hohe Wert in A. 1 sollte wegen der geringen Zahl der Belege in diesem Abschnitt nicht fiberbewertet werden." (55) "Da in A. 3 nur ein und in A. 4 kein (to-Attributsatz belegt ist, sind die Werte ffir den Anteil der (to-Attributsfitze an allen (to-Sfitzen erst ab A. 5 ernst zu nehmen." (79). 6 Such circumspection, however, is not evident everywhere: in fact P. displays it only when the
58
REVIEWS
results are unusual. In the case of jako in final clauses he notes a decline from 30 percent down to 3.5 without hinting at the fact that the 30 percent relate to three occurrences only (149). Another reason (and in m y eyes this is the main reason) for the fact that the number of occurrences cannot always be taken at face value is the subdivision of the population. P. divides the text into centuries, but the centuries are not all allotted the same amount of space in the ms. The last two periods contain almost 30 percent each, roughly the same as the first four periods taken together (4). The picture does not change when the number of clauses is taken as the basic unit (42). This invalidates many of the findings for periods A. 1-A. 5. Generally speaking statistical significance can only be tested if the number of occurrences is 5 or higher. In the case of A. 1 this limit is attained by 7 out of 165 theoretically possible and out of 37 actually attested combinations (183). The subdivision into centuries is thus rather questionable from the point of view of statistics. It is even more questionable from the textologist's point of view. Chronicle texts are compilations and the language of the sources is not altered significantly by the compiler (or compilers). A compiler, on the other hand, is free to add texts anywhere in the chronicle, and these additions would then reflect his (contemporary) language and style. In the case of the Nikon chronicle it has been shown that Daniil's additions to the text contain stylistic peculiarities that influence linguostatistics. 7 And the possible influence of the sources is acknowledged by P. himself in the case of pone~e: "Vielmehr k6nnte es so sein, dab ein in die Chronik aufgenommener Text sich sprachlich von seiner Umgebung abhebt und unter anderem einen untypisch hohen Anteil von jako-Kausals~ttzen aufweist." (109). In view of this P.'s subdivision of the text into centuries is very questionable indeed. If at all, the text would have to be subdivided into its sources and not into centuries. But then the sources are not homogenous either since their "genre" (factography or narration etc.) is responsible for differences, too. P. knows this: ,,Bei einem Schnittverfahren nach inhaltlichen Gesichtspunkten liefSen sich fundierte Aussagen fiber die syntaktisch-stilistische Struktur der Abschnitte machen. Bei dem hier gew/ihlten Verfahren scheinen Strukturverschiebungen zwar durch, eine Lokalisierung untersuchter Erscheinungen auf einer Berichtszeitskala l~il3t aber keine etwa ffir einen Literaturwissenschaftler verwertbaren Schlfisse zu." (42) Unfortunately the same holds for linguists. The conclusions reached by P. may be invalidated by the fact that the Nikon chronicle is not a homogenous statistical population and that possibly interesting informations are shrouded by an artificial subdivision of the text. I doubt very much whether the Nikon chronicle is a text that can be
REVIEWS
59
analysed by using statistical procedures simply because the chronicle does not represent a homogeneous statistical population. To analyse a text statistically that is not suitable for such analysis is of little use. The situation is aggravated by the fact that the statistical procedures applied are, to say the least, modest. P. operates with absolute, theoretical, and relative frequencies. He refrains from testing the results as to their significance, not because the numbers are too small for such texts but, as he says, because the significance is so obvious (13, n. 25). 8 As it stands we have to take P.'s word for it. ! have the impression that at least some of the quantitative deviations are statistically not significant. It would have been reassuring to have a test of significance for the major categories where such a test could be performed quite easily. In some instances P.'s line of reasoning strikes me as being very unstatistical. Thus he indicates (correctly) that an "AusreiBerwert" (his terminology, p. 179) in A. 5 is not to be overestimated since the basis for it is furnished by three occurrences only. He does not mention the fact, however, that A. 3 and A. 4 rely on a weak material basis, too (1 and 3 occurrences). This, unfortunately, is not evident from the tables since P. indicates relative frequencies when treating individual conjunctions; absolute numbers are only given at the beginning of each chapter. If the attentive reader wants to know how reliable the relative frequencies are he has to go back. Sometimes this does not help either since the tables at the beginning of a chapter are not complete and then one has to resort to the cumulative tables at the end of the book. 9 The other major problem of P.'s book is of a more linguistic nature. P. does not reflect in great detail on the grammatical conceptions underlying his analysis. Basically he uses a syntactic conception: clauses are classified according to the part of speech they represent in the sentence structure (3). The parts of speech, however, are defined according to the traditional system that is not homogeneous. Thus we obtain clauses that are defined syntactically (subject, predicative, object and attributive clauses) and others that are defined according to content (the adverbial clauses). This system is not as unproblematic as P. would have it (3). Often a clause can be classified according to both principles and may therefore belong to two different types of clauses simultaneously. This was evident to P. as can be judged from scattered remarks.~° Unfortunately P.'s conclusion is not that the classification is at fault but that additional distinctions are needed, distinctions that are, needless to say, arbitrary again. P. does not present a grammatical system of his own but quotes a variety of grammars. It is hard to understand why he bases his compound theory partly on grammars of the contemporary German language, espe-
60
REVIEWS
cially in difficult cases (cf. 73 seq.). It is true that the syntax of different languages is comparable, but that does not mean that there is o n e syntax for all languages.11 Using grammars of German or contemporary Russian in this case is another instance of making facts fit a theory. A more satisfactory solution would be to abstract the theory from the text. Otherwise one arrives at categories that do not correspond to the linguistic reality. A good example of this are the predicative clauses (27-33). Their definition is culled from Schaller, and he in turn is clearly influenced by Latin grammar tradition. It becomes quite clear, however, that such a conception is not applicable to the material without making many arbitrary decisions. Thus it seems difficult to draw a clear line between subject clauses and predicative clauses. The distinction is, it appears, "at times difficult" (28). Talking about subject clauses P. even admits that it is impossible to distinguish them "unambiguously" (15) from predicative clauses. Since it is difficult or even impossible to distinguish the two types of clauses, why distinguish them at all? Either our methods of analysis are insufficient, and then arbitrariness is a bad way out of a bad situation, or else the distinction is not relevant for the language and then it should not be made at all. Furthermore factual evidence adduced in the book shows that P. himself sometimes does not know how to distinguish predicative clauses from object clauses, cf. examples 77 and 197.12 These are the fundamental objections to be raised with regard to P.'s book. They are rather serious since they question the very foundations of the analysis: the text as object of analysis, the method chosen for the analysis and the categories used to present the results. Before assessing the book as a whole I shall deal with a few details that are also relevant. 19 seqq. In the chapter on subject clauses the numbers of predicate noun clauses are added to those of the subject clauses and together they form the basis for the diagrams, purportedly to "stabilize" the values. Why not treat them together? Is this not a tacit admission that the distinction is arbitrary and not founded on textual evidence? 51 (and similarly 90 and 155) P. says that the curve resembles a normal, or Gaul~ian, curve (his reference to tab. 249 is incorrect; it should be 250). I do not know why the author says this: it gives me the impression that he wants to emphasise the statistical nature of the distribution. This, of course, is nonsense: the tables P. is referring to do not indicate values a variable may take and it is therefore useless to look for normal distribution. 64 The parallelism of the development j a k o - (to and glagolati/re(i- govoriti/skazati is very interesting. P. does not comment on the paralle-
REVIEWS
61
lism, he merely states the fact. I suspect that this is an instance of the "Russification" of the language used in the chronicle. 88 The Nikon chronicle quotes (s.a. 6951) a canon of the Laodicean council introduced, as usual, by jako. It is questionable to treat such instances as examples of attributive clauses; in any case they may falsify the statistical picture: many quotations of the kind influence the percentage since only in A. 3, A. 4, and A. 7 the absolute numbers ofjako are higher than 20. 109 The sudden drop in the frequency ofpone~e in A. 3 intrigues P., and he hints at the converse development in the case ofjako (cf. 105 where there is no reference to 109). The explanation that we have a temporary drop is not convincing since this is not very likely to happen. Maybe one of the sources prominent in A. 3 is responsible for the development. 119 The observations regarding the substitution of CS conjunctions by their "Russian" counterparts are very convincing. I wish there were more conclusions of that kind in the book. 123/4 P. maintains that aJ(e li establishes, as do other combinations with a~(e, a reference to the context. In the case of ag(e ubo, however, the conjunction is supposed to separate the clause from the context. This also raises the problem how "complex conjunctions" are to be treated. In principle P. analyses all possible combinations separately but does not, as far as I can see, justify this procedure. In at least one instance he even violates the principle consistently, and that is in the case of zane(~.e) (112). A similar but isolated case is pone~e (bo) (69). It remains unclear, furthermore, why ague is treated as one conjunction (ag(e li, ubo, by, ge etc. being considered as variants), whereas (to and (toby, da and daby, jako and jako~e are kept separate. (Egda, by the way, is treated the same way as ag(e, cf. 158-161.) If the reason for this distinction is orthographic (depending on whether the conjunction is written as one or as two words) then P. merely reproduces the editor's usage, and this usage does not tell us anything about seventeenth century Russian. 135 budi is treated as a mere variant of budetb. Why this? Would not eda and egda deserve a similar treatment? 170 It is unclear why example 887 (Egda 5e) should be a representative of
vnegda ~e. 174 and 178 The artificiality of the separation of conjunctions with and without particles is demonstrated by 903 and 916. P. failed to provide a summary at the end. This is regrettable since the
62
REVIEWS
book is thus reduced to a heap of statistics, diagrams and examples. Interesting general observations or tendencies of development may be indicated in the sections but they are never put together to form a coherent picture. I have the impression that such tendencies are present in the Nikon chronicle (witness the development one might call "Russification") but it would take P.'s expertise to corroborate this impression. The final assessment, I am afraid, cannot be very favourable. Undoubtedly P.'s book is the result of a great amount of hard work, and his effort at providing exhaustive coverage of the material deserves praise. Unfortunately this effort does not bear fruit since the results suffer from the inadequacy of the methodology employed. The book is very useful as a quarry for examples of subordinate clauses but the numerical part of it must be used with great caution. And this is to be regretted: what good are numbers if they cannot be trusted without prior verification? Saddest of all: statistics live up to their bad reputation once more ("statistics may prove anything"). And this need not happen: statistics, I think, are an indispensable tool for historical linguistics and progress in this linguistic field will depend on the correct use of this tool. If we linguists do not learn to master statistics we shall forever be stuck to the impressionistic "method" of presenting chance observations as representative of the language as a whole. I look forward to the time when historical linguistics will have overcome these problems and will be able to draw reliable informations from ms. sources and to proceed to verifiable generalisation.
NOTES ~ In the case of certain categories of mss. from Russia the matter is not as simple; cf. infra. 2 It was published simultaneously as a dissertation (G6ttingen 1987) and as a book in the Berlin series. I used a copy of the dissertation but as far as I know the main body of the book is the same in both cases. The author is referred to as 'P'. 3 Cf. Br~uer's book on subordinate clauses in OCS and Old Russian (BrS.uer 1957). 4 The numbers in brackets refer to the pages in P.'s book. 5 The problem of the statistical nature o f langue and parole is very complex and cannot be treated in detail here. Suffice it to refer to l l e r d a n ' s remarks in the revised version of his book (Herdan 1966, VI). My own position is somewhat different. I take a text ideally to be a r a n d o m sample of the complete parole of an individual and not of langue. (I am too much of a linguist to consider langue to be a statistical population.) 6 In this case P. is very eclectic: A. 5 has only 3 occurrencies and its validity is also questionable. 7 Cf. on this Kloss 1980, 112-130. 8 Elsewhere the reasoning is different: on p. 133 he says that the theoretical frequencies are too low to allow for a test. 9 It is true that P. sometimes refers the reader to these tables (cf. 17, 34, 41) but it is unfair practice to give relative frequencies and to c o m m e n t on the unreliability of these frequencies only if they do not fit into the general picture of certain developments. Either all relative frequencies based on few occurrences are marked as being unreliable or none. The author's eclectic procedure betrays a tendency to make facts fit the theory.
REVIEWS
63
10 Cf. the reference on p. 160 (temporal egda) to p. 83 (attributive egda). 11 In my eyes it is not even possible to apply concepts of the grammar of contemporary Russian to the language of the Nikon chronicle. ~2 In the former only the clause contains a predicative, and I do not see how subject raising changes anything.
BIBLIOGRAPHY Br~iuer, H.: 1957, Untersuchungen zum Konjunktiv im Altkirchenslavischen und im Altrussischen. Teil 1. Die Final- und abhiingigen Heischesiitze, Wiesbaden ( = Ver6ffentlichungen der Abteilung fiir slavische Sprachen und Literaturen des Osteuropa-lnstituts an der Freien Universitiit Berlin l 1). Herdan, G.: 1966, The Advanced Theory of Language as Choice and Chance, Berlin etc. (= Kommunikation und Kybernetik in Einzeldarstellungen 4). K~occ 13. M.: 1980, Hu~conoeocuft caoO u pycc~ue :temonucu X V I - X V I I ae., MocKBa.
Universitiit Bamberg
ROLAND
MARTI
64
REVIEWS
CTaRaUKOBa, E~eua B.rla~IMrlpOaHa, B.~uauue amleumnofi c u c m e . ~ b l n a ~onoaoeu,~ec~y~o (ua ~uarnepuaae ucmopuu Onyx ~oue.~ 'muna o' a pycc~o~ a~t,oce). )IuccepTarana na concraHrie y~eno~ cxeneHrl ~an~in~laTa qbri~o~oru~ec~nx nays, Moc~Ba 1984, 246 CTp. CTa)IHUKOBa, E~erta B~a~2wMupoana, Bauauue atcgenmno~t cucmeMbl ua ~onoHoeu~ec~y~o (ua ~uamepua~e ucmopuu Onyx ~oue~ 'muna o' a pycc~o~t a 3 ~ e ) . AaTopeqbepaT ~iccepTattnn Ha COUCKaHueyqeHO~ cTeneHn ~ar~arl~aTa qbn~o~ornaec~nx Hays, Moc~aa 1984, 24 cTp. Tyra~, E~ena B~a)IriMupoBna [= CTa)IHHKOBa], "Paa~unenrie ~IByX~OHeM 'Tuna o' B cTapOBe~uropyccKo~ pyKonncrx XVI B." B ~H.: £13b~oaaa npa~mu~a u meopua a3m~a, MOCKBa 1978, B~,In.2, 70--81. Tyrafi, Eaena B.rla~II,IMUpOBHa [= CTa~IHI,IKOBa], "K BOnpocy 06 9Boa~orauri )IByx qbOHeM 'vnna o' B XVI B.", ~u~o~oeuuecnue naynu 1983, No. 2, 62-68. That E.V. Stadnikova's dissertation on middle Russian manuscripts distinguishing two/o/phonemes, c l o s e d / 6 / a n d open/~/, remains unpublished is a misfortune for Slavic accentology, qhis work provides valuable new data and an excellent analysis of the material, and demonstrates a thorough knowledge of the scholarly literature. The abstract and published articles listed above outline certain important aspects of Stadnikova's research, but the wealth of carefully organized data presented in the actual dissertation should be made available. Stadnikova's dissertation consists of an introduction (3-33), a description of her sources (34-65), a discussion of the distinction of two/o! phonemes in sixteenth-century Russian written records (66-121), and an analysis of the further historical evolution of this opposition (122-67). There is a brief summary of her conclusions (168-71), followed by two appendices, the first a dictionary of morphemes providing data on the representation of /o/ phonemes (172-227), and the second a list of relevant literature (228-46). Her original contributions include not only important new data, but the presentation of these data as a beautifully organized practical reference work that should be on the desk of every Slavic accentologist. Furthermore, the value of Stadnikova's dissertation goes beyond accentology, as the title suggests; her investigation of the influence of accentuation on non-accentual features will be of interest to every student of Russian historical phonology.
Russian Linguistics 13 (1989),
REVIEWS
65
As Stadnikova observes in her introduction, the existence of Russian dialects that distinguish two/o/phonemes has been known since the nineteenth century (7), but there have been few attempts to account for this phenomenon until recently. Her description of the history of the problem omits no relevant sources, and she provides a survey of earlier dialect (7-8, 15-16, 20-23) and manuscript studies (16-18, 23-24), a discussion of earlier analysis (7-15, 18-20), and an introduction to her theoretical framework (26-29). The inadequate attention devoted to the question o f / o / p h o nemes in Russian in earlier works derives both from a failure to recognize the value of manuscript studies for such investigations and from a prejudice that the issue was intractable (24-25). Stadnikova demonstrates that this pessimism is unwarranted, and that the considerable recent progress in Slavic accentology, particularly extending the work of R. Jakobson and V.A. Dybo (26), provides a foundation for resolving such questions. Within this theory, a closed/6/phoneme is the reflex of inherited/o! under stress (and, therefore, only in orthotonic words), while open /~/ is the reflex of inherited/o! not under stress (in both orthotonic words and enclinomena) and o f / o / from earlier/e/, /~,/, or/~,/ (27). A common Russian seven-vowel system, which distinguished the reflexes of inherited/6/and/e/ as well a s / 6 / a n d / ~ / , was subsequently simplified in most Russian dialects and in the literary language (20-21). Of crucial importance for Stadnikova's study of accented manuscripts is her recognition of the optional expression of the opposition between the two/o/phonemes. Following N.N. Durnovo and A.A. Zaliznjak, she acknowledges t h a t / 6 / m i g h t be freely replaced b y / ~ / i n manuscripts, while pointing out that neutralization does not occur in the other direction: /a/ is not normally replaced b y / 6 / ( 2 9 , 42). The ratio of the number of places where/6/ is actually spelled to the number of places where it is expected in the system under investigation is called the coefficient of the expression of/6/. Stadnikova's first chapter describes the orthographic systems of eight sixteenth-century manuscripts and printed books and uses both extralinguistic and linguistic and orthographic features to localize them in several dialect areas. Extralinguistic features include information about the scribe and the place where the text was copied. Linguistic features include accentual features (based on the dialect maps in section 2 of Zaliznjak's 1985 monograph, Ot praslavjanskoj akcentuacii k russkoj) and different distributions of the two/o/phonemes. The core data in the dissertation, taken from ten systems in eight manuscripts and books, are distributed as follows. Of the undated and anonymous publications traditionally called uzko~riftnoje (Uzk.), sredne~riftnoje
66
REVIEWS
(Sred.), and ~iroko~riftnoje (gir.) (Nos. 1-6 in A.S. Z~rnova, Knigi kirillovskoj pegati, izdannye v Moskve v X V I - X V I I vv. Svodny katalog, 1958.) and probably published between 1553 and 1564, only the first shows a consistent distinction of/o/phonemes. This text has a coefficient of 96.4 % and displays both eastern and western accentual features, suggesting that it can probably be attributed to a Central area (35-43). In the Skitski Paterik portion of a large manuscript dating to the middle of the sixteenth century (Pat.), Stadnikova identifies two scribes (Pat. A. and Pat. B.), both of whom show dialect features suggesting the western portion of the northwestern dialect area; Pat. B., in particular, shows features peculiar to Pskov. The coefficient for Pat. A. is 97 %, while for Pat. B. it is 70 % for roots (43-49). A Lestvica manuscript from 1558 (Les.) has a coefficient of 86 % for roots and can be assigned to the south or southwest of Moscow (49-52). A gospel that can be attributed to the first quarter of the sixteenth century (Muz.) and has a coefficient of 91-92 % shows northwestern (particularly Novgorod) features, with occasional eastern influences (52-57). A miscellany from the 1560s shows Muscovite features in the parts attributable to two scribes (X1. A. and X1. B.), with some features in X1. B. that point to an area southwest or west of Moscow. The coefficient for X1. A. is 93 %, while for X1. B. it is 80 % (58-60). A gospel from 1549 ((~er.) has a coefficient of 60 °/0 and shows a system similar to that in Muz., representing the eastern edge of a northwestern area (60-62). Stadnikova identifies five orthographic systems used to reflect the distinction b e t w e e n / 6 / a n d /o/in her core manuscripts: 1) a kamora system, where /6/ is represented by ~ and ~ and /o/ is represented by o and o9 (X1. B. and (2er.); 2) an omega system, w h e r e / 6 / i s represented by o9 and /o/ is represented by o (Les.); 3)an "uzkoliternaja sistema," where /6/ is represented by narrow o and /~/ is represented by broad o (XI. A.); 4)a "~irokoliternaja sistema," where /6/ is represented by broad o and /~/ is represented by narrow o (Uzk., Pat. A., and Pat. B.); 5) and a combination of types 1 and 3, w h e r e / 6 / i s represented either by narrow o or by any ~ or &, while /~/ is represented by broad o or co without kamora (Muz.). While the first two types of systems were previously noted by other scholars, one of Stadnikova's original contributions is the discovery that broad and narrow o may also be used to distinguish different/o/phonemes (6). The data from the core systems are supplemented by a variety of material from texts analyzed by other scholars and from dialect studies (62-65). The supplementary manuscripts, which date from the fourteenth to seventeenth centuries, attest kamora, omega, and ~irokoliternaja systems, as well as combinations of an omega system with uzkoliternaja and girokoliternaja ones, and they are attributable to northwestern, southwestern, southern,
REVIEWS
67
western, and central zones. In one manuscript used by L.L. Vasil'jev in his famous 1929 study o f / o / phonemes in Russian manuscripts, Stadnikova observed that there is actually a combined uzkoliternaja and omega system, while Vasil'jev had noticed only the omega system (64). Dialect data are from northeastern (totemski govor), central (Leka), and southeastern (Novoselka, zadonski govor) zones. The second chapter arranges the material from the principal sources to illustrate that the basic distribution o f / o / p h o n e m e s in all morphological classes (roots, suffixes, and desinences) is in accord with Jakobson's and Dybo's theory, and this chapter provides more examples of the accentual behavior of some morphemes than the appendix. The only weakness of this section is occasional inattention to details, so that the discussion of the suffix -~,cma-(o) (88) correctly cites Dybo's observation that words with this suffix are orthotonic, but fails to comment on the occurrence of barytone forms where oxytona are reconstructed for Common Slavic. For example, Stadnikova's presentation suggests that ~ cpOObcma~ (88) is orthotonic in conformity to the CmS1 pattern; although it is orthotonic, the place of stress attests a Russian innovation, which, furthermore, is not reflected in the Russian literary language for this lexeme. (On the development of-t, cma-(o) in Russian, see Zaliznjak 1985, 81-82 and 149-50.) While the second chapter emphasizes the inherited pattern of the opposition b e t w e e n / 6 / a n d / a / t h a t underlies all the texts under investigation, the third chapter traces the evolution of this opposition in the different systems reflected in the written sources and highlights individual deviations from a common original distribution. The evolution of these systems can be considered a development from a system with a strictly phonological distribution of/o/phonemes towards one where they are in free variation, producing simplified six- and five-vowel systems (124); the attested intermediary stages show variation restricted to certain morphemes. These developments may result from analogy within a single dialect, due to influences either between orthotonic words and enclinomena in different accentual paradigms or between the different elements of a single paradigm, in which some forms may be stressed on the stem and others on a desinence. This type of analogy accounts for the great fluctuation between /6/ and /a/ in desinences (161). Another source of change is influence between neighboring dialects, particularly in the central area, which, as a geographic and political center, was exposed to wider influences than peripheral dialects (123-24). Stadnikova treats specific developments individually, with copious examples from written sources (126-66). Principal innovations in the distribution of/o/phonemes include the replacement o f / 6 / b y / ~ / i n the nNAsg
68
REVIEWS
desinence -6; the appearance o f / ~ / i n the Dpl desinence -6m~ and, sporadically, o f / 6 / in the homophonous Isg desinence; the appearance o f / ~ / in orthotonic Gpl forms like ep66r, and, conversely, o f / 6 / i n former enclinomena like ep66u, representing a leveling of variation within a root; the appearance o f / ~ / in orthotonic long forms like nbeaeo and /6/ in short forms like n6so (originally an enclinomenon); the emergence o f / 6 / i n forms like m6, tom6, ~tm6, crab under the influence of neuters ending in/6/, spreading to K6~cc)o, rnbu; considerable variation between /6/ and /~/ in desinences as a result of various analogical influences; the emergence o f / ~ / i n Isg -6to as a result of the definalization of stress; the appearance o f / ~ / i n 66:~cuu, e6um,, 66~uu (66~,tuuu), edpuu (e6ptuuu), and other words as a result of a late retraction of stress; and various other confusions o f / 6 / a n d /a/, including some spellings o f / 6 / as a reflex of/~,/ (168-69). The effect of all these changes is a tendency toward the elimination of regular alternations o f / 6 / a n d / a / w i t h i n individual morphemes. Specific dialect features include stressed/~/in nNAsg -6 (south), fIsg -bin (northeast), Dpl -6m~ (west), Gsg -e6 (south and southeast), and unstressed /6/ in Gsg -co (west). Finally, while vacillation between /6/ and /a/ is widespread, it is most characteristic of the central dialects (170). One potential difficulty with the analysis undertaken in chapter three is distinguishing situations where a spelling o f / ~ / in place of expected /6/ represents a step towards the elimination of the opposition in a certain morpheme from situations where it represents a neutralization that is expected according to the principle of optional expression. The most reliable evidence for the actual elimination of an opposition is a situation where the replacement o f / 6 / b y / ~ / i s accompanied by a replacement of/~/ by/6/, since the latter replacement is unambiguous. While the application of such considerable data to demonstrating the original distribution o f / 6 / and /~/ in Russian and accounting for its further evolution is a major advance for Slavic accentology, the greatest practical contribution of Stadnikova's dissertation lies in the organization and presentation of her material. The dictionary that constitutes her first appendix is an excellent guide to the distribution of the two/o/phonemes in individual morphemes (roots, suffixes, and desinences), and is based on manuscript and dialect information from the fourteenth through twentieth centuries. An entry consists of a head word with nested subheadings, all supplied with indications of the sources that attest both expected and unexpected/o/phonemes. For example, the entry for mO6Oab, where /~/ is expected as a reflex of /~,/, reads as follows (198):
REVIEWS
69
.,rho6o6~, n npoa3B. (*~) -6- qac., AMdp., TOT. Xa.E., 3a.; 6/o-Mep. 3x/llx, C6. 16x/24x, FIaT.A 15x/6x+lx, FIaT.E 8x/19x, )2oc. 4x/4x+19x, Ylec. llx/39x+4x; o-l-I. Ix, ~nT. lX+2X, H. 1306, Xp. 6x, Y3K.2766, Yler., P~I3.Cp. 6 B ~ypn. 2x.
The note (*~,) refers to the origin of t h e / o / p h o n e m e in this word. It is followed by a list of places w h e r e / 6 / i s attested, including the core manuscript XI.B., the secondary manuscripts Cas. and Amf., and the dialect studies Tot. and Zd. This is followed by a list of sources where an alternation between /6/ and /o/ occurs, with information on the number of occurrences. The plus sign in some counts of citations with /o/ separates examples that occur with accent marks, on the left, from those that occur without accent marks, and are therefore less reliable, on the right. A list of sources where o n l y / ~ / i s attested follows, and the entry concludes with a reference to two occurrences of /6/ in a 1932 article by Durnovo. The appearance o f / 6 / w h e r e / ~ / i s expected in this lexeme is treated in greater detail, with copious specific citations, in chapter 3 (138-39). Since Stadnikova's dissertation has not been published and remains inaccessible to most readers, the abstract and the two articles listed above are the best sources of information about her work. The abstract contains a general summary of the contents of the dissertation; it presents the basic principles observed, provides brief descriptions of the manuscripts and books used, reproduces some of the important tables, and outlines the conclusions. While it offers a convenient account of Stadnikova's methodology and conclusions, the absence of the valuable dictionary and lengthy lists of citations make it of less practical use. "Razli6enije..." consists of a description of Muz., which, as noted above, combines an uzkoliternaja sistema with a kamora system, a distribution first discovered by Stadnikova. In addition to offering evidence of a complex and previously unknown orthographic system f o r / o / p h o n e m e s , this manuscript has an extremely high coefficient for the representation o f / 6 / . As expected, it is highest in affixes and native roots (96.4 % after a single consonant letter, 91.4 % after two consonants) and lower in nominal desinences (where there are more models for analogical change) and borrowed roots. (The specific figures differ somewhat from those given in the corresponding section of the dissertation, apparently because different data were excluded from the computations as unreliable.) Individual peculiarities, including differences between Muz. and the manuscripts discussed by Vasil'jev, are illustrated; the most important of these is that Muz. is more likely to spell /~/ where /6/ is expected in nominal desinences than Vasil'jev's manuscripts. This article can serve as a summary of the portion of chapter one of the dissertation devoted to Muz.
70
REVIEWS
"K voprosu..." discusses the phonological developments that produced the general Russian opposition between /6/ and /a/ and then traces the subsequent loss of this opposition in specific positions, with examples from the manuscripts, books, and dialects used in the dissertation. Thus, this article presents some of the general Russian information from chapter one of the dissertation as well as an outline of the conclusions developed in chapter three, although it is not, of course, as richly documented. To summarize, Stadnikova's research in the history o f / o / p h o n e m e s in Russian is a major contribution to Russian historical accentology and to the historical phonology of Russian in general. Although the copious data included in the dissertation have not been published, the abstract and the two articles discussed above provide a useful introduction to her sophisticated theoretical framework, her conscientious methodology, and the valuable conclusions she reaches. Stadnikova's dissertation attests the same high standards as other recent accentual studies by linguists associated with the Institute of Slavic and Balkan Studies in Moscow (including A.A. Zaliznjak, who directed Stadnikova's dissertation, V.A. Dybo, R.V. Bulatova, and K.K. Bogatyrev), and its publication would be a valuable service to Slavic accentology.
Harvard University
DAVID J. BIRNBAUM
REVIEWS
71
Mayo, Peter, The Morphology of Aspect in Seventeenth-Century Russian (Based on Texts of 'The Smutnoe Vremja'), Slavica Publishers, Columbus, Ohio, 1985, XI+234 pp. Peter Mayo's monograph has its origin in his doctoral dissertation presented at the University of Sheffield, England. It is a detailed analysis of the morphological category of aspect in texts from the first half of the 17th century. Since previous research has concentrated mainly on the administrative language, Mayo has selected a text corpus representing the "bookish-ecclesiastical style of the literary language" for his painstaking scrutiny of 37,000 verbal forms. Mayo's book consists of an 'Introduction' (pp. 1-19) followed by five core chapters presenting 'Non-prefixal pairs' (pp. 20-28), 'Prefixal pairs' (pp. 29-56), 'Parallel prefixation' (i.e. "the creation of compound aspectual pairs from two groups of simple verbs in which the addition of a prefix did not necessarily result in perfectivization", p. 57) (pp. 57-106), 'Prefixal-Suffixal pairs' (e.g. -ey6umu: u3ey6umu/uJey6~amu) (pp. 107-72) and 'Biaspectual verbs' (pp. 173-92). The monograph rounds off with 'Conclusions' (pp. 193-210), 'Notes' (pp. 202-06), 'Bibliography' (pp. 207-17) and a very useful 'Index' (pp. 218 34) of all verbs referred to in the book. In the 'Introduction' Mayo surveys briefly the often conflicting views about the origin and development of the Russian literary language and places the language of his text corpus in the context of this development. Then he approaches the problem of aspect. Paragraph 1.2 'The origins and development of the category of aspect in Russian' (pp. 5-12) is a succinct expos6 of a complicated matter. From the general Mayo proceeds to the particular: 1.3 'The category of aspect in the smutnoe vremja texts' (pp. 12-19). The five core chapters are similarly constructed. The bulk of each chapter consists of lists of aspectual pairs, cited in the context. It adds to the interest of this "dictionary of aspect" that every aspectual pair is considered in a diachronic perspective by comparison with modern Russian. At the beginning of each chapter the authors defines the type of aspectual pairs to he treated and discusses problems of classification. In the chapter 'Prefixal pairs' Mayo, unlike Maslov and Isaeenko, sides with his fellowcountryman Forsyth in the recognition of prefixal pairs (e.g. ~uOT~mu/yauO~mu, p. 37). The last chapter is a lucid presentation of 'Biaspectual verbs'. These verbs are presented under three headings: "a) verbs still biaspectual but no longer so in modern Russian; b)those previously biaspectual but no longer so in modern Russian; and c)some compound verbs which also show signs of fluctuating aspect" (p. 175).
Russian Linguistics 13 (1989).
72
REVIEWS
Not quite surprisingly Mayo comes to the conclusion that "[...] in broad terms the findings of this study agree with those of [...] Lopuganskaja (1967) for the administrative style" (p. 193) for the same period. The merit of Mayo's study resides not primarily in new theoretical insights, but in its thorough and scholarly treatment of an impressive quantity of data. Unlike many lightfooted theoretical writings on Russian aspect, Peter Mayo's empirical study will have lasting value and serve as a useful point of reference for future studies on aspects within a diachronic perspective. Universitetet i Bergen
ALF GRANNES
REVIEWS
73
Pycc~ue napoc)n~,te eoaopt, t. flunesoeeoepa~u~tectcue ucc:teOoeanun, n o ~ pe~t. P. I4. A ~ a H e c o s a , M o c K , a , H a y ~ a , 1983, 141 c~p. C 6 o p n n r 1JOCBfl~tleH aKTya.qbHb~M BoI~pOCaM coBpeMeHHO~ OtH4caTeY~bHO~ ~ m a n e r T o n o r u n n COCXOnT n3 15 cxaxe~. Pa6oxs~ c 6 o p a n r a ~ a r T n q e c r n pacnpe~e~amTC~ n o qeTblpeM OCHOBHblM pa3~eaaM: CTaTbfl C . B . BpoM~efi ' O M o p ~ o ~ o r n q e c ~ o M Bb~nyc~e ~ n a ~ e ~ T o ~ o r n q e c r o r o aT~aca p y c c r o r o ~ 3 s i r a ' (cxp. 3-24) c n y ~ u T O6m~M n p e ~ a c n o a n e M , Tar r a t o n a ~acaeTc~ p a 3 n ~ x a o n p o c o ~ ~ a ~ e r T n o ~ M o p ~ o ~ o r n n ~ MOp~OHO~Or~n a CBfl3~ C 3aBep~eH~eM BTOpOFO BbInycKa ~uaaenmoaoeuuecnoeo amaaca
pyccnoeo n ~ m a ; a x o p o ~ p a 3 a e a (cxp. 25-81) COCTOnT n3 mecTn CTaTefi, p a c c M a T p n a a ~ m n x p a 3 ~ e ~ o ~ e x n q e c ~ n e fla~ennfl p y c c r n x ~na~e~xoa; xpexn~ p a a ~ e n (CTp. 82--107) a r n ~ q a e x nflTb cza~e~, n o c a s m e n n s i x n e r o TOpbIM ~ o p ~ o a o r n q e c r n ~ Oco6eHHOCTflM pycc~nx ~ n a ~ e r T o a ; B nocne~heM, qeTaepxoM p a 3 ~ e a e aaTopbI xpex cxaTe~ n a T e p e c y ~ T c n OT~e~bnS~Mn a o n p o c a M n n e ~ c n r o ~ o r n n . H p n ~ o ~ e n n s i e CeMb ~apT n a a ~ c ~ p n p y ~ x e p p n x o p n a n b n o e pa3MemeHne n e c r o a s r n x H3yqennblX anaaerxHbIX BapHaHTOB H ~ a ~ T npe~cTaB~eHne o6 aT~ace. 1. BcvynnTenbnaa CTaTba C. B. BpOMJ~efi (CTp. 3--24) 3HaKOMI4T qHTaTeYlfl C zOn~ro qvo 3 a a e p m e n H b ~ ~ o p ~ o n o r n q e c r n ~ a~myc~oM ~uaaenmonoeu~ecnoeo amaaca pyccnoeo n3bma ( ~ A P f l ; abIn. II); 3xo n o a o e a o c x n ~ e Hne c e r x o p a ncTopnn fl3bI~a n a n a a e ~ T o n o r n n ~HCrnTyTa p y c c r o r o s3b~a AH CCCP aa~eTc~ npoaon~enne~
n e p a o r o a s m y c r a (~oneTn~a)
n r p e T ~ e r o (CanTarcnc-nercnra). A a T o p , caMa y q a c x a o a a a m a ~ a cocT a a n e n n n a a n n o r o aT~aca, n a n a r a e T np~nunns~ pa6oTSI n a a n n ~ , OCnOBanUO~ Ha Teopnflx P. ~ . A a a H e c o a a , CCb~ancb n p n aTOM Ha 6 o ~ b m o e ~ o a n ~ e c x a o npnMepoB. B 3 a r ~ q e H n e OTMeqaeTcfl, qTO apeaab~ MOp~Onornqec~nx, ~ o n e T n q e c r n x n cnnTa~cnqec~nx OTanqnTe~sns~x qepT coan a t a n T , XeM CaMb~M n o a ~ p e n ~ n n o 6 ~ e n p n 3 n a n n b ~ e pycc~ne ~naaerTns~e n3oraoccb~; a n p n n o ~ e n n n ~ a a n n o ~ CTaTbe ~aeTcfl n o n u u ~ ~apT a x o r o a x o p o r o a s i n y c r a ~ A P f l c y r a 3 a n n e u a a T o p o a .
nepeqenb
2. E.F. 13ypoaa, 'FIpoxexHqecKHe r:~aCHble a no3~4t~m4 n e p a o r o npe~ly~tapnoro c n o r a a p y c c r n x r o a o p a x ' (CTp. 25--35): no~Ipo6Ho onHcbmaeTcn ~tB~IeHHe, TI4HHqHOe ~SIfl lOFO-3al-la~HO~ ,/II4a~IeKTHO~ 3OHbl H )2.1Dt lOFa [IcKoBCKOI~ O6flaCTI4; CXO~CTBO a p e a a a npoTeTr~qecrnx F.rIaCHbIX l - t o npe-
~lyztapnoro c n o r a c TeppHTopne~, r~le Ha6n~o)laeTcfl ~IHCC~IMrUIflTrIBHoe aKaHbe~ ]loKa3bIBaeT, qTO o 6 a flB.rleHH~/ I4CTOpl4qecKI4 CBfl3aHbl.
P . ~ . Flayqbom,Ma, 'HeroTopb~e OCO6ennocTn cau~lxri a ceaepr~opycc~nx r o a o p a x ' (cxp. 35--44): aBxop paccMaxp~Baex aZO eme Mano n3y~enHoe a
Russian Linguistics 13 (1989).
74
REVIEWS
~tna:~erTax a a ~ e n n e , cpaBanBaa e r o c C~4CTeMaMtl p y c c r o r o anTepaTypHOFO H y r p a n n c r o r o
fl3blKOB.
B.H. TenaoBa, ' O BOra~n3Me 3 a y ~ a p H o r o KOHe~HOFO OTKpblTOFO c ~ o r a noc~e Taep~b~X cor~acHbIX B a r a ~ m u x r o B o p a x p y c c r o r o fl3b~a' (crp. ~ 5 4 ) : y c ~ a n a ~ n ~ a e ~ c ~ , ~ o ~ ~ a r o ~ n o a n u n ~ p e ~ y a n p o a a a a ~ [~] a a ~ e T c a 6o~ee nCrOHHUM, n o c p a a a e n n ~ c [a]. K . O . 3 a x a p o a a , ' B u a ~n n e p e x o ~ a > o a 3 a y a a p n o M n o ~ o x e a n n ? (ha MaTepna~e ~HHHblX o ~ o n q a n n ~ r ~ a r o ~ o a I c n p a x e a n f l a r o a o p a x B ~ a a n M n p c r o - H o a o ~ x c r o ~ r p y n n u ) ' (CTp. 55--60): a a T o p n p e ~ n o ~ a r a e T , qTO a paccMaTpnBaeM~iX r o a o p a x , B ~aHno~ n o 3 n a n n , ~ r a n b e a a ~ e T c a ne p e 3 y ~ x a x o u ~ o n e x n q e c r o r o n 3 ~ e n e n n n /e/ a /o/, a TO~brO pe3y~bxaxoM rpa~uaxnqecro~ ana~ornn. ~.~. Kacaxrnn, ' O c o 6 e n n o c x n Bo3~e~CTBnfl M a r r o r o 3a~Hefl3b~qnoro c o r ~ a c n o r o , c x o ~ m e r o n o c a e x a e p a o r o c o r a a c a o r o , na n p e a m e c x a y ~ m n ~ r ~ a c n u ~ a p y c c r n x r o a o p a x ' (crp. 60-72): o 6 c y ~ a a ~ x c a yc~oanfl ~ r a n ~ a a a x o ~ n o ~ n u n n a caftan c a c c n ~ n a a u n e ~ c o r a a c a u x ; n c n o a ~ 3 o a a a a o6mnpnaa ~nxepaxypa. H . H . H m e n n q r o a a , ' M n n r a o r e o r p a ~ n q e c r n ~ n cxaxncxnqecrn~ ana~n3 CTfl~ennfl B p y c c r n x r o B o p a x (o ~ o ~ n ~ e ~ c a o ~ n o a x o ~ e ~ n c c ~ e a o B a n n ~ ~ 3 u r o a u x ~ a ~ e n n ~ ) ' (cxp. 72-82): a a x o p n p e a n o ~ a r a e x , qvo p a ~ a n q n o r o p o a a p a c x o ~ e H n f l a pe3y~bTaTax c x a ~ e n n f l y r ~ a r o a o a n n p n a a r a Te~bnblX CBn~eTe~hCTBy~T O TOM, qTO, ~oHeTnqecrn~ n o n p o n c x o ~ e unto, n p o u e c c c x f l ~ e n n a 6 ~ c c a u o r o n a q a a a ~ o p ~ o ~ o r n 3 n p o a a n n u ~ ; ~ p o ~ e x o r o , r a t h e p a c x o ~ e n n f l ~ o r y x O6~CnflTbCfl n a o 3 a e ~ c x a n e ~ n a n p n ~ a r a x e ~ b n b i e M e c x o n M e a n o r o cr~onennfl. 3. t A.Id. C o n o r y 6 , ' O CrlrtI
TOpbIX HCKOHHblX Tnnax c~Ionertrifl; ocTa~bnbxe qbopMbI Br~noqaroTcfl B aTOT n p o u e c c y ~ e B 6 o n e e nO3~Hee Bpe~a, o qeM CBn~eTen~,CTByeT TO, qTO OHri He npe~cTaBn~OT T a r n x m n p o ~ n x a p e a n o a , r a ~ o6tune qbopM~t P-~-H naae~e~. H . A . 13yrpnncra~, ' Y a a p e n n o e o r o a q a a n e -u a n p e a a o ~ n O M na)~e~e CylIIeCTBHTeJIbHbIX BTOpOFO T v t n a CK3IOHeHH$t B p y c c r n x ~na~e~Tax' (CTp. 89-92): ~aeTca no~tpo6noe onncanHe T e p p n T o p n a ~ b u o r o p a c n p o c T p a n e n n a a a a n o r o o r o n n a n n f l ; aBTop npeano~IaraeT, HTO oKonqanHe ( - u ) a a ceBepo-3anaaHo~ TeppnTopnn nMeeT p a 3 n o e n p o n c x o ~ e r m e : ( 1 ) M o p qbonornqecroe ( n o n BYIrI~IHneM M~trKO~ pa3HOBH~HOCTH CK.rIOrteHH~I Ha Taepay~o) a npe~te~ax n c r o a c r o r o a p e a ~ a p a c n p o c T p a n e n n a ~ a n n o r o
REVIEWS
75
OKOHqaHrlfl rl (2) qboHeTnHecKoe (B pe3ynbTaTe o 6 m e r o n e p e x o a a ~ B [//]) B npe~e~ax Yla)IorO-OHe~CKOFO apea~ia. H . A . 3-InnoBcKa~, 'tl~opMa pO~HTe.qbnOFO n a ~ e ~ a MHO~eCTBeHHOFO qHc~a cymecTanTen~Hb~X ~ e H c r o r o p o ~ a c OCHOBO~ Ha n a p n b ~ TBep~bI~ c o r ~ a c H s ~ ' (CTp. 93-99): caMoe pacnpocTpaHeHHoe a r o a o p a x ~aaeHne TUn 6 a 6 y ~ n a > 6 a 6 y m n o a ; cpaBnnTenSnO p e t e u a 6 a ~ a e T c s ynoTpe6neHHe OKOHqaHH~ -e~ B TOM ~ e na~eme. PaBBnTHe ~OpM C OrOHHaHHeM -Oa CBfl3aHO C TeppnTopne~ ~FO-BOCTOHHbIX FOBOpOB pyccKoro fl3b~ra, r a e 3Tn ~OpMbl BnepBbIe HOflB~HCb Ha TeppnTopnn pfl3aHCKHX FOBOpOB. 3 . H . By~aToaa, T . ~ . C v p o r a H o ~ a , ' F n n o T e 3 a C.H. O6HOpCKOrO O n p o n c x o ~ a e n n n OKOHqaH~ -y~, - a ~ , -au, -aMu B ~opMax npn~araTe~bHblX B caeTe aaHHmX ~ n a n e ~ T o a o r n q e c r o r o a x 3 a c a pyccKoro fl3bIra' (CTp. 100--104): O 6 n o p c r n f i O6~aCHa~ aTH oronqaHnfl ~ H o p y c c ~ o r o HapeqHfl ~HCCHMH~flTHBHbIM HpHH~HHOM opFaHH3a~HH BOKa~H3Ma, ~e~CTBOBaBmHM B n e p n o ~ BO3HHKHOBeHnfl aKanbfl; BonperH 3THM B3r~fl~aM aBTOpb~ a o r a 3 s I B a ~ T , qTO B 6eBy~apHb~X oronqaHnflX B a ~ H e ~ m y ~ p o ~ b cb~rpa~o a3anMO~e~cTane 6eBy~apHbIX O~OHHaHH~ cyH1eCTBHTenbHSIX ~ npH~araTe~bHb~X; TarnM o6pa3oM, 3ByrOaO~ BH~ pacCMaTpnBaeMb~X OKOnqaHnfi MOt onpe~e3flTbCfl ~ a r pa3~qHbIMn ~OHeTHqeC~HMH ~aKTOpaMH, TaK H COqeTaHHeM ~OHeTHqeCKHX H Mop~o~oFHqeCKHX HpHqHH; AnccnMn~a~ns ~n~aeTc~ 3 ~ m b O~OOfi ~3 3V~X np~q~n. E.B. HeM~eHKO, ' K ~ o n p o c y o6 nHHOBa~aX B ~ o p @ o ~ o r n a n p ~ a c v a f i (o nO~HmX ~OpMaX n pycc~nx r o n o p a x ) ' (cvp. 105-107): p a c c M a ~ p ~ a e v c s y n o v p e 6 3 e ~ n e n o ~ n o ~ @OpMbl c v p a A a v e ~ o r o npn~acv~s npome~mero ~peMeH~ YaM, r~e o x n ~ a e v c a ~ p a v ~ a a ~ o p M a (~ npea~Kav~n~ofi ~yH~~ n ) ; a ~ v o p RyMaev, qvo a~eKTHBa~H~ He BCeV~a 06~flCH~eT TaKOe flB3eHHe; Hepe~KO 3TH o6OpOTM o 6 ~ a ~ a m T F3aFO~bHbIM xapaKTepOM: cp. KOMy Kdm~ ~aOenbt~? ~CTaHaB~HBaeTC~ CHHTaKCHqeCKH~ n a p a ~ 3 e 3 a 3 M c ynoTpe6~eHHeM ~e~CTBHTe~bHOFO HpHqaCTH~ H p o m e ~ m e r o BpeMeHH THHa ona e ~ e a~epa npu6~amaa; 3TO M o p ~ o ~ o r a ~ e c ~ o e ~ o ~ o o 6 p a 3 o ~ a H a e MOF~O BO3HHKHyTb Ha OCHOBe n p e a n ~ a v a n ~ o r o a e e n p n ~ a c v ~ a na -emu. 4. IO.C. A 3 a p x , 'CymecTBnxenbnb~e na -a c 3rcnpeccHanbiMn cydpqbn~caMn Trma 'r~acH~+3a~nene6nb~ cor~acn~' a pyccr~x roaopax' (CTp. 108-121): paccMaTpnaamTc~ c n o ~ a v n n a ~pamnxa < 6pare, no:t~xa < ~ o d o m b , ~ a ~ e n ~ e a < ~ a M e ~ b ncxop~qecKn ~ cHnxponno C TOqKH 3peHn~ M o p ~ o 3 o r n n (CK~OHeHn~), C~OBOO6pa3oBann~ ~ CMb~C~a. O . H . Mopaxo~cKan, ' C n o c o 6 b l H a n M e n o ~ a H ~ ~ e p e 6 ~ z B a a q a ~ H ~ i e n e p n o a ~ ~n3Hn' (CTp. 121 132): npnBoanTCa nHTepecHb~fi nepeqenb pa3Hb~X T~HOB o p r a H n 3 a ~ n ~ ~aHHOFO ceMaHT~qeCKOFO npocTpaHCTBa.
76
REVIEWS
O.E. KapMaHoaa, 'Ha3aaHne aeqepHnx co6pannfi Mo~o~e~n ~ pyccicux roBopax' (cxp. 132-140): cTapb~e, Tpa~mtHOnnb~e na3aannn o6aa~aaroT 6o~btuofi yCTO~qnBOCT~,~O n npo)Io~a~ow qbyHI~2nOHnpOaaTb B ~aa~erTax, HalIOJIHflflCbHOBblM co)Iep~aHrleM.
Universitb de Toulouse-Le Mirail
MAURICE COMTET
REVIEWS
77
L'Hermitte, R., Science et perversion idbologique. Mart. Marrisme. Marristes. Une page de l'histoire de la linguistique soviktique. Institut du Monde Sovirtique et de i'Europe Centrale et Orientale. Cultures et socirtrs. De L'Est 8. Centre National de la Recherche Scientifique-Universit6 de Paris-Sorbonne. Paris: Institut des l~tudes Slaves, 1987, 102 pp. N. Ja. Marr died in December of 1934 at the height of his fame as the founder of Marxist linguistics in the Soviet Union. Marrism existed for fifteen more years. In 1950 Stalin declared the "new doctrine of language" a distortion of Marxism, Marr's pupils-understandably enough-agreed with his opinion, and until the dictator's death Soviet linguistics was busy rehashing Stalin's slim brochure, vituperating Marr, and relearning the basics of comparative philology. For some time after 1953 linguists in the U.S.S.R. followed the main European and American trends without any catastrophic developments, Marr's name lost its ominous ring, and gradually his former pupils forgot their fears and salvaged a good deal of his teachings. Me~raninov lived long enough to see his own works published and republished. The followers of Marr were not repressed under Stalin and with one exception remained in their high positions, though Filin, Kacnei'son, Budagov, Guxman, Desnickaja, as well as a number of less prominent scholars, had to change their orientation considerably. In the sixties, Marr was felt to be a victim of Stalin's purges and received a measure of sympathy due the other victims of the terror. When Marrism was in the ascendant, the West treated it with indifference or contempt. The notorious discussion of 1950 attracted some interest outside the Soviet Union, and Margaret Schlauch, for instance, admired its results very much indeed, but, on the whole, the West would have known next to nothing about Marr and his ideas, had Lawrence Thomas not written his excellent book The Linguistic Theories of N. Ja. Marr (1957). For some reason, Marrism has become rather popular as an object of investigation in recent years. In 1982 Jan Bjcrnflaten published a book about him: Marr og sprhkvitenskapen i Sovjetunionen, 1982 (cf. also chapter 4 of K.H. Phillips' Language Theories of the Early Soviet Period, 1986), and now Ren6 l'Hermitte has offered a monograph on the same period in Soviet linguistics. L'Hermitte's study has the following content: Introduction; I. Nikolaj Jakovlevi6 Marr (La naissance de la "japhrtologie," Apparition de la "nouvelle throrie du language," Les rractions en U.R.S.S., De Marr aux marristes, La guerre politico-idrologique est drclarre, Mort de N. Ja. Marr); II. L'aprrs-Marr et le triomphe du marrisme (Un certain modus vivendi, La situation en 1947, Guerre froide et lyssenkisme, Nouveile offensive des
Russian Linguistics 13 (1989).
78
REVIEWS
marristes, La "discussion" de 1950...et l'intervention de Stalin); III. Mort et r+surrection (relative) du marrisme (Vingt ans apr~s); IV. L'audience du marrisme en dehors de I'U.R.S.S. These four chapters are followed by a conclusion (a page and a half), an index of names, and a short list of relevant titles (the rest of the bibliography has to be pieced together from the footnotes). The monograph is written in an emotional style typical of L'Hermitte's other works, with numerous exclamatory and interrogative sentences. It presents an accurate report of Marrism, but it adds nothing to our knowledge of this phenomenon. L'Hermitte mentions neither Thomas nor Bj~rnflaten, which is unfair. If Marr is worthy of a new book today, this book should contain a brief section gathering up the most important facts and a detailed discussion of his activities from our own perspective. Certain questions are still of interest. Is there anything in Marr's works on the structure and history of language that has outlived their author? (His Caucasian studies are apparently still important, if one can judge by the publication of his book Basksko-kavkazskie leksi(eskie paralleli. Akademija Nauk GSSR, Tbilisi, Mecniereba, 1987. 192 pp). Could one publish a small volume of his selected writings for the benefit of the present-day scholars? Which of his ideas (if any), divested of Marxist rhetoric, were fruitful for the development of linguistics? Did he permanently influence Soviet linguistics through the works of his pupils? How do Meg~aninov's general and syntactic ideas fare today? What accounts for a moderate Renaissance of Marr in the late sixties and the seventies among Soviet scholars? How typical is the Marrian episode of twentieth-century linguistics as a whole? It would also be interesting to learn what made L'Hermitte himself write his book. In a word, a serious assessment of Marrism might be useful. We are not chroniclers of the past, we are its historians. We want to know where we stand in relation to our predecessors. If Marrism was nothing but a horrifying perversion, a poisonous fruit- among hundreds- ripened by the totalitarian sun, let it stay in its limbo and let the dead bury their dead.
University of Minnesota
ANATOLY LIBERMAN