The Prague School and Verbal Morphology, A Trend in European Structuralism

 

Contents Introduction 1. The 'Classical Period' of the Prague School 2. Later Developments Conclusion Footnotes Bibliography

 

Introduction

All metaphysical questions are historical questions, and all metaphysical propositions are historical propositions. Every metaphysical question either is simply the question what absolute presuppositions were made on a certain occasion, or is capable of being resolved into a number of such questions together with a further question or further questions arising out of these.

Collingwood, An Essay on Metaphysics, p. 49

 

Linguistics in the twentieth century has the absolute presupposition, stemming from Ferdinand de Saussure' s Cours de linguistique générale, that language should be studied as a system; this implies a step towards abstraction and is a complete reversal of what Pos 1939:71 calls the nineteenth century's nominalism which was concerned with the sheer description and accumulation of facts, isolated from one another (cf. Trnka 1948: 154). This tendency towards abstraction implies another presupposition, namely that it should be possible to provide a formal analysis of language and was best illustrated in the inter-war period by Trubetzkoy's Grundzüge der Phonologie. This last presupposition was carried out by J. Cantineau, in accordance with a major trend in American structuralism, to the point of virtually eliminating meaning from linguistic description. Another presupposition, which will be recurrent throughout this paper, is that the structure of 1anguage should be described in terms of binary features (e.g., langue/parole, markedness/unmarkedness).

This essay will deal with works on structural verbal morphology which use the formal apparatus originally devised by Trubetzkoy and his collaborators of the 'Cercle linguistique de Prague' for the analysis of phonology; in accordance with the very concept of structuralism, this means viewing language as a system of oppositions. This will entail a presentation of the so-called 'classical period' (1929-1939) of the Prague School, which culminated in the publication of Trubetzkoy's Grudzüge der Phonologie, and of the influence that these seminal ideas had on other members of the group when faced with the description of particular aspects of the verbal syntax. The second part will be devoted to an analysis of J. Cantineau's generalisation of Trubetzkoy's ideas about phonology to other fields of linguistics. There will also be an exposé of how the teachings of the Prague School have been synthesized by the Spanish scholar Martin Sánchez Ruipérez in a coherent doctrine which he applied to the description of Greek verbal morphology(l). This essay could also be viewed as a study of the concept of markedness from the early 1930's to the early 1950's, in a fraction of European structuralism.

The notion of a formal apparatus used in describing morphological phenomena will be considered a crucial one for the purpose of the present study. As for the goal of morphology, one can use a definition given in 1958 by representatives of the new Prague School :

The role of structural morphology is (1) to state morphological oppositions (e.g., that of the number of substantives, that of the common case — adnominal case, etc. in English) and their neutralisations (e.g., ... the neutralisation of grammatical genders in plural in German... ), (2) to state the phonemic means (often homonymous) implementing the morphological oppositions of a language, such as prefixes, suffixes, and alternations of phonemes (so-called morphonemics). (Vachek1960:50)

 

1. The 'Classical Period' of the Prague School

The first meeting of the 'Cercle linguistique de Prague' was held on 6 October 1926, under the presidency of Vilém Mathesius. Historically, the Prague movement can be divided into three periods : the first one, called the 'Classical Period', from 1929 to 1939, when occupation and war interrupted the activities of the Prague Circle; in fact the date 1939 coincides with the death of N. S.  Trubetzkoy. The second period is one of stagnation, characterised in the late forties by ideological squabbles around N. J. Marr's doctrine (Vachek 1966:13; Lepschy 1975:890). There was a revival of interest in the late fifties and early sixties which led to the publication of a new journal, Travaux linguistiques de Prague (1964) (on this ternary division, cf. Vachek 1961:67). In the present paper, we will be concerned only with the classical period, which acknowledged the combined influence of Baudouin de Courtenay and Ferdinand de Saussure (Mathesius 1932:6). Saussure's influence is also acknowledged by Trnka 1958:36 and Vachek 1966:18 who mentions the importance of S. I. Karcevskij, a pupil of Saussure, in the first days of the Prague Circle (cf. also Martinet 1953: 577) but the importance of this influence has been questionned by G. Lepschy 1970:53. One may also add the influence of the Czech Joseph Zubatý and the Russians Šachmatov and Ščerba (Trnka 1958:36).

From the very beginning, the Prague linguists were mainly concerned with phonology, though they also approached questions such as literary criticism, problems of the standard language, etc.; they intended to test their theses first in phonology and afterwards at higher levels of language (Vachek 1966:77-8).

Many scholars contributed during these years to the theory of phonology : for example, Martinet's contributions to the discussion of neutralisation and segmentation (Martinet 1936, 1939) are important historically but for the present purpose, it will suffice to set out the main theoretical assumptions made by the Prague linguists as they appear in Trubetzkoy's Grundzüge der Phonologie (published posthumously in 1939) and how they have been applied during this period to the analysis of morphology. This simplification in the presentation of the Prague linguists' doctrine is justified by the fact that Trubetzkoy was the most impressive personality in the group (Lepschy 1970:57).

Trubetzkoy calls phonological oppositions those phonetic oppositions which, in a given language, serve to differentiate meanings (Trubetzkoy 1976:33). The sounds which enter in such a relation, he calls phonemes(2). A phoneme has phonological content inasmuch as it is a member of a phonological opposition (Ibid.:69). These phonological oppositions can be classified according to the logical relation uniting their two members : such a classification is crucial in evaluating how the phonological system works (Ibid.:76). A privative opposition is one in which one member is characterised by the presence of a mark or feature, the other by the absence of the mark : for instance, in the pair b/p, b is characterised by the feature 'voice', p by its absence; the member of the opposition characterised by the presence of the mark will be called marked term, the other unmarked term. Trubetzkoy adds that this type of opposition is extremely important in phonology (Ibid.:77). A gradual opposition is one in which its members (here, by definition, can be more than two) are characterised by different degrees of the same feature : for example, o and in German have different degrees of opening. Trubetzkoy adds : "Les oppositions graduelles sont relativement rares et moins importantes que les privatives" (Ibid.). We shall have the opportunity to criticise this judgement later and see how it must be amended if the theoretical framework outlined by Trubetzkoy is to be applied to higher levels of linguistics. The third category is composed of equipollent oppositions, in which the two terms are logically equivalent, in other words these two terms cannot be viewed as having a different degree of the same feature nor as the affirmation or the negation of a feature. These last oppositions are held to be the most numerous in every system (Ibid.). And Trubetzkoy adds : "La valeur équipollente, graduelle ou privative d'une opposition phonologique dépend donc du point de vue auquel on se place pour la considérer. Mais on ne doit pas croire que cette valeur soit purement subjective et imaginaire : par la structure et le fonctionnement du système phonologique la valeur de chaque opposition est la plupart du temps donnée objectivement et sans équivoque" (Ibid.:78).

According to the extent of their distinctive power, oppositions can be either constant or neutralizable. In certain circumstances, one of the members of the opposition can lose its distinctive feature; in that case, there remains only what the two members have in common, in other words the "archiphoneme"; this concept appeared for the first time in Jakobson 1929, cf. Vachek 1966:22. The member of the opposition which appears in the neutralised position is to be considered as the unmarked term, this being only possible in cases of privative oppositions. On the other hand, if the neutralisable opposition is not privative but gradual, then it is always the extreme member of the opposition (the member which presents an extreme degree, either minimally or maximally, of a given feature) which appears in the neutralised position (Ibid.:85).

The publication earlier, in 1927, of Karcevskij's monograph Système du verbe russe constituted the first attempt, within the Prague Circle, to deal with morphology. The overall impression is that of an inventory of forms, the task of the linguist being to discover regularity amidst apparent chaos : "Cependant ne serait-il pas permis de considérer toute classification simplement comme un moyen pratique et conventionnel, dépourvu de toute valeur scientifique, de cataloguer les phénomènes linguistiques, comme un procédé destiné à mettre dans une langue un peu de cet ordre que la nature lui a refusé, afin de faciliter par là l'étude de cette langue ? Mais une langue est un système, et prétendre y introduire de l'ordre, c'est simplement avouer que nous ne la connaissons pas suffisamment. Penser autrement et / admettre qu'une langue puisse demeurer dans le chaos, c'est renoncer à la possibilité d'une science du langage" (Karcevskij 1927:43-4).

Though Karcevskij operates with notions such as neutralisation and the assumption that all grammatical categories should be binary, there is no systematic attempt at defining them. When dealing with the Russian infinitive, Karcevskij introduces the important notion of "zero-value" but, unfortunately, he does not develop this concept. But this treatment of morphology was very influential, as it was later acknowledged by Jakobson : "None of the experts in syntax would ever by-pass Karcevski's classification of elementary combinations (Saussure's syntagmes). Meillet was right in considering Karcevski's analysis of Russian verbal categories as a remarkable performance and, as V. Vinogradov points out, it is this work on the system of the Russian verb that gave such an impetus to investigations of verbal derivation" (Jakobson 1956:496).

But probably the earliest and most important systematic work on morphology during that 'classical period' and the most typical of the Prague School approach (Vachek 1966: 118, n. l5) remains Jakobson's "Zur Struktur des russischen Verbums" (1932), an article that looks rather sketchy, with many references to other Russian linguists but only a few examples. Jakobson starts by accepting Trubetzkoy's system of correlations (p. 76) and uses them in studying the values of the grammatical forms : the Russian verb contains two correlations of verbal aspect and two correlations of verbal voices. In the aspectual correlation, "perfective" is the marked term and "imperfective" the unmarked term; inside the imperfective group, there is another correlation, which appears only in the preterite : iterative (marked) / non-iterative forms (unmarked). An equivalent of the phonological concept of archiphoneme is also introduced : "Die merkmallose Form fungiert im sprachlichen Denken als Repräsentant des Korrelationspaares; darum werden als gewissermassen primäre Formen empfunden : die Imperfektiva gegenüber den Reflexiva [..]" (Jakobson 1932:83).

Jakobson characterises the asymmetry of the correlative grammatical forms as the antinomy of the signalising value of an A ("die Signalisierung von A") and its non-signalising value ("Nicht-Signalisierung"). Two signs may be used for the same referent, but in one case, the first sign will ascribe a mark to that referent , while in the second case this mark will remain unmentioned : for example, in Russian, the word for 'heifer' can be either télka or telénok. Although the two words apply to the same referent, only in the second case is the meaning left incomplete.

From the asymmetry of the correlative forms follows also another antinomy : the general and the partial meaning of the unmarked form, in other words the non-signalising of A and the signalising of non-A. Thus the same sign possesses two different meanings, so that in one case the mark of the referent remains unprecised, the presence of this mark being neither affirmed nor denied; in the other case, the absence of the mark is manifested. For example, the word telénok can either refer to a calf without any mention of its sex or can also refer to a male calf. Here Jakobson introduces a significant difference from Trubetzkoy's framework. Whereas at the phonological level the correlation is between the presence of a feature and its absence, here the correlation is more complex : the un-marked term does not mention the presence nor the absence of a given feature (in contradistinction to the phonological unmarked term which is always characterised by the absence of a feature; this analysis is different from the one given by Ruipérez, cf. infra). This analysis is also adopted by de Groot 1939:111.

 

If we take as an example of phonological neutralisation the German final devoicing of consonants ('Auslaut-verhärtung'), we get the following scheme,

to which can be compared

E. Seidel 1936 is worth noting in so far as it is a departure from Jakobson' s theory : "[...] ich kann Jakobson's [...] Gegenüberstellung von Perfektiv (=merkmalhaltig) zu Imperfektiv (=merkmallos) nicht völlig zustimmen. Mir scheint die Merkmallosigkeit eine der Funktionen zu sein neben den übrigen" (Seidel 1936:118). Seidel states that both the perfective and the imperfective aspects in Russian can function as marked and unmarked. In its ultimate consequences, this merely means abandonning.the attempt of applying to morphology and syntax the formalism devised for phonology in favour of a sheer classification. Martín Sánchez Ruipérez is severe in his critique of Seidel's attempt, saying that it is an absurd one (Sánchez R. 1954: 19).

B. Havránek also wrote an article on the same topic (Havránek 1939) which has been described as a minor contribution (Vachek 1966: 87). The structuralist formalism is much less developed than in Jakobson 1932; actually, the only reference to markedness is this quotation : "Donc l'aoriste [en vieux slave] se présente comme la forme du temps passé 'non-marquée' et l'imparfait comme la forme 'marquée', en employant la terminologie de la linguistique structurale" (Havránek 1939: 227). But Havránek's analyses are structuralist in so far as he integrates them in a global description of the verbal system.

Even if one is to include a couple of related articles (such as : Karcevskij 1932, Mathesius 1932, Trnka 1932), the production of the Prague linguists in the domain of morphology and syntax seems rather sparse. However one must emphasise the fact that they wanted to establish their theory on a solid phonological basis which, in the beginning, could not but entail a neglect of other linguistic levels in favour of a deeper insight in the analysis of the phonological component. Nevertheless, the basis on which future works could be built was laid in Jakobson's paper on the Russian verb; only after World War II could these ideas be developed, partly by modern representatives of the Prague School (Vachek 1966:86) but, for an obvious linguistic reason (the works being written in Czech), they will not be dealt with here and we will, in the stead, focus our attention on some Western European linguists who used the conceptual framework established by Trubetzkoy and Jakobson.

 

2. Later Developments

The publication in 1947 of the French translation of Trubetzkoy's Grundzüge was a major event since it contributed largely to the diffusion of the Prague theories among Western linguists. Following an indication given by Trubetzkoy himself ("tous ces principes de classement ne valent pas seulement pour les oppositions phonologiques, mais aussi pour n'importe quel système d'oppositions : ils ne contiennent rien de spécifiquement phonologique", Trubetzkoy 1976:94), Cantineau published some years later an article (Cantineau 1952) which is, possibly, the first attempt to generalise Trubretzkoy's ideas on phonology to the whole field of linguistics. However, it will be seen that the Prague School doctrine was distorted on a major issue by the influence of the American descriptivism. I shall now give a summary of Cantineau's views, introducing a critical analysis wherever necessary.

Cantineau calls 'opposition significative' the opposition made by two signs whose 'signifiers' (to use Culler's translation of 'signifiant ') are different. What differentiates this from a phonological opposition is that here whole signs, both signified and signifier, are part of the opposition whereas the members of a phonological opposition do not have meaning or, at least, meaning is not taken into account (which is a strange interpretation of Trubetzkoy 1976:33, already quoted). A classification of these 'oppositions significatives' is possible as it has been possible in phonology, according to the same logical principles.

Trubetzkoy's distinction between proportional oppositions and isolated oppositions is used as such and is held to be the basis of grammar since it allows for a clear distinction between grammar and lexicon : grammatical oppositions are proportional because the formal and semantic relation existing between the members of a given opposition also appears in at least another opposition in the same language; lexical oppositions are isolated. A gramrnatical opposition which becomes isolated ceases to be part of the grammar and enters the domain of the lexicon.

Amongst grammatical oppositions, a privative opposition will be one in which the signifier of one member will be characterised by the presence of a formal feature ("marque formelle") which is absent in the signifier of the other member. Cantineau illustrates this point by an example taken from Greek : in the verb λείπω, the unaugmented aorist forms, imperative λίπε, subjunctive λίπω, optative λίποιμι, are in a privative opposition with the corresponding present forms : λεῖπε, λείπω, λείποιμι which are marked by the root vocalism -ε- whereas this mark is absent from the aorist. As we shall see presently, Cantineau would disagree with the following critique since it will be made on semantic grounds. However, I believe it is not possible to give an account of morphology and syntax without any intervening semantic interpretation: for example, using only formal criteria cannot provide an analysis of the following Latin sentence : animal gramen pascitur because there is no nominative-accusative distinction in the neuter and since the word order is free (example from Martinet 1964:28). On the other hand, Cantineau does not exclude the methodology of the American linguists who interpret the French masculine adjectives rond, oblong, droit, gris, lourd, gros as lacking a mark present in the corresponding feminine forms ronde, oblongue droite, grise, lourde, grosse : though in principle the mark of a privative opposition should be the feature added to the unmarked member, it is not at all impossible to view it as a feature removed from the marked member. But such an analysis, which admits masculine as the marked member, though most economical from a morphophonemic point of view is counterintuitive : it does not take into account everyday uses of the masculine for both genders, e.g. L'homme est mortel vs. La femme est mortelle where the second example has not the generalising value of the first; the same holds for a pair like chien/ chienne, where the feminine will never be used to refer to the species. This raises the question of semantics in the grammar : should semantic -factors be taken into account in the definition of 'mark'? On this point, Cantineau's position is very clear : "la définition de la marque doit rester purement formelle, sur le plan du signifiant. On repoussera, comme contraire à la méthode, toute définition sémantique dans laquelle la marque serait cherchée sur le plan du signifié" (Cantineau 1952:29). This is, of course, in accordance with the general presupposition made by twentieth-century linguists that language should be described in formal terms and is a major characteristic of American descriptive linguistics, though it is wrongly assumed that Bloomfield rejected meaning in favour of a purely formal analysis (Bloomfield 1935:ch.9); this trend is most evidenced in the work of Z. H. Harris, e.g. "morphemes are not distinguished directly on the basis of their meanings or meaning differences, but by the results of distributional operations upon the data of linguistics"(Harris 1960 [=1951]: 363), though, even there, meaning is not completely excluded : "In determining the morphemes of a particular language, linguists use, in addition to distributional criteria, also (in varying degrees) criteria of meaning difference. In exact descriptive linguistic work, however, such considerations of meaning can only be used heuristically, as a source of hints, and the determining criteria will always have to be stated in distributional terms" (Harris 1960:365, footnote 6). Judging by the number of references to works of Harris, Cantineau was well aware of this position though 'Les oppositions significatives' does not yield any internal evidence of Cantineau's having read Methods in Structural Linguistics (published one year earlier). In any case, it is not possible to exclude meaning completely and one has to admit meaning as a criterion when faced with choosing between two competing analyses, as in the example already given of gender in French.

Moreover, if the mark can be seen either negatively or positively, this leads to the ultimate consequence (noted by Sánchez Ruipérez 1953:7) that it is no longer possible to identify the marked term only on the level of the signifier, i.e. by formal means, a consequence certainly not realised by Cantineau.

Cantineau also uses the phonological notion of equipollent opposition defined here as an opposition between two equivalent members, i. e. both characterised by the presence of a feature; he adds that these are the most frequent oppositions, an assertion which seems strange when one considers the examples given : Engl. foot/feet, goose/geese, German sie brechen/sie brachen. If one were to introduce semantic criteria, one would come to the conclusion that there is a correlation of privative opposition in the following pairs, each pair being proportional with one another: Engl. book/books, table/tables, tooth/teeth, German sie lieben/sie liebten, sie brechen/ sie brachen. The intuition of any native speaker will be that tooth and teeth stand in the same relation to one another as book and books. In other words, Cantineau's analysis does not reflect the native speakers' judgement about the working of their own language, though it is only fair to add that this kind. of structuralism is not intended to give a model of a native speaker's competence. In any case, one is led to question the necessity of using equipollent oppositions when dealing with higher linguistic levels; Cantineau's position may well stem from a misunderstanding of the following passage in the Grundzüge :"Les oppositions équipollentes sont dans chaque système les plus nombreuses" (Trubetzkoy 1976:77), where it is not clear whether Trubetzkoy refers only to phonological systems or to linguistic (or even semiologic) systems in general. In this connection, it is worth noting that Martín Sánchez Ruipérez does not use this concept in his analysis of the Greek verb.

Cantineau's concept of neutralisation is similar with the one Jakobson employs in "Zur Struktur des russischen Verbums" : an opposition is neutralised when its two signs have the same signifier though the signifieds are different, e. g. there is no number mark in the following pair in (spoken) French : il mange/ils mangent. So, implicitly, Cantineau tacitly accepts Jakobson's interpretation of neutralisation (cf. supra).

Cantineau concludes : "On voit [...] que la plupart des problèmes de grammaire peuvent être traités par la méthode des oppositions. Celle-ci est une méthode générale de classement formel. C'est par hasard qu'elle a été appliquée pour la première fois aux sons du langage. Elle permet d'étudier d'une façon plus approfondie une partie quelconque de la langue — de même que bien d'autres problèmes concrets" (Cantineau 1952:40).

The most thorough attempt to apply the Prague School phonological formalism to other fields of linguistics is probably Estructura del sistema de aspectos y tiempos del verbo griego antiguo, análisis funcional sincrónico (1954), by Martín Sánchez Ruipérez, professor of classics at the University of Salamanca (Spain) (3). The goal here is the same as for Cantineau : to apply the methods used in phonology, since they are of general value (Sánchez Ruipérez 1954:6 and 10). But such a classification will not take into account the signified : an opposition of signs is impossible without a corresponding opposition of signifieds whereas, on the contrary, an opposition of signifieds is possible even without a corresponding opposition of signifiers; in an article on the same subject written one year earlier, Sánchez Ruipérez says that "... the neutralization of morphological oppositions takes place also on the level of the signifié" (Sánchez Ruipérez 1953: 244). This implies furthermore that "the determining factor of the neutralization must be of semantic, not of phonic, value" (ibid.:24-5). This is a major difference from the view taken by Cantineau and, ultimately, this position is in accordance with the first major work of the Prague School on verb, Jakobson 1932.

Every feature distinguishing a sign at the level of the signified is called 'noción pertinente'; 'sentido' or sense applies to a value ('noción pertinente') realised in 'la parole'. This distinction does not appear in Jakobson 1932 nor Cantineau 1952.

Sánchez Ruipérez's treatment of isolated and proportional oppositions is the same as Cantineau's : the former pertains to the domain of the lexicon, the latter to the grammar. A grammatical opposition established, at the level of the signifier, by means of morphemes (which are defined as the signifiers of grammatical categories within a word) is called a morphological opposition (Sánchez Ruipérez 1954:11-12).

Privative morphological oppositions have a special character, different from privative phonological oppositions, a fact already noted by Jakobson 1932. This is due to their being an opposition of signifiers, and not only signifieds. A simple privative opposition will be represented by the formula

Ax / A

where A is a semanteme (or lexeme) and x a morpheme. The marked member Ax expresses the notion of the opposition shown by the morpheme (positive value). The unmarked member possesses a double function : as unmarked, A remains indifferent to the distinctive notion of Ax (neutral value); but as the term opposed to Ax, A can also express the negation or the absence of the notion x (negative value) (4). Sánchez Ruipérez (p. 17) adds that this is a significant difference from Jakobson 1932 : Jakobson says that the unmarked term does not mention the presence nor the absence of a given feature whereas Sánchez Ruipérez assigns to it two values, neutral and negative. Sánchez Ruipérez's position is justified empirically in his article on the neutral aspect of present indicative in classical Greek (1953).

Taking as an example the case of aspect in Greek, this conception of privative opposition parallels the analysis given by Martinet of the phonemes /t/, /d/, /n/ :

(Sánchez Ruipérez 1953:242; the difference from the scheme given supra is worth noting).

This means that the perfect is, as far as aspect is concerned, in opposition to the present-aorist block in that it expresses "la acción verbal después de su término" (p. 47), e.g. νενόσηκα "at a given moment, I fell ill and I am still ill". In contradistinction, both aorist and present express action in se (p. 45). By their neutral value as unmarked members in that opposition, present and aorist can be used instead of perfect: for example, in Herodotus IV 190 θάπτουσι δὲ τοὺς ἀποθνήσκοντας οἱ νομάδες the present participle (underlined) is used instead of the perfect, though the meaning is clearly that they bury the dead not the dying. There is a similar privative opposition between present (marked) and aorist (unmarked) within the unmarked member, non-perfect.

A gradual opposition is introduced to explain the values of the Greek moods but it is not developed further since the book is primarily concerned with tense and aspect (p. 15, especially footnote 1). In this respect, subjunctive and optative are differentiated only in the intensity with which they express the notion of mood. But this treatment should also include the imperative; one would then be led to posit objective expression of reality vs. subjective disposition of speaker, this last member being composed of a gradual opposition : imperative — subjunctive — optative, each term being a greater distanciation from the reality expressed by the lexeme (this is tentatively suggested in order to include the imperative. but would need to be developed and illustrated with examples). In phonology, it is theoretically possible to reinterpret a gradual opposition as a privative one, for example the opposition u/o as being close/non-close (depending on the phonological system described, cf. Trubetzkoy 1976:79) but the same is not true for morphological oppositions, since both members have a positive value whereas, if it were a privative opposition, one of the members would have a zero-value with its associated two uses already discussed.

Finally, contrary to Cantineau' s opinion, no morphological equipollent opposition is introduced by Sánchez Ruipérez : they are excluded on a semantic basis : "en una hipotética oposición morfológica equipolente, las nociones expresadas por cada uno de los términos, no siendo ni grados distintos: ni la afirmación o negación de una misma noción, serían valores independientes unos de otros, de tal modo que, dado uno, sería imposible conocer cuál es el otro o cuáles son los otros, contradiciendo así el carácter que debe tener toda oposición gramatical significativa" (p. 16).

This has the result that, of the three types of phonological oppositions defined by Trubetzkoy, only two seem to apply at higher linguistic levels.

As opposed to Cantineau, Sánchez Ruipérez, by his use of semantic criteria in analysing morphological oppositions, is in direct filiation from the classical theses of the Prague School, e.g. Pos 1939:75 : "L'opposition en elle-même et dégagée de tout facteur matériel, est de nature éminemment logique : c'est une relation qui ne se constate pas, mais qui se pense. Les opposés sont deux, mais d'une façon particulière; leur dualité n'a pas le caractère indéterminé et contingent de deux objets arbitrairement réunis par la pensée. La particularité consiste en ce qu'étant donné l'un, la pensée déduit l'autre, ce qui n'est pas le cas de la dualité contingente."

 

Conclusion

In the early days of the Prague School it was held that the system of a language should be uncovered by the linguist since it is pre-existing, in other words the linguist, with the help of his classificatory tools, does not put things into order but only reveals an order already present.

For phonological description, Trubetzkoy devised a kind of formalism which afterwards came to be applied to morphology, but never to the same extent; this use of formalism is part of a trend in modern linguistics which ultimately culminates in the use of mathematical devices. The system of (privative) oppositions in morphology is characterised by its binarism, a feature best exemplified in Jakobson's works : given a system of the following type

____A____

B (b - c)

A is in opposition to B (which stands for b and c as a whole) and within B, b is in opposition to c. Of the three kinds of oppositions postulated by Trubetzkoy in Grunzüge, only two, privative and gradual oppositions, seem workable at higher linguistic levels, or at least in morphology. The third, equipollent opposition, has been rejected on semantic grounds, in accordance with the practice of the first Prague linguists who always gave equal importance to form and meaning.

 

Footnotes

1. It should be added here that the difference between morphology and syntax is somewhat attenuated in the Prague movement : "According to the Prague functionalists morphology is concerned with the analysis of the word, whereas syntax is mainly the analysis of the sentence into its constituent relationships (e.g. subject — predicate, etc.). In contradistinction to the views of the former linguists, the Prague School holds that morphology and syntax cannot be linguistically contrasted to each other as two disciplines concerned with 'parole' and 'langue' respectively, because even syntax deals not only with 'parole' but also with 'langue', in attempting to discover normothetic laws, whose individual actualizations take place in utterances" (Vachek 1960:50, s.v. morphologie et syntaxe).

2. The notion of phoneme itself has evolved since the early days of the Prague School when the phonemes were defined by Trubetzkoy and Jakobson as "des images acoustico-motrices les plus simples et significatives dans une langue donnée" (First Congress of Slavicists, 1929; also TCLP 1, 1929, 10-11), terms which are reminiscent of Baudouin de Courtenay's "Lautvorstellung". This psychologistic approach was abandoned in the "Projet de terminologie phonologique standardisée", TCLP 4., 1931, 309-323, for a definition as a phonological unit not dissociable into smaller and simpler (and Trubetzleoy will add later : successive) phonological units (this does not take into account the notion of distinctive feature.) For the history of the concept of phoneme, see Trubetzkoy 1976: 36 ff., esp. 41-46 and Vachek 1966, 43-50.

3. Most of his data come from E. Schwyzer's Griechische Grammatik (II/ Syntax und syntaktische Stilistik).

4. "En el término no caracterizado de una oposición morfológica privativa, junto al valor neutro o de indiferencia a la noción distintiva, el signo posee el valor negativo  consistente en la indicación de la ausencia o negación de la noción básica" (Sánchez Ruipérez 1954: 18).

 

Bibliography

Bloomfield, L. 1935. Language, London, George Allen and Unwin Ltd.

Cantineau, J. 1952. 'Les oppositions significatives', Cahiers Ferdinand de Saussure 10, 11-40.

Charisteria Guilelmo Mathesio quinquagenario ... oblata. 1932. Prague.

Collingwood, R.G.. 1972. An Essay on Metaphysics, Chicago, Henry Regnery Company.

Culler, J. 1976. Saussure. Glasgow, Fontana Modern Masters.

Daneš, F. and J. Vachek. Prague Studies in Structural Grammar today, TLP 1, 21-31.

Garvin, P.L. 1963. 'Czechoslovakia', Current Trends in Linguistics, 1, 499-522.

--------, 1964. A Prague School Reader on Esthetics, Literary Structure and Style, Georgetown University Press.

--------, 1969. 'The Prague School of Linguistics' in Archibald A. Hill, ed., Linguistics Today, New York and London, Basic Books Inc., 229-238.

Groot, A.W. de. 1939. 'Les oppositions dans les systèmes de la syntaxe et des cas', Mélanges Bally, 107-127.

Harris, Z.S. 1960. Structura1 Linguistics, Chicago and London, University of Chicago Press.

Havránek, B. 1939. 'Aspect et temps en vieux slave', Mélanges Bailly, 223-230.

Holk, A.G.F. van. 1964. 'Functional Syntax and Syntactic Operations', MSLL 17, 1964, 37-46.

Ivić, M. 1965. Trends in Linguistics, The Hague, Mouton.

Jakobson, R. 1929 'Remarque sur l'évolution phonologique du russe, comparée à celle des autres langues slaves', TCLP 2.

--------, 1932. 'Zur Struktur des russischen Verbums' in Charisteria G. Mathesius, Prague, 74-84 = J. Vachek, ed., 1964-, 347-359.

--------, 1936. 'Beitrag zur allemeinen Kasuslehre', TCLP 6, 240-288.

--------,1939. 'Nikolaj Sergejevič Trubetzkoy' in Sebeok 1966:526-542. Reprinted from Acta linguistica 1.64--76.

--------, 'Serge Karcevski', in Sebeok 1966:493-496. Reprinted from Cahiers Ferdinand de Saussure 14. 9-13.

Karcevskij, S. 1927. Système du verbe russe, essai de linguistique synchronique, Prague.

--------, 1932. 'Sur la structure du substantif russe', Charisteria G. Mathesisus, Prague, 65-73 = Vachek 1964:335-346.

Lepschy, G.C. 1970. A Survey of Structral Linguistics, London, Faber and Faber.

--------, 1975. 'European Structuralism: Post-Saussurean Schools', Current Trends in Linguistics 13, 887-902.

Leroy, M. 1963. Les grands courants de la linguistique moderne, Brussels and Paris.

Malmberg, B. 1964. New Trends in Linguistics, Stockholm and Lund.

Martinet, A. 1936. 'Neutralisation et archiphonème' TCLP 6, 46-57.

-------- 1939. 'Un ou deux phonèmes?', Acta linguistica 1, 94-103.

--------,1953. 'Structural Linguistics' in A.L. Kroeber, ed., Anthropology Today, Chicago, 574-586.

--------, 1959. 'Quelques traits généraux de la syntaxe', Free University Quaterly. 2, 115-129.

--------,1960. 'Elements of a Functional Syntax', Word 16, 1-10.

--------, 1964. 'The Foundations of a :Functional Syntax' , MSLL 17, 25-36.

Mathesius, V. 1932. 'La place de la linguistique fonctionnelle et structurale dans le développement général des études linguistiques', Časopis pro moderní filologii, 18, 1-7.

--------, 1936. 'On some Problems of the Systematic Analysis of Grammar', TCLP 6, 95-107 = Vachek 1964 :306-319.

--------, 1936b.'Deset let Pražskeho linguistického kroužku', Slovo a slovesnost 2, 137-145; translated in Vachek 1966:137-151, under the title 'Ten Years of the Prague Linguistic Circle'.

Pos, H.J. 1939. 'Perspectives du structuralisme', TCLP 8, 71-78.

Projet de terminologie phonologique standardisée, TCLP 4, 1930, 309-323.

Robins, R.H. 1967. A Short History of Linguistics, London, Longman.

Sánchez Ruipérez, M. 1953. 'The Neutralization of Morphological Oppositions as Illustrated by the Neutral Aspect of the Present Indicative in Classical Greek', Word 9, 241-252.

--------, 1954. Estructura del sistema de aspectos y tiempos del verbo griego antiguo, análisis funcional sincrónico, Salamanca, Theses et studia philologica salmanticensia.

Sebeok, T.A. 1966. Portraits of Linguists, A Biographical Source Book for the History of Western Linguistics, 1746-1933, Indiana University Press, vol. 2.

Seidel, E. 1936. 'Zu den Funktionen des Verbalaspekts', TCLP 6, 111-129.

Thèses, TCLP 1, 1929.

Trnka, B. 1930. 'On the Syntax of the English Verb from Caxton to Dryden', TCLP 3.

--------, 1932. 'Some Thoughts on Structural Morphology', Charisteria Mathesius, 57-51 = Vachek 1964 : 329-334.

--------, 1946. 'Vilém Mathesisus' in Sebeok 1966 : 477-488 : Časopis pro moderní filologii 29, 3-13.

--------, 1948. 'Jazykozpyt a myšlenková struktura doby', Slovo a slovesnost 10, 73-83. Translated in Vachek 1966 : 152-165.

-------- et al. 1958. 'Prague Structural Linguistics', Philologica Parguensia 1, 33-40 = Voprosy jazykoznan'ija 3, 1957, 44-52.

Trubetzkoy, N.S. 1929. 'Sur la morphonologie"', TCLP 1,85-88 = Vachek 1964:183-186.

--------,1936. 'Die Aufhebung der phonologischen Gegensätze', TCLP 6, 29-45.

--------,1976. Principes de phonologie, Paris, Klincksieck. Translation of Grunzüge der Phonologie, TCLP 7, 1939.

Vachek, J. 1936, 'Phonemes and Phonological Units' TCLP 6, 235-239.

--------,1961. 'À propos de la terminologie linguistique et du système de concepts linguistiques de l'École de Prague', Philologica Praguensia, 4, 65-78.

--------, 1964. A Prague School Reader in Linguistics, Bloomington, Indiana University Press.

--------, 1966. The Linguistic School of Prague, Bloomington and London, Indiana University Press.

-------- and J. Dubský. 1960. Dictionnaire de linguistique de l'école de Prague, Utrecht and Anvers,

Spectrum.

Waterman, J.T. 1970. Perspectives in Linguistics, Chicago, 2^nd ed., The University of Chicago Press.

Addenda

Bierwisch, h. 1971. Modern Linguistics. Its Development, methods and problems, Janua Linguarum 110, The Hague, Mouton.

Faye, J.P. and L. Robel, eds., 1969. 'Le Cercle de Prague', Change 3, Paris, Seuil.

Garroni, Emilio (ed.). 1966. Il circolo linguistico di Praga. Le tesi del '29, Milan, Silva.

Møller, C. 1936. 'Thesen und Theorien der Prager Schule', Acta Jutlandica 8:2.

Vachek, J. 1970. 'On some less known aspects of the early Prague linguistic school'. Actes du X^e Congrès International des Linguistes, vol. 2, 333-7.