Roman Jakobson and the Birth of Structuralism

W. Keith Percival

[This is a preview of an article of mine entitled "Roman Jakobson and the Birth of Linguistic Structuralism" to appear shortly in the journal Sign Systems Studies, Tartu, Estonia. Note that the published version will differ significantly from the one presented here. While comments and inquiries from interested readers are obviously welcome, I should prefer that this web version not be cited verbatim.]

A difficult problem facing historians of twentieth-century linguistics is gauging the impact on the field made by Ferdinand de Saussure's posthumous book Cours de linguistique générale (1916). The by now traditional notion is that this was such a revolutionary book that all forward-looking linguists immediately began drawing inspiration from it and from it alone.

While this perspective certainly contains a grain of truth it overlooks a number of factors. First, Saussure was only one of several linguistic theorists active in the early twentieth century. The crucial question, therefore, is why so many linguists chose to regard themselves as followers of Saussure rather than some other linguistic theorist (or theorists) and how this exclusive loyalty to Saussure came about. Second, Saussure was arguably part of a broad trend in the history of ideas beginning at the end of the nineteenth century and the beginning of the twentieth that rejected the "positivism" prevalent among academics in the humanities and the social sciences.[1] It is not surprising that when these ideas began to affect linguistics the influence did not come from Saussure alone. For many linguists Saussure's Cours merely confirmed and reinforced what they already believed.

That this was the case is shown most clearly in the career of Roman Jakobson (1896-1982), who had already begun to absorb non-positivist ideas in the second decade of the twentieth century, long before copies of Saussure's Cours de linguistique générale reached Russia. One important philosophical source of these ideas was the phenomenology of Edmund Husserl (1859-1938), which reached Jakobson via the Russian philosopher Gustav Špet (born in 1879, executed by the KGB in 1937), who had studied with Husserl in Germany. Moreover, non-positivist ideas were also being espoused in the same period (viz. 1910-1920) by Jakobson's many artistic and literary friends, and they likewise could not have got these ideas from reading Saussure's Cours.

Given facts of this kind, we should perhaps invert the relationship between Saussure's book and twentieth-century linguistic theorizing and try to discover what specific themes in it came to be singled out and promoted by linguists when the theories propounded in that book began to be assimilated.

In the case of Roman Jakobson, we must reckon with an additional complicating factor, namely that he was exposed to Saussure's ideas in two stages. First, he had a colleague and friend, Serge Karcevski (Karcevskij), who had studied in Geneva, where he absorbed Saussurean doctrine from Charles Bally, Saussure's friend and immediate successor. Having returned to Moscow in 1917, Karcevski passed some of these ideas on to the linguists there. Thus, on the first page of a paper on the poetry of Khlebnikov, first presented in Moscow in May 1919, Jakobson voiced the Saussurean distinction between synchronie and diachronie, but without actually using Saussure's own terminology. Then, in the same paragraph, he cites Saussure's idiosyncratic term poussière linguistique, an expression which curiously did not occur anywhere in the Cours and was not mentioned in the famous courses on general linguistics that Saussure had given at the University of Geneva from 1907 to 1911. Jakobson's 1919 lecture appeared in print in Prague two years later, after he had left Russia for good.[2]

So here we have a group of linguists influenced by some of Saussure's ideas without having ever seen the Cours de linguistique générale. Sometime in the early 1920s, Jakobson, now in Prague, sent his colleagues back in Russia a copy of the second (1922) edition of the Cours which Albert Secheahaye had sent him from Geneva.

As for the term "structuralism," it did not begin to be used by linguists until the late 1920s, over a decade after the Cours de linguistique générale first appeared. In fact, it first occurs in the writings of the Linguistic Circle of Prague (Cercle linguistique de Prague), which had been founded by Vilém Mathesius and Roman Jakobson in 1926. An important early document was a manifesto presented collectively by a group of the members of the Circle at a congress on Slavic philology held in Prague in October 1929.

In the same year (viz. 1929), Roman Jakobson gave the following much-quoted thumbnail definition of structuralism: "Were we to comprise the leading idea of present-day science in its most various manifestations, we could hardly find a more appropriate designation than structuralism. Any set of phenomena examined by contemporary science is treated not as a mechanical agglomeration but as a structural whole, and the basic task is to reveal the inner, whether static or developmental, laws of this system. What appears to be the focus of scientific preoccupations is no longer the outer stimulus, but the internal premises of the development; now the mechanical conception of processes yields to the question of their functions" (Roman Jakobson, from an article headed "Romantické všeslovanství — nová slávistika" in the Czech journal Čin dating from 31 October 1929 and translated into English and re-published in Roman Jakobson, Selected Writings, vol. 2: Word and Language [The Hague & Paris: Mouton, 1971], p. 711). This may be the earliest use of the word "structuralism" by a linguist.

In the 1920s, furthermore, we already see members of the Linguistic Circle of Prague relating their ideas to Saussurean theoretical positions whenever they could. An early example occurs in an article by Serge Karcevski himself entitled "Études sur le système verbal du russe contemporain" published in Prague in 1922 in the first volume of the journal Slavia (pp. 242-268). Karcevski, who (as mentioned above) had studied under Charles Bally at the University of Geneva from 1916 to 1917, hence shortly after Saussure's death, begins his article with the following quotation from the Cours (2nd edition, p. 124):

"La langue est un système dont toutes les parties peuvent et doivent être considérées dans leur solidarité synchronique." 'Language is a system all parts of which can and should be considered in their synchronic inter-dependence.' Note that Karcevski uses Saussure's idiosyncratic term "synchronique" here, instead of the more traditional "statique." In this connection, see Cours de linguistique générale, 2nd ed., pp. 176-177 and Tullio de Mauro's commentary in his endnotes 256 and 257 of his critical edition of the Cours (Paris: Payot, 1984), p. 470.

The term "structuralism" was, however, not created de novo by linguists in the late 1920s but had already been in use among psychologists a generation earlier. At that time, it referred to the theories professed by the British-born American psychologist Edward B. Titchener (1867-1927), who had studied at the University of Leipzig with the founder of experimental psychology Wilhelm Wundt (1832-1920). The essence of Titchener's approach consisted in using evidence from introspection to determine the irreducible elements of consciousness and establish the ways in which these elements combine.

In an earlier period still, moreover, the terms "structure" and "structural" had been current in various scientific and technical fields all the way back to the first half of the nineteenth century. It is instructive to consult the relevant entries in The Oxford English Dictionary. An example of the early use of the word structure in connection with language occurs in the statement of aims of the Philological Society of London, newly founded in 1842: "[...] the investigation of the Structure, the Affinities, and the History of Languages."[3]

As we have seen, these terms were introduced into twentieth-century linguistics by Roman Jakobson. We do not know exactly why he decided to use them to refer to his approach to linguistics and that of his Prague colleagues in this fashion. It is clear that he had had broad intellectual interests from the beginning of his career, and that he was well read in all the various Slavic literatures and as well as in the entire philosophical, artistic, literary, and linguistic literature current in the early decades of the twentieth century. It is conceivable that he had heard about psychological structuralism from colleagues in psychology (Christian von Ehrenfels, for instance), but it seems more likely that he decided to adopt this terminology for reasons of his own.[4].) However, while that starting point is tantalizingly shrouded in mystery and may for ever remain, there is no doubt that Jakobson disseminated the notion of structuralism to his fellow linguists in western Europe in the 1930s. Later, his association with Claude Lévi-Strauss in New York in the early 1940s was of crucial importance in the spread of some of the key ideas of structuralism to anthropology and from there to other social sciences (one thinks in this connection of Jean Piaget in psychology). It was also from Jakobson that the American linguist Zellig Harris got the term "structural" that appears in the title of his famous Methods in Structural Linguistics, published in 1951. Perhaps it is high time for linguists to begin commemorating Roman Jakobson, rather than Ferdinand de Saussure, as the founder of structural linguistics.

The use of terms like "structure," "structural," etc. was, I might also suggest, a manifestation of the strong desire that linguists have felt since the mid-nineteenth century to adopt words redolent of the biological and natural sciences. Think, for instance, of terms like "analysis, "organism" and "morphology," borrowed by comparative philologists in the nineteenth century from biology and geology, two fields with unimpeachably scientific legitimacy at that time. Words of this kind are an important component in the scientific window-dressing that professional linguists have constructed for themselves over the past two centuries. As such, it is not necessary that they have unequivocal meanings. In this instance, the vagueness of the term "structure" may even have contributed to its original cachet. Indeed, the very fact that it was non-committal may have made it an ideal slogan for the promoters of various new brands of linguistic research that otherwise had little in common.

Significantly, moreover, the advent of structuralism coincided chronologically with the appearance of the Saussure myth, i.e. the notion that Saussure was the founder of modern linguistics. Perhaps, therefore, the two were different sides of the same coin. If that is so, it may be that in a strange sense Saussure really was the founder of structuralism: the self-styled structuralists themselves saw to that. Although it can be convincingly shown, as I believe I did in my 1981 article in Semiotica, that no immediate Saussurean paradigm was unleashed by the publication of the Cours de linguistique générale in 1916, [5] it may nevertheless be true that the vogue of the term "structuralism" was connected with the need felt by the devotees of that book to believe fervently in various ideas that they read into it and also by their equally fervent desire to imagine that they were members of a single coherent scientific movement opposed to the linguistics of their immediate predecessors. While we certainly know something about how these beliefs first took shape, the question of how they were disseminated and came to be accepted by later generations of professional linguists is still far from clear. We need to track the use made by linguists of the cluster of terms surrounding the concept "structure." Imagine comprehensive entries for these terms in the style of the Oxford English Dictionary or Le Grand Robert!

Needless to say, the phenomenon of "structure talk," i.e. the widespread use of the term "structure" in many different disciplines, raises a number of interesting questions. Here are just a few of them: How has linguistic structuralism nowadays come to be associated with the trend away from historical to descriptive linguistics? In any event, what factors caused this "synchronic turn" in linguistics? Were there perhaps three (not two) fundamental developments in twentieth-century linguistics, namely the rise of structuralism, the turning away from historical studies, and the vogue of Saussure's Cours de linguistique générale. How did structuralism subsequently spread to the social sciences, literary studies, philosophy, and even physics? To what extent was the ground prepared in the social sciences by the earlier use of the term "structuralism" in psychology and linguistics? What led to the expansion of structuralism to include talk of "deep structure," which was not originally part of it? What is the relation between European structuralism and the structuralism of the New World? To what extent can generative grammar be considered part of the structuralist movement? What role have various brands of structuralism played in the rise of critical theory, poststructuralism, and postmodernism in literary studies and philosophy? Can parallels be drawn between "structure talk" and more recent talk involving the slippery term "postmodern"? In general, what can we discern behind these various terminological pandemics? Was there more to structuralism in linguistics than a bizarre kind of discourse?

Last but not least, let me mention an important ingredient that must be added to any well-balanced account of the origins of Prague School structuralism, namely the influence of the Russian Formalist school of literary analysis, mediated first and foremost by Roman Jakobson, who moved from Moscow to Prague in 1920. In addition, of course, the members of the Linguistic Circle of Prague went on to develop their own theories of literary analysis, but these differed from those of the Russian formalists; hence the two movements, though historically linked, cannot be equated.[6]

Facts of this kind lead me to raise the awkward question of the relation between linguistics and literary criticism in the period up to the emergence of linguistic structuralism in Prague in the late 1920s. In this area, it seems abundantly clear that linguistic theory was tributary to early twentieth-century modernist movements in literature, not to mention the fine arts. I realize that historians of linguistics may still be loath to investigate links between their own field and the study of literature and art, which they, perhaps understandably, regard as fundamentally unscientific and hence of no possible relevance to the history of their own discipline. But sooner or later the problem will have to be faced.

There were links at different times between structuralism and various contemporary philosophical trends. As we have seen, early structuralism in Czechoslovakia owed much to the phenomenology of Edmund Husserl. On the other hand, later structuralism, especially in France, came to play an important role in the acrimonious debates between the phenomenologists and the existentialists, the followers of Martin Heidegger on the one hand and those of Jean-Paul Sartre on the other.[7] Even stranger perhaps is the fact that in the United States structuralism was associated with extreme empiricism (the philosophical position often called "logical positivism"). A similar orientation was adopted by the Danish linguist Louis Hjelmslev. He was also an outspoken empiricist, like his American counterpart Leonard Bloomfield. That structuralism could assume such a variety of philosophical positions is a clear indication that it was not a rigidly defined philosophical persuasion. The term was flexible enough to be

As regards future research in this area, I might tentatively suggest the following as a reasonable agenda. First, the relations, personal and institutional, between the various structuralist schools will need to be investigated. Second, the initial local theoretical tradition of each school of linguists will need to be established. Third, the impact of Saussure's Cours on each group will need to be tracked. Fourth, the impact of other influential contributions to theoretical linguistics will need to be re-examined pari passu. Fifth, the relations between linguists and scholars/scientists in other fields will need to be focused on. Sixth, the wider ramifications of linguistic theorizing will need to be brought into the picture.[8] In this entire area, if I may return to the question raised at the beginning of this essay, the vogue of the Cours de linguistique générale may well turn out to function like Ariadne's thread.



  1. The most important voice was the philosopher Edmund Husserl, whose works were known to and appreciated by the early structuralists, including Jakobson; see, for instance, Elmar Holenstein, Roman Jakobson's Approach to Language: Phenomenological Structuralism (Bloomington & London: Indiana University Press, 1976). On the impact of the anti-positivist movement in the world of literary criticism, see René Wellek, "The Revolt against Positivism in Recent European Literary Scholarship," in William S. Knickerbocker, ed., Twentieth Century English (New York: The Philosophical Library, 1946), pp. 67-89. I often feel that "positivism" is rather an inadequate word to describe the approach to linguistics of the late nineteenth-century practitioners, such as the so-called Junggrammatiker (Neogrammarians), to whom Jakobson and his colleagues were staunchly opposed, but I adopt it here for want of a better alternative. On the hostility to positivism and preference for holistic thinking in Russia, see above all Patrick Sériot, Structure et totalité: Les origines intellectuelles du structuralisme en Europe centrale et orientale (Paris: Presses Universitaires de France, 1999), to which I have been greatly indebted. For an important primary source, the reader may wish to consult Roman Jakobson's seminal article "Über die heutigen Voraussetzungen der russischen Slavistik," Slavische Rundschau, 1:8 (1929), 629-646, esp. pp. 631-634. Go back to the text.

  2. R. Jakobson, Noveishaia russkaia poeziia: Nabrosok pervyi (Prague: Tipografia "Politika" v Prage, 1921). For a complete re-edition of this early book of Jakobson's conveniently provided with a facing German translation, see Jurij Striedter & Witold Košny, eds., Texte der russischen Formalisten, vol. 2: Texte zur Theorie des Verses und der poetischen Spache (Munich: Wilhelm Fink Verlag, 1972), pp. 18-135, at p. 18. For an English translation, see Edward J. Brown, Major Soviet Writers: Essays in Criticism (Oxford University Press, 1973), pp. 58-82. Saussure used the expression poussière linguistique to refer to features of the language of a particular period that were relics of earlier stages of the same language and were no longer productive. This notion plays an important role in Jakobson's argument about the relation between the poetry of Pushkin and that of his own contemporaries Vladimir Mayakovsky (1893-1930) and Velimir Khlebnikov (1885-1922). Go back to the text.

  3. See Fiona Carolyn Marshall, "Edwin Guest: Philologist, Historian, and Founder of the Philological Society of London," The Henry Sweet Society for the History of Linguistic Ideas, Issue No. 42 (May 2004), p. 17. Go back to the text.

  4. Jakobson himself mentions Christian von Ehrenfels in his Selected Writings, vol. 2: Word and Language (The Hague & Paris: Mouton, 1971), p. 716. Elmar Holenstein comments on the term "structural" in his book Roman Jakobson's Approach to Language: Phenomenological Structuralism (Bloomington & London: Indiana University Press, 1976), p. 14 and mentions its prior use by Edward Titchener and (earlier still) by Wilhelm Dilthey. René Wellek says: "The word structure, which in Czech is a technical, foreign term with none of the associations of building that it has in English, occurs in Mukařovský's Máj in 1928." A History of Modern Criticism: 1750-1950, vol. 7: German, Russian, and Eastern European Criticism, 1900-1950 (New Haven & London: Yale University Press, 1991), p. 442. I might point out that the English word structure had also begun to lose many of its associations with building by the eighteenth century (see the entry on this word in The Oxford English Dictionary). This broadening of the word's meaning was doubtless prepared by the metaphorical use of Latin structura in literary contexts (see the relevant entry in any unabridged Latin dictionary such as Forcellini). See also Wellek's remarks on the crucial role played by the term "structure" in Russian formalism, op. cit., p. 321. On the relation between Prague School literary criticism and earlier Russian formalism, see Peter Steiner, From Formalism to Structuralism: A Comparative Study of Russian Formalism and Prague Structuralism (Doctoral dissertation, Russian and East European Literature, Yale University, New Haven, Connecticut, 1976). Also especially helpful is Jindřich Toman, The Magic of a Common Language: Jakobson, Mathesius, Trubetzkoy, and the Prague Linguistic Circle (Cambridge, Massachusetts & London: MIT Press, 1995), and above all Patrick Sériot, Structure et totalité (Paris: PUF, 1999). Go back to the text.

  5. See "The Saussurean Paradigm: Fact or Fantasy?" Semiotica, 36 (1981), 33-49. Go back to the text.

  6. On the formalists, see Victor Erlich, Russian Formalism: History, Doctrine, Slavistische Drukken en Herdrukken, 4 (The Hague: Mouton, 1955, and later editions). On Jakobson's involvement with the so-called Futurist poets, see Roman Jakobson's posthumous book My Futurist Years, compiled and edited by Bengt Jangfeldt and Stephen Rudy, translated with an introduction by Stephen Rudy (New York: Marsilio Publishers, 1997). For a seminal treatment of the Futurist movement in Russia see Vladimir Markov, Russian Futurism: A History (Berkeley & Los Angeles: University of California Press, 1968). To some historians of linguistics, the Futurists may seem even further afield than the Formalists, but Jakobson was himself a poet, and some Futurist notions did indeed carry over into his linguistics. On that factor, see Jakobson's own reminiscences in My Futurist Years, esp. pp. xii-xvi. Go back to the text.

  7. On later developments, see J. G. Merquior, From Prague to Paris: A Critique of Structuralist and Post-Structuralist Thought (London: Verso, 1986). Go back to the text.

  8. I do not wish to imply here, of course, that no progress whatever has so far been made. An excellent survey of the various European structuralist schools of linguistics is, for instance, Jörn Albrecht, Europäischer Strukturalismus: Ein forschungsgeschichtlicher Überblick (Tübingen: Francke Verlag, 1988). See also John E. Joseph, From Whitney to Chomsky: Essays in the History of American Linguistics (Amsterdam & Philadephia: John Benjamins, 2002), esp. pp. 54ff. On structuralism outside linguistics a vast secondary literature has grown up, needless to say, but most of it is devoted to ideological promotion. Moreover, little of it can be understood by professional outsiders, unfortunately. Go back to the text.