Book announced at https://linguistlist.org/issues/34.2136

EDITOR: James McElvenny
TITLE: The Limits of Structuralism
SUBTITLE: Forgotten Texts in the History of Modern Linguistics
PUBLISHER: Oxford University Press
YEAR: 2023

REVIEWER: Geoffrey Sampson

SUMMARY

This book aims to shed light on the difficult concept of “structuralism” by reprinting publications by seven scholars active in the early decades of synchronic linguistics (all but the latest of these reached print at least ninety years ago), with each introduced by a present-day academic. “Reprinted” is not quite the right word, because most of these contributions were originally in other languages and are published in English translation here for the first time. Helpfully, original page numbers are inserted in the translations to facilitate comparison with the originals.

After Chapter 1, in which the editor James McElvenny introduces the concept of structuralism and the aims of his book, the seven “reprinted” authors are:

Ch. 2: Peter Du Ponceau on the grammatical structure of North American languages, introduced by Floris Solleveld

Ch. 3: Franz Boas, “The classification of American languages”, introduced by Margaret Thomas

Ch. 4: Georg von der Gabelentz on the idea of language typology, introduced by McElvenny

Ch. 5: Antoine Meillet on grammaticalization, introduced by John Joseph

Ch. 6: Roman Jakobson on “The Eurasian language union”, introduced by Patrick Sériot

Ch. 7: Louis Hjelmslev on what he called “linguistic correlations”, introduced by Lorenzo Cigana

Ch. 8: Emile Benveniste on “Structure of language and structure of society”, introduced by Chloé Laplantine.

In the first chapter, McElvenny spells out what I understand by “structuralism” in linguistics in terms of two famous quotations, which describe a language as “un système où tout se tient” and say “dans la langue il n’y a que des différences _sans termes positifs_”. That is, a language should not be seen as a heap of disconnected elements but a structure of relationships where a change at any one point in the system will carry in its train changes elsewhere in the system, and what matters about an element is not its individual properties but how it relates to other elements. (Compare a currency, where what matters about a coin of a given pattern is that it is equivalent to two of another pattern or five of a third pattern, but the fact that British “silver” coins used to be made of a copper/nickel alloy and are now nickel-plated steel has no relevance to the functioning of the currency.)

Before reading this book, I confidently believed that the two French quotations I have just given both came from Saussure’s _Course in General Linguistics_, so it was a shock to learn from McElvenny that the “où tout se tient” wording does not occur there (though the “que des différences” quotation does). McElvenny attributes the former phrase to Antoine Meillet (1903), though E.F.K. Koerner (1996–97), who has gone into this issue thoroughly, points out that Meillet had already used “un système où tout se tient” in an 1893 paper. Meillet had been a student of Saussure’s in Paris in the late 1880s, and as McElvenny points out, Meillet felt that he was only giving expression to an idea which was already inherent in the 1879 _Mémoire_ that made Saussure’s name. Georg von der Gabelentz made a very similar remark in a paragraph inserted in the 1901 edition of his _Sprachwissenschaft_ (not included in the 1891 edition): “Jeder [sic] Sprache ist ein System, dessen sämmtliche Theile organisch zusammenhängen und zusammenwirken. Man ahnt, keiner dieser Theile dürfte fehlen oder anders sein, ohne dass das Ganze verändert würde.” Evidently this idea was “in the air” among linguists at the turn of the twentieth century, and Saussure may well have been responsible for introducing the idea, but it was Meillet who introduced the best-known expression of it into the published literature of linguistics.

The word “Limits” in McElvenny’s title is explained on his first page by saying that his book aims “to survey the conceptual boundaries of structuralism, to see where its borders may lie and how permeable they may be”; he asks “Who was part of this movement? Who was not?” This explanation is needed, because some of the reprinted authors are much less clearly members of the structuralist movement than others. Franz Boas is an important figure in the history of synchronic linguistics, but I would not have thought he had much or any connexion with structuralism in the Meillet/Saussure sense. However, McElvenny argues that Boas was a leader of a separate current of thought called “American structuralism”, distinct from the European, Saussurean tradition. McElvenny writes that “the American structuralists had a natural affinity for seeing languages as patterned formal systems” (though I wonder what linguists that would not apply to). On the other hand, Margaret Thomas in her introduction to the Boas chapter, while recognizing “American structuralism” as a separate intellectual current, denies that Boas was a structuralist even of that stripe.

The postulation of a distinct, not very clearly defined, “American” structuralism is one way in which this book displays a certain American bias. Another is that McElvenny sees little doubt that the central figure in the history of structuralism was Roman Jakobson. (And Patrick Sériot, introducing the Jakobson chapter, says that it is a “received opinion” that Jakobson “was an early promoter, if not the founding father of structuralism”.) McElvenny has found a passage in a Czech magazine from 1929 in which he believes Jakobson was the first person to use the word “structuralism” in the relevant sense. What matters, though, is not the first use of the word, but how the current of thought developed, and I would not see Jakobson as a central figure there. Jakobson was significant to Americans as the vehicle through which Prague School structuralism came to the USA, when he emigrated to New York in 1941, but I never noticed that Jakobson himself made large contributions to structuralism as an intellectual movement.

By now, the leading ideas in Saussure’s _Course_ have acquired such prestige that there is a danger of their being taken as self-evidently right (in which case structuralism would hardly merit identification as a distinct intellectual movement). So, in considering a book like McElvenny’s, it will be as well to begin by reminding ourselves that structuralist principles are in fact often questionable. To my mind they are not necessarily wrong, but they exaggerate the incidence of phenomena which do crop up in languages but are not the whole story.

Thus, John Ohala (2005) pointed out that the complex array of hidden psychologically-real relationships between phonemes postulated by structuralist phonology – matrices of pluses and minuses, “feature geometry”, etc. – are largely redundant because the facts they are used to explain often follow from the physical (articulatory, aerodynamic, or acoustic) properties of the sounds, which are factors comparable to the metal content of coinage and hence should be irrelevant in Saussure’s eyes. André Martinet (1955) held that sound-changes are controlled by the “functional yield” (or “functional load”) of phonemic oppositions, an abstract, relational property, but when Robert King (1967) tested this hypothesis in detail he found that “functional load, if it is a factor in sound change at all, is one of the least important of those we know anything about”. (See also Sampson 2017: ch. 12 on a recent attempt to resurrect Martinet’s idea.) Those who have extended structuralist principles to syntax believe that a native speaker’s usage is governed by complex ranges of algebraic grammar rules of which the speaker is unconscious, but empirical research casts doubt on whether natural languages have grammars (Sampson 2017: ch. 3). And so on. But mainstream linguistics continues to maintain a structuralist mindset, paying little attention to objections like these.

When one considers how little notice is taken of those who express scepticism about structuralist axioms, it is hard not to conclude that many academics and others simply _want_ language to be a phenomenon whose outward surface conceals a wealth of unobservable but psychologically real principles or rules, open to study only by initiates. McElvenny brings out the way in which, after structuralism was inaugurated near the beginning of the twentieth century, it “spread throughout the humanities and social sciences, reaching a peak mid-century” – or later: as I recall it was in the years about 1965 or 1970 when any new book with “structuralism” in the title was sure to find an eager audience among young people who were convinced that linguistics and related subjects were going to reveal to us previously-unguessed mysteries about the human mind. People might not have any expertise of their own in the relevant subjects, but they had a strong will to believe. It may not be a stretch to see structuralism, with its insistence that what matters about language and other human activities is abstract, unobservable mental patterns below the surface of physical reality, as one of the innocent-looking headwaters of what became a broader intellectual current that has ended with the humanities side of universities giving us a “post-truth” society, in which those who assert empirical facts such as the reality of immutable biological sex risk being “cancelled” or losing their jobs, in a way that had seemed to become unthinkable in the centuries since Urban VIII cancelled Galileo.

So we need to approach McElvenny’s (or any) account of structuralism in a sober frame of mind, asking ourselves not just whether the work of the movement is accurately described but whether that work did in fact lead to better understanding of the nature of human language, or of other aspects of mental life to which structuralist methods were applied.

EVALUATION

In evaluating a book of this type, relevant questions are whether the contributions are wisely chosen – are the authors adequately representative, and do the items selected from their output help readers to understand their contribution to the structuralist movement; and also, how well the chapter-introducers interpret the authors they introduce.

One surprising selection from the output of an author who certainly merits a place in the book is by far the most recent of the contributions, that by Emile Benveniste. This transcribes a talk given by Benveniste at a meeting which the Olivetti company convened in 1968 to burnish its public image. (Olivetti was then a successful typewriter manufacturer, the Apple of its day: its machines were works of art.) At that date farsighted industrialists knew that a computer revolution was round the corner, and Olivetti was already trying to carve out a place for itself within the new digital world. The meeting assembled a line-up of established academic “names” to give talks under the overall heading of “Languages in society and in technology”; I wasn’t there, but I imagine elegant surroundings and an invited audience of influential though not necessarily intellectual citizens. Benveniste’s contribution was the kind of thing that was probably expected: lots of buzzwords, not much solid content. I wonder whether Benveniste himself would have chosen to resurrect it fifty-odd years later.

(Meanwhile, IBM and Apple eventually rolled over Olivetti as they did over others. “Olivetti” today is merely a brand owned by Telecom Italia.)

A surprising omission, on the other hand, is no chapter by Nikolai Trubetzkoy. This is all the odder, when the chapter by Roman Jakobson, dating from 1931, is about the concept of _Sprachbund_, which I associate more with Trubetzkoy than with Jakobson. (This is the concept that a set of languages may show family resemblances not because they descend from a common ancestor but because, being spoken in contiguous territories, they have grown similar to one another.) The existence of _Sprachbünde_ is uncontroversial today, but it contradicted Neogrammarian orthodoxy and was a novel idea in the early twentieth century.

Rather than discussing the _Sprachbund_ concept in general, Jakobson’s chapter is almost entirely about one alleged example, and not a particularly plausible case: a _Sprachbund_ covering the languages of what had been the Russian Empire and, when Jakobson was writing, was the Soviet Union. It emerges that belief in this linguistic phenomenon was part of a much broader, rather sinister-sounding ideology (Vladimir Putin would love it) called Eurasianism, which asserted that the whole of that territory was (quoting Patrick Sériot’s introduction) “a _natural totality_ … characterized by a certain number of elements which united it [including] ethnic, economic, anthropological, human, geographical, cultural, linguistic, etc. features”, and naturally separate from the rest of Europe; according to Jakobson “the historical destiny of Eurasia confirms its indissoluble unity”. Eurasianism was apparently a formal association founded in Sofia in 1921 by four Russian émigrés, Trubetzkoy being one of them.

It is not for me to assess the non-linguistic aspects of Eurasianism (though I do wonder how far Jakobson continued to teach the historical destiny of the Soviet Union when he was a citizen of the 1950s’ USA). But the linguistic aspect is odd enough. He tells us that Soviet languages (as I shall call them for short) are united by sharing two properties: they are not tone languages, and they show a phonemic contrast in some or all consonants between what in Russian are called “hard” and “soft” variants, soft consonants being what modern phoneticians would call either palatal or palatalized consonants, depending whether closeness of tongue to hard palate is a primary or secondary articulation.

Jakobson notes that although these properties come close to coinciding with the Soviet borders, Eurasianism has been “aggressive” enough to extend the hard/soft contrast to Polish to the west and Japanese to the east (he claims that Japanese /rya/ is two phonemes rather than three, /ry/ being a soft counterpart to hard /r/). I suppose the case of Polish could be explained by the fact that much of Poland was once part of the Russian Empire. But Czech, he repeatedly says, lacks this contrast – as it should, since the Czech lands historically belonged to the Austrian rather than Russian empire. This is bewildering. If Japanese /rya ~ ra/ count as soft v. hard, I cannot imagine why Czech _řa ~ ra_, or _ti di ni_ v. _ty dy ny_, do not. I have noticed more than once before that linguists interested in Russian seem to get into muddles about the terms “palatal” and “palatalized” – see my review (Sampson 2023) of Bakró-Nagy et al. (2022), an otherwise excellent book – and this is surely another example. Jakobson does mention some languages of eastern Siberia which lack the hard/soft contrast; Sériot (footnote to p. 144) seeks to explain this away by saying that Jakobson was only really interested in the languages of European Russia. But if the Eurasianist idea about language boils down to “all Soviet languages are of type T, except for the ones that aren’t”, this is not much of a claim.

(Since the hard/soft contrast is so central to Jakobson’s chapter, it is unfortunate that where he illustrates a long series of minimal pairs, on p. 182, almost all the soft signs seem to have gone missing in the romanized transliterations.)

Remarkably, Jakobson explicitly asserts that the various Soviet languages have mutually assimilated not simply because their speakers influenced one another, but for “teleological” reasons: the languages were striving to express their shared Eurasianity. Polabian was a Slavonic language spoken in otherwise German-speaking territory which went extinct in the eighteenth century; it did have hard and soft consonant-pairs despite being far from Russia, but according to Jakobson “A language doomed to disappear often allows itself risky phonological experiments which are inaccessible to a language that is destined for wide expansion.” All this is certainly enlightening about the quality of Roman Jakobson’s thought, but does it inform us about the structuralist movement?

The writers who introduce the chapters have interpreted this task rather variously. Some of the introductions are useful brief accounts of how the chapter introduced relates to the wider state of linguistic thought at the time, and some go further, using the chapter introduced as a peg on which to hang a wide-ranging survey of the author’s work. (McElvenny’s introduction to the von der Gabelentz chapter is three times as long as that chapter.) Half of Lorenzo Cigana’s twenty-page introduction to the Hjelmslev chapter is about its publication history – an early MS was intended as a conference contribution but was rejected, and we are told exactly which committee members voted against it, how Hjelmslev arranged to publish a reworked version elsewhere, and so forth: material that might have been allowed to lapse into the oubliette of history without much loss. More than one chapter-introduction contain passages which are expressed too metaphorically for me to gather a clear meaning from them. I do not know what to make of a remark like Sériot’s “the emergence of European structuralism between the two World Wars is similar to a painful birth. One can distinguish different lineages, different lines of force.” When a woman’s labour is painful, it isn’t because she is unsure who the father is.

Some chapter-introductions contain factual errors. For instance, the first chapter after the editor’s introduction, by Peter Du Ponceau, is largely about polysynthesis in Amerindian languages. The introducer Floris Solleveld tells us that Du Ponceau had favourite examples of words in European languages which share some of this polysynthetic character; the first example he quotes is “Italian _nolo_ ‘I do not know’ ”. The word _nolo_ is Latin for “I do not want” – as Du Ponceau says where he uses the example in the piece Solleveld has translated. (_Nolo_ in Italian means “freight”.) We might all on occasion slip a mental cog and produce an error as crass as this, and then be so hypnotized by our own MS that re-reading fails to show us the mistake. But it is an important part of an editor’s job in a case like this to save contributors from themselves. (When I was a young man, OUP’s editors could be relied on to query such points if they slipped through as far as them, but I realize that those days are long past.)

Footnoting a passage in the Meillet chapter about the grammar of a French Bible quotation, “laissez venir à moi les petits enfants” (the Authorized Version has “Suffer little children to come unto me”), John Joseph objects that Meillet’s analysis is misguided because both the French and the English versions are only loose translations of the Greek original, which has _ta paidia_ with no word for “little”. This is quite beside the point, since Meillet was discussing the grammar of the French sentence (he did not mention the original); but in any case Greek _paidia_ is the plural of _paidion_ which is a diminutive of _pais_, “child”, so “little children” is a perfectly literal rendering.

At one place in Lorenzo Cigana’s translation from the French of the Hjelmslev chapter, Hjelmslev exemplifies an issue about uses of the imperative in a footnote which begins, in Cigana’s version, “Cf. Latin _ubi data occasionst, rape clepe tene_ (Plaute, _Pseud._ 138)”. Cigana has not realized that Hjelmslev’s “Plaute” is the French name for Plautus, and that _occasionst_ is a copying error for Plautus’s _occasiost_, a colloquial contraction of _occasio est_. The line from Plautus’s _Pseudolus_ means “Where an opportunity has been given (i.e. if you get a chance), rob, filch, grab, …”. Hjelmslev is making the straightforward point (much easier to understand than a lot of what he writes in this article) that although the verbs _rape clepe tene_ are singular imperative forms, in Plautus’s play they are not functioning as commands to a hearer but as a description of the behaviour of bystanders who are not being addressed. This has evidently passed Cigana by. Yet, if he does not take the trouble to follow Hjelmslev’s concrete examples, how can he hope to lead us through the mazy abstractions of Hjelmslev’s theorizing?

True, in Hjelmslev’s case some readers may feel that it is scarcely worthwhile to spend too much time trying to make sense of his strangely abstruse algebraic analyses of simple language phenomena. I am put off when a scholar boasts, using the royal we, that “Our work has no forerunners”, as Hjelmslev does here, and I sympathize with the view attributed by Cigana to Trubetzkoy that Hjelmslev’s work is full of “useless subtleties”. (But then, I have not offered to interpret Hjelmslev’s theory of linguistic correlations to the scholarly world.)

I am glad to have been put right about the authorship of the “où tout se tient” quotation. But this book did not leave me with a sense that I had learned much that I did not know before about the structuralist movement, or even perhaps that there is a great deal more to be learned.

REFERENCES

Bakró-Nagy, M., et al., eds. 2022. The Oxford Guide to the Uralic Languages. Oxford University Press.
von der Gabelentz, G. 1901. Die Sprachwissenschaft: ihre Aufgaben, Methoden, und bisherigen Ergebnisse (2nd edn). Tauchnitz (Leipzig).
King, R.D. 1967. “Functional load and sound change”. Language 43.831–52.
Koerner, E.F.K. 1996–97. “Notes on the history of the concept of language as a system ‘où tout se tient’ ”. Linguistica Atlantica 18.1–20.
Martinet, A. 1955. Economie des changements phonétiques. Francke (Bern).
Meillet, A. 1893. “Les lois du langage I: Les lois phonétiques”. Revue internationale de sociologie 1.311–21.
Meillet, A. 1903. Introduction à l’étude comparative des langues indo-européennes. Hachette (Paris).
Ohala, J.J. 2005. “Phonetic explanations for sound patterns: implications for grammars of competence”. In W.J. Hardcastle and J.M. Beck, eds, A Figure of Speech: a Festschrift for John Laver. Lawrence Erlbaum.
Sampson, G.R. 2017. The Linguistics Delusion. Equinox (Sheffield).
Sampson, G.R. 2023. Review of Bakró-Nagy et al. 2022. Linguist List 34.742, <linguistlist.org/issues/34/34-742/>.
Saussure, F. 1879. Mémoire sur le système primitif des voyelles dans les langues indo-européennes. Teubner (Leipzig).

ABOUT THE REVIEWER

Geoffrey Sampson graduated in Oriental Studies at Cambridge University in 1965, and studied Linguistics and Computer Science as a graduate student at Yale University before teaching at the universities of Oxford, LSE, Lancaster, Leeds, and Sussex. After retiring from his Computing chair at Sussex he spent several years as a research fellow in Linguistics at the University of South Africa. Sampson has published in most areas of Linguistics and on a number of other subjects. His recent books include "The Linguistics Delusion" (2017), "Voices from Early China" (2020), and "God Proofs" (2022).