Saussure and Chomsky

Foreword/Préface

It is very good news that contemporary specialists of the works of Saussure and Chomsky get back to a serious comparison between them. Even though a number of divergences may seem obvious, it is particularly refreshing that some key convergences are actually being uncovered, which were long hidden under the confusing layers of radical and sometimes superficial readings of Saussure. Such convergences were also waiting for some more recent evolutions of the Chomskyan paradigm to emerge more clearly.

Saussure is still often viewed by functionalists as a naturalist (for langue is an entirely psychic object organized with general principles in the brain, and, as Constantin writes, “langue is a part of language”, which is crucial since language is a “faculty” for Saussure) and by naturalists and formalists as a relativist or a pre-functionalist (because “langue is social in essence and independent of the individual” and “langue is a social institution”). I think that such opinions are built upon a very confusing usage of the term langue in writings like the Course, and more ideological than balanced interpretations of Saussure’s philosophy of language by following scholars.

Indeed, the concept of langue seems to be about two clearly different things. First (and foremost, perhaps), langue is the fundamental and universal set of principles determining natural language. It is a system of signs, whatever the actual language this system is filled with. Second, langue refers to specific languages; it is in that sense for example that langue is only fully realized by the mass of speaking subjects within a community.

Some quotes fuel the confusion. In the Introduction of the Course, we read that langue is “a treasure deposited by the practice of speech by the subjects who belong to the same community, a grammatical system virtually existing in each brain” (CLG, En. Trans., 1983: 13), where the two notions of langue seem to co-exist in the same sentence (and what follows in the text confirms the impression). Saussure himself seems hesitant about separating the fundamental principles of langue from the notion of particular languages. One explanation might be that langue-as-a-principle (a “system of signs”) is realized only in a given natural language. In any case, Saussure creates his particular notion of langue for the reason that according to him “language” is not a scientific object; yet langue ←7 | 8→makes it possible to understand the most essential apparatus of language. In that sense, langue-as-a-principle is not a social thing.

Then, is it reasonable to view the notion of langue taken as this set of universal principles as relating to the “language faculty”? Saussure in the third course calls language an instinct (Saussure, 1993: 69) that can’t be studied: “The fact is that we cannot explore the pigeonholes inside our brain”¹, so langue must be an idealized artifact standing for what, in the brain, can explain the possible shapes of languages. The answer is thus yes if one contemplates langue as the inner principles of language, i.e. that every language is a system of signs having differential values and that it is a system virtually existing in each brain (which cannot be understood otherwise than forming the essence of language), and which is grounded on a deeper faculty of association and coordination which drives acquisition (CLG: 29). But the answer is no, on the contrary, if one thinks that the faculty of language is rather conceived of by Saussure as restricted to the faculty of externalization by parole. This can be subject to infinite debates, the Saussurean corpus being fragmentary and reconstructed, but it makes at least sense that langue-as-a-principle is the essence of language, reason for which general linguists, i.e. theoretical linguists, should study it, not parole, not diachrony, not sociolinguistics, etc., according to Saussure.

To naturalists and formalists in general, cross-linguistic variation is interesting in itself but above all inasmuch as commonalities can be drawn, in turn opening to a better understanding of the basic principles and constrains at play in human natural language. That’s certainly what Saussure, with his original expertise in cross-linguistic comparison, both diachronic and synchronic, managed to do when he induced that language is fundamentally a system of signs. But, arguably, Saussure was not so much interested in syntax as in the lexicon, morphology and phonology (but see Rizzi in this volume on the similarities between Saussure’s morphology and Chomsky’s syntax; what is more, Saussure’s morphology, where words are not mere compounds but, so to say, higher-order units where a word like désireux should not be considered as désir + eux but rather, as he puts it, désir x eux – see Anderson 2018 – may be compared with merge). Words are the locus where phonological changes occur, and a way to make sense of such changes is to figure out their generality and consequences over a whole synchronic state of a language. Saussure’s intuitions in this respect appeared when he was 16 and he was attending his Greek class and made him discover the nasalis sonans. In the ←8 | 9→course of his reflections, this intuition only grew in the same direction and ultimately lead to the discovery of the system of vowels in IE languages, of laryngeals and ultimately lead him to the theory of langue as a system of signs. He had that day the intuition that the exception of athematic verbs in Greek, who have a flexion in -atai/-ato, while regular verbs have a flexion in -ntai/-nto were actually transformations on the basis of an original n but placed in a consonant context so that it had to be pronounced with a vocalization (hence the nasalis sonans), which over time moved and stabilized as an a. His idea developed further into the notion that this change from n to nasalis sonans to a was not an isolated phenomenon but rather part of a wider pattern of diachronic change operating on these successive synchronic stages, an idea that lead him ultimately to conceiving languages as organized networks of solidarities.

Unfortunately, from this rather commonsensical view that every language encompasses lexical and morphological arbitrary conventions, a number of scholars have been very quick in concluding that Saussure’s pre-structuralism paved the way to a relativist view of language (even though this would not be equated with the Whorfian tradition). That English river has a different meaning ←9 | 10→from French rivière, for the reason that rivière is limited in signification by the presence of fleuve, is an arbitrary, conventional, fact of language. Combine this to the notion of a sign in Saussurean terms, and you have the conclusion that the concept of “river” and that of “rivière” are constructed by the language, inasmuch as the signified is bound to the signifier by the structure of values within a language. The same reasoning applies to grammatical features like gender, number, etc. Saussure famously explains that language provides a system of “values” to the otherwise amorphic and chaotic thought. But nothing in Saussure’s thinking prevents from pinpointing the more atomic components of conceptual meaning, and this is even quite obvious, even though somewhat implicit, in Saussure’s own considerations. This is a route that Hjelmslev and others will open within structural semantics, not without limitations, but the important thing is that all this discussion is about categorization, not reference, and is only relatively incidental to what “concepts” proper can be in the mind. Thus the structure provided to thought by language is one of classification, which does not affect the grasping of reality; in that sense, language has no influence on “worldviews”, only on conceptual classifications. Reference is indeed not a topic of concern for Saussure, but this prudence of his has been overinterpreted beyond measure by subsequent scholars who made the unsubstantiated and harmful claim that Saussure “rejected” reference, adding a solipsist strawman that was to fit nicely with the relativist overinterpretation of the Course.