The book continues a tradition of Saussure-exegesis in which German and French authors, such as Godel, Amacker, Engler, Wunderli and myself engaged, a tradition less apparent in English speaking countries where authors normally limit their interpretative efforts to the Cours de Linguistique Générale as published by Charles Bally and Albert Sechehaye (for an exception, cf. Thibault, cf. Nerlich 1997). All this might change as English translations of the first, second and third course of lectures on general linguistics that Saussure gave at Geneva are now available.
Although we cannot do justice to the (sometimes overwhelming) richness of detailed quotation and interpretation displayed in this book, we shall try to summarise it in its broad outline. Its aim is twofold: (1) to reconstruct the various meanings that the terms arbitrary and arbitrariness assumed in Saussure's work over time, that is in his early notes and in his three courses on general linguistics held at the University of Geneva between 1907 and 1910/11, with the last and third course being the most important record of Saussure's conception of the arbitrariness of the sign; (2) to analyse how the concept of arbitrariness was interpreted and, in some instances, distorted by later generations of linguists (who, one has to say in their defence, only had the Cours in the form published by Bally and Sechehaye at their disposition). So as to achieve the first task, the author had to interpret the concept of arbitrariness in the context of various narrower and wider semantic fields, that is its interconnection with other central concepts in Saussure's theory of language, such as lanue and parole and so on, and to contrast it with the complementary concept of motivation. The structure of the book follows neatly from its two goals.
After a short introduction the second chapter, entitled "Der sprachtheoretische Begriff der Arbitrarität", discusses Saussure's concept's of langage, faculté du langage, langue as nomenclature, and other conceptions of langue, parole and the interrelation between langue and parole. The most noteworthy part of this chapter is perhaps the lengthy discussion of the faculty of langage, which draws on a cornucopia of Saussurian notes, and the interpretation of this concept by later writers, such as Peter Wunderli. She writes: "Die 'faculté du langage' ist damit ein Komplex, der eine Reihe von 'untergeordneten' Fähigkeiten einschließt, d.h.. für die Konstituierung der 'langue' und für deren konkreten Vollzug auf der Ebene der 'parole'. Die 'faculté du langage' funktioniert sowohl bei der Rezeption als auch bei der Übertragung der sprachlichen Äußerungen, und trägt so zur Entstehung des Systems und zum sprachlichen Gebrauch bei." (p. 35).
The third chapter, entitled "Die Konzeption des sprachlichen Zeichens", deals with the concepts of image acoustique and concept, signifiant and signifié, their schematic representations, and their terminological history. After about 1894 Saussure had played with such terms as sème, symbole, mot, signe, aposème, forme, sôme, contre-sôme, anti-sôme, para-sôme, and parasème (cf. pp. 82-83). Garciá should have stressed more that this terminological tinkering actually brought about a real theoretical breakthrough in Saussure's thoughts, as the following quote (not fully given by Garciá:86) demonstrates: "Différence [ou a]vantage du terme de sème sur celui de signe . 10 <Pas essentiel [b.].> Sig[ne p]eut être = geste direct, c'est-à-dire hors d'un système et d'une convention. - <Sème = signe faisant partie d'un système [b.]>. - Sème = 10 signe conventionnel, - 20 signe faisant partie d'un système <également conventionnel>" (N15, CLG/E, II: 36). And: "Mais du reste il serait faux de dire que nous faisons une question très capitale de sème au lieu de signe. - Vérité est que parasème et aposème sont des notions capitales" (ibid., 3310.13) The aposeme is the material envelope of the seme, but what is more important, the term paraseme indicates that signs can only be studied as part of a conventional (arbitrary) system of signs. What this note shows is that around 1894/97 Saussure arrived at the conclusion that to give linguistics a sound methodological and theoretical foundation one had to abandon wholeheartedly the view that there is a natural link between sign and object and replace it by the view that the linguistic sign can only function as a sign because it entertains systematic links with other signs in a (conventional) system of signs. The thesis about the arbitrariness of the sign followed naturally from this insight.
Chapter 3 goes on to deal with such difficult notions of unité linguistique, entité linguistique and identité linguistique and throws important light on their definitions. It closes with a discussion of concepts introduced by some of Saussure's followers, such as moneme, morpheme and syntagm. The scene is now set for dealing with the problem of arbitrariness in detail.
Chapter 4 treats the arbitrariness of the sign itself and the link between the two sides of the sign. Chapter 5 deals with possible counter-arguments such as onomatopoeia and interjections. As BN was reading Pinker in parallel with Garciá, we cannot resist quoting the following fact that Saussure, had he known about it, would certainly have used with delight to shut up some of his critics: "Because of that arbitrariness, there is no hope that mnemonic tricks might lighten the memorisation burden, at least for words that are not built out of other words [what Saussure called relative arbitrariness]. Babies should not, and apparently do not, expect cattle to mean something similar to battle, or singing to be like stinging, or coats to resemble goats. Onomatopoeia, where it is found, is of no help, because it is almost as conventional as any other word sound. In English, pigs go 'oink'; in Japanese, they go 'boo-boo'." (Pinker 1994: 152). To get back to our summary: Chapter 6 deals with a central claim of the book, that is arbitrariness as a semiological principle. Chapter 7 discusses arbitrariness in the context of the change and stability of linguistic signs. Chapter 8 studies the relation between arbitrariness and value; chapter 9 relative and absolute arbitrariness, chapter 10 arbitrariness and diachrony, and chapter 11 arbitrariness as irraison. Chapter 12 is devoted to the discussion of arbitrariness after Saussure, in the Geneva School, glossematics, Martinet, Benveniste and in the context of the physei-thesei debate.
The book closes with a summary, in which the author lists the various meanings of arbitrariness and warns the reader not to use these divers acceptations indiscriminately as it has so often been done in the past. Garciá distinguishes between the following levels at which Saussure situates and uses the concept of arbitrariness: first of all there is semiological arbitrariness, which characterises any semiological system of signs whatsoever, but in particular systems of linguistic signs; secondly there is linguistic arbitrariness, which characterises the link between the two constituents of the binary sign; however, depending on the context, arbitrariness might be attributed to the signifiant, the signifié, the sign as a whole or the system of la langue; thirdly there is relative arbitrariness, or the morphological motivation of signs in a language; and lastly there is the fact that, from a diachronic perspective (the point of view of phonetic change), changes in signifiants are never motivated.
We would have liked to have seen a more detailed discussion of the arbitrariness thesis in the context of langue not only as a semiological system, but also as a social fact, especially in relation to the notion of conventionality. The sign as a whole exists only as a social fact. It is therefore 'independent' of the objects it designates, as Saussure pointed out in his early notes. There is no intrinsic link, especially not a natural one, between the sign and the object (cf. the quotes given on pages 83 and 84). The link between sign and object is purely conventional, but not conventional in the sense of some nomenclaturists, that is to say as based on a social agreement of any sort. There is agreement in some sense, that is, we all agree in our uses of a certain sign for a certain object, although, Saussure stresses, we never in fact agreed on using just that sign for that object (as one could say in German Übereinstimmung is not the same as Übereinkunft). The term convention is here used in the sense of a social practice, a custom, a use. In his later writings Saussure moves away from stressing the arbitrary and purely conventional link between the sign and the object and focuses on the inner constitution of the sign, on the link between signifiant and signifié. He now tries to dispel another confusion, associated this time not with the term convention, but with the term arbitrary. He stresses that arbitrary does not mean capricious, dictatorial, dependent on the individual will, the intentionality of those who use a sign. The speakers of a language never really have a choice; they cannot, as Wittgenstein would say, establish a private language. Private and convention or private and rule do not go well together. A rule, a convention, a sign must be social to be able to function in a community, where we rely on the fact that we can convert meanings into sounds when we speak and sounds into meaning when we listen and all this according to a certain conventional code. As Bréal said already in his arguments against those who regarded signs as living organisms: "Nos pères de l'école de Condillac, ces idéologues qui ont servi de cible, pendant cinquante ans, à une certaine critique, étaient plus près de la vérité quand ils disaient, selon leur manière simple et honnête, que les mots sont des signes. [...] ils n'ont pas plus d'existence que les gestes du télégraphe aérien ou que les points et les traits (.-) du télégraphe Morse". (Bréal 1924 [1897]: 255)
In short, arbitrary means neither 'conventional' in the sense of social contract, nor 'intentional' in the sense of individual choice. The link between signifiant and signifié only exists because the sign is used in this way by a social mass, and it is conventional in this sense only. The link between the sign as a whole and the object designated, or more specifically between the signifié and the signifiant can therefore never be a motivated one. That does not mean however that semiological systems based on the principle of arbitrariness are static or immutable. In fact, the opposite is the case: As there is no inherent necessity in language other than its arbitrariness, languages can change, they are always 'mutable', adaptable to the uses we make of them over time.
For Saussure, to acknowledge the arbitrariness of the sign is of paramount importance. Only when we accept this principle can we see how a language functions and changes. It is only because the semiological and social rules governing languages as systems of signs are arbitrary, that languages are not submitted to individual human will and whimsy. These rules are imposed on us by the language we learn to use. They are not freely accepted by us, instead we are trained to follow them when we learn a language. However, although these rules are taught normatively, they are not experienced as normative, but rather as a natural law (arbitrary means at the same time necessary). We cannot simply imagine alternatives, we take the rules for granted, they are self-evident for us, and what is more, we usually do not even notice them or that we follow them. This self-evidence and the certainty with which we follow the rules blindly, lies for us beyond the justified and the unjustified (cf. Heringer 1985:271). Every language is autonomous because it is arbitrary, and that means that it sets its own limits as well as the limits of our world.
Whether this is the right way of looking at the nature of language is, naturally, up for debate, as we have indicated at the beginning of this review. Whereas generativists (in the broadest sense) would wholeheartedly accept the doctrine of arbitrariness, cognitivists would not. Although their work (especially Langacker 1987) is based on Saussurean terminology (signifier, signified, etc.), their conception of the linguistic sign is more compatible with what Saussure (not Peirce) called symbol (they therefore speak quite often of symbolisation, 'symbolic constructions' etc.) We shall therefore conclude this review with a quote from Garciá's book in which Saussure writes about the symbol in his last and third course: "(1135) A propos du mot de symbole: (1136) Nous avons grand scrupule à employer ce terme. (1137) Le symbole a pour caractère de n'être jamais complètement arbitraire; le symbole n'est pas vide. Il y a un rudiment de lien entre idée et signe, dans symbole" (CLG/E: 155) (Garciá:106).
As an author who writes in various language which are not her native ones, BN sympathises with Garciá's plight, and admires her impeccable German. However, the French quotes in particular should have been checked thoroughly by a native speaker.
Heringer, Hans-Jürgen (1986): "Not by Nature nor by Intention: The Normative Power of Linguistic Signs" In; Linguistic Dynamics. Ed. Th. Ballmer. Berlin/New York, p. 251-278.
Langacker, Ronald W. (1987). Foundations of Cognitive Grammar, vol. 1: Theoretical Prerequisites. Stanford, Calif.: Stanford University Press.
Nerlich, Brigitte & David D. Clarke (1997). Review of Paul J. Thibault. Re-reading Saussure. The Dynamics of Signs in Social Life. London & New York: Routledge, 1997. Newsletter of the Henry Sweet Society for the History of Linguistic Ideas 28, 18-22.
Pinker, Steven (1994). The Language Instinct. The new science of language and mind. Harmonsworth: Penguin Books.
Saussure, Ferdinand de (1967-1974) Cours de linguistique générale. Édition critique par Rudolf Engler. Wiesbaden: Harrassowitz, 2 vols. (= CLG/E)
Sinha, Chris (1996). "Signifying Subjects". Paper presented at the Symposium Semiotic Systems and Semiosis organized by Christiane Moro and Cintia Rodriguez. IInd Conference for socio-cultural research Vygotsky-Piaget, Geneva 11-15 September 1996. (MS)
Tomasello, Michael (1995). "Language is Not an Instinct". Cognitive Development 10, 131-156.