Some remarks on Chomsky’s reading of Saussure¹

Frederick J. Newmeyer
University of Washington, USA
fjn@u.washington.edu [Link]

Abstract

The point of departure of this paper is a puzzling reference to Saussure in Chomsky’s book Knowledge of language (Chomsky 1986). Taken at face value, this reference appears to reverse dramatically Chomsky’s earlier negative evaluation of a central aspect of Saussurean thought. However, I argue that, properly interpreted, the quotation does not in fact embody a changed assessment of Saussure on Chomsky’s part. Along the path to my conclusion, questions will inevitably arise regarding the compatibility of Saussure’s views of language with those of Chomsky’s. I deal with such questions in a somewhat scattershot manner, because the relationship (historical and intellectual) between Saussure’s ideas and those of Chomsky’s is too difficult and too open-ended for a short communication such as this one. For Chomsky, at least, we have an unbroken paper trail stretching over sixty years, documenting his positions and their historical evolution. But ascertaining what Saussure’s position was on a variety of central questions is a perhaps insurmountable challenge, given that he published very little and his personal notes (which continue to be discovered) are often too fragmentary to allow unambiguous interpretation.

Keywords: CLG, Knowledge of language, Chomsky, Saussure, history of linguistics ideas.

Algunas observaciones sobre la lectura chomskyana de Saussure

Resumen

El punto de partida de este trabajo es una referencia desconcertante a Saussure en el libro de Chomsky intitulado Conocimiento del lenguaje (Chomsky 1986). Tomada a primera vista, esta referencia parece invertir dramáticamente la anterior evaluación negativa de Chomsky de un aspecto central del pensamiento saussureano. Sin embargo, sostengo que, correctamente interpretada, la cita no encarna de hecho una evaluación cambiada de Saussure por parte de Chomsky. A lo largo del camino que lleva a mi conclusión, inevitablemente surgirán preguntas acerca de la compatibilidad de las concepciones de Saussure con las de Chomsky. Me ocupo de tales cuestiones de una manera un tanto dispersa, porque la relación (histórica e intelectual) entre las ideas de Saussure y las de Chomsky es demasiado difícil y demasiado abierta para una comunicación corta como ésta. Para Chomsky, por lo menos, tenemos una trayectoria ininterrumpida de artículos de por lo menos sesenta años, que documenta sus posiciones y su evolución histórica. Pero determinar cuál era la posición de Saussure sobre una variedad de cuestiones centrales es quizá un reto insuperable, ya que publicó muy poco y sus notas personales (que siguen siendo descubiertas) son a menudo demasiado fragmentarias para permitir una interpretación inequívoca.

Palabras clave: CLG, Conocimiento del lenguaje, Chomsky, Saussure, historia de las ideas lingüísticas

1. Introduction

The point of departure of this paper is a puzzling reference to Saus- sure in Chomsky’s book Knowledge of language (Chomsky 1986). Taken at face value, this reference appears to reverse dramatically Chomsky’s earlier negative evaluation of a central aspect of Saussurean thought. However, I argue that, properly interpreted, the quotation does not in fact embody a changed assessment of Saussure on Chomsky’s part. Along the path to my conclusion, questions will inevitably arise regarding the compatibility of Saussure’s views of language with those of Chomsky’s. I deal with such questions in a somewhat scattershot manner, because the relationship (historical and intellectual) between Saussure’s ideas and those of Chomsky’s is too difficult and too open-ended for a short communication such as this one. For Chomsky, at least, we have an unbroken paper trail stretching over sixty years, documenting his positions and their historical evolution. But ascertaining what Saussure’s posi- tion was on a variety of central questions is a perhaps insurmountable challenge, given that he published very little and his personal notes (which continue to be discovered) are often too fragmentary to allow unambiguous interpretation. Most problematically, Saussure did not himself write the seminal book that bears his name on the title page (Saussure 1916/2005). In fact, the two colleagues of Saussure’s who compiled the book did not attend the lectures on which it was based and examination of the notes by the students who did attend the classes has revealed a number of discrepancies with the published volume. All of this is well-known, of course, but it is worth stressing at the outset as a sort of advance apology for what might appear to the reader to be an excessive reserve on my part to reach definitive conclusions on Saussure’s ‘true’ opinions, and hence their compatibility with Chomsky’s².

The passage in Chomsky (1986) with which I am most directly concerned is the following:

It should be noted that familiar characterizations of ‘language’ as a code or a game point correctly toward I-language, not the artificial construct E-language. A code is not a set of representations but rather a specific system of rules that assigns coded representations to message-representations. Two codes may be different, although externally identical in the message-code pairings that they provide. Similarly, a game is not a set of moves but rather the rule system that underlies them. The Saussurean concept of langue, although far too narrow in conception, might be interpreted as appropriate in this respect. (Chomsky 1986: 31)

Chomsky appears in this passage to be identifying Saussure’s langue as a rule system that characterizes a speaker’s I-language, where I – (for ‘internalized’) language is ‘some element of the mind of the person who knows the language, acquired by the learner, and used by the speaker-hearer’ (Chomsky 1986: 22). E – (for ‘externalized’) language, on the other hand, is language ‘understood independently of the properties of the mind/brain’ (p. 20)³.

The problem to be addressed here is that not many pages earlier in Knowledge of language, Chomsky had unambiguously (it seems to me) assigned langue to E-language! The first paragraph of §2.2 of the book, entitled ‘Externalized language’, mentions that the term ‘grammar’ has been used in a variety of ways, after which Chomsky moves directly to approaches that (mistakenly, in his view) focus on E-language. One of them is ‘Saussurean structuralism’:

Structural and descriptive linguistics, behavioral psychology, and other contemporary approaches tended to view a language as a collection of actions, or utterances, or linguistic forms (words, sentences) paired with meanings, or as a system of linguistic forms or events. In Saussurean structuralism, a language (langue) was taken to be a system of sounds and an associated system of concepts; the notion of sentence was left in a kind of limbo, perhaps to be accommodated within the study of language use. (Chomsky 1986: 19)

The sentences that follow refer to the linguists Leonard Bloomfield and Zellig Harris and to the philosophers David Lewis and Willard Quine, all of whom, according to Chomsky, conceived of grammars as species of E-language. And on the page immediately after his remarks about langue being appropriately thought of as a system of rules, Chomsky remarked that ‘Saussurean structuralism had placed Jespersen’s observation about “free expressions” outside of the scope of the study of language structure, of Saussure’s langue’ (Chomsky 1986: 32). Since Jespersen, along with Humboldt, were the only historical figures cited in the chapter whose conception of language Chomsky deemed compatible with his notion of I-language, it seems clear to me that Chomsky was implicitly characterizing Saussure as being (merely) interested in E-language.

In other words, it would appear, superficially at least, that Chomsky sandwiched a positive appraisal of Saussure in between two negative appraisals (with respect to the same issue) in the same chapter of the same book. The positive appraisal —if that is what it was— stands out as an extraordinary one from Chomsky’s point of view. To the best of my knowledge Chomsky had never before referred to Saussure’s langue as a system of generative rules. The following quote from Aspects of the theory of syntax 21 years earlier states his longstanding position quite clearly⁴:

The distinction I am noting here is related to the langue-parole distinction of Saussure; but it is necessary to reject his concept of langue as merely a systematic inventory of items and to return rather to the Humboldtian conception of underlying competence as a system of generative processes. (Chomsky 1965: 4)

In the above passage, langue is characterized as an ‘inventory’, not as a rule system.

The paper is organized as follows. In §2 I trace Chomsky’s remarks on Saussure from their first mention in 1963 to the 1986 book that forms the nexus for this discussion. Section 3 raises (and then dismisses) the possibility that the account of Saussure in Anderson (1985) might have led Chomsky to more positively evaluate the former’s theoretical framework. In section 4 I make the case that Chomsky was only rhetorically attributing to Saussure the idea that langue was conceptualizable as an I-language based system of rules, a position that I reinforce in §5 by making reference to analogous rhetorical strategies in Chomsky’s earlier work. Section 6 traces the more ‘Saussurean’ direction of Chomsky’s recent work and §7 is a brief conclusion.

2. Chomsky’s evaluation of Saussure through the decades

If we look at the published record, Chomsky has been remarkably consistent with respect to his evaluation of Saussure, though he has highlighted or downplayed certain aspects of the latter’s thought over the years. Chomsky’s first published references to Saussure from the early 1960s could be described as ‘positive, but with negative undertones’⁵:

In a work that inaugurated the modern era of language study Ferdinand de Saussure (1916/1966) drew a fundamental distinction between what he called langue and parole… It is the child’s innate faculté de langage that enables him to register and develop a linguistic system (langue) on the basis of scattered observations of actual linguistic behavior (parole). Other aspects of the study of language can be seriously undertaken only on the basis of an adequate account of the speaker’s linguistic intuition, that is, on the basis of a description of his langue. This is the general point of view underlying the work with which we are here concerned… Our discussion departs from a strict Saussurian conception in two ways. First, we say nothing about the semantic side of langue… Second, our conception of langue differs from Saussure’s in one fundamental respect; namely langue must be represented as a generative process based on recursive rules. It seems that Saussure regarded langue as essentially a storehouse of signs… (Chomsky 1963: 327-328)

In other words, Saussure was on the right track, but his taxonomic preoccupations prevented him from appreciating the rule-governed creativity central to language. Nevertheless, this passage was positive enough that the editor of the critical edition of the Cours could describe it as ‘une veritable profession de foi saussurienne’ (De Mauro 1972: 400)⁶.

Chomsky’s Aspects, published in 1965, was his first book-length study entirely devoted to syntax since Syntactic structures (Chomsky 1957). Despite his noting that there are parallels between the competence-performance distinction and that between langue and parole (Chomsky 1965: 4), Chomsky primarily focuses on the syntactic deficiencies of Saussure’s approach to language:

It is worth noting that this naïve view of language structure persists to modern times in various forms, for example, in Saussure’s image of a sequence of expressions corresponding to an amorphous sequence of concepts… (7-8)

Chomsky continued in this vein throughout the 1960s:

There are, then, many varieties of taxonomic syntax; in particular, the few remarks that de Saussure offers concerning syntax indicate that he accepts this position… (Chomsky 1966: 22)

The great Swiss linguist Ferdinand de Saussure, who at the turn of the century laid the groundwork for modern structuralist linguistics, put forth the view that the only proper methods for linguistic analysis are segmentation and classification... In fact, Saussure in some respects even went beyond this in departing from the tradition of philosophical grammar. He occasionally expressed the view that processes of sentence formation do not belong to the system of language at all… (Chomsky 1968: 17)

Harris, and other methodologists of the 1940s, were developing an approach to linguistic analysis [including syntax — FJN] that one can trace at least to Saussure. (Chomsky 1969: 94, reprinted in Chomsky 1972: 194)

In other words, the more that syntax came to be central to Chomsky’s view of linguistic theory, the more critical he became of Saussure, whose remarks on this topic are, at best, ambiguous⁷. Nevertheless, at the time (and until about a decade later) Chomsky’s remarks vis-à-vis Saussure were interpreted positively enough that a historiographer of linguistics could assert —quite implausibly, in my opinionthat Chomsky’s Aspects model was ‘triggered by the appearance of the English translation of [Saussure’s] Cours in 1959 and therefore owes much to Saussure’s work’ (Koerner 1973: 12).

John Joseph (2002) has carried out a close reading of Chomsky’s references to Saussure from the 1960s through the 1980s, dividing them into several time periods, each characterizable by Chomsky’s overall ‘attitude’ to Saussure (see also Koerner 2002 for similar remarks):

(1) a. Chomsky (1963) the Saussurean

b. Chomsky (1962-64): Reaching further back

c. Chomsky (1965-79) the Anti-Saussurean

d. Chomsky (1986) and after: the Neo- Saussurean?

Joseph attributes Chomsky’s changing evaluation of Saussure to Chomsky’s desire to situate his view of language in the most propitious historical context. According to Joseph, in 1963 Chomsky strove ‘to align himself with a pre- Bloomfieldian precursor’ (Joseph 2002: 148). In the second period (1962-64) he began to identify his views with that of a more explicitly rationalist forebear, namely Humboldt, while at the same time drawing away from Saussure, whose ideas were ‘widely considered to be the cornerstone of structuralism, from which Chomsky now wished to distance his own position as much as possible’ (p. 150). This position intensified in the late 1960s and 1970s, as ‘Chomsky’s agenda had become just the opposite of what it had been in his 1963 article: to lump Saussure and the neo-Bloomfieldians together in one great “modern linguistics” demonology, framed by the Descartes-to Humboldt tradition and its generative revival’ (p. 151). Joseph drew the conclusion of Chomsky’s ‘neo-Saussureanism’ of 1986 from his face-value interpretation of the long quotation from Knowledge of language cited above and from Chomsky’s rather gentle critique of Saussurean syntactic theorizing, in which ‘the notion of sentence was left in a kind of limbo’. I present Joseph’s explanation of Chomsky’s apparent change of heart below in §6, in the context of a fuller discussion of Chomsky’s remarks about Saussure in his 1986 book.

In any event, I feel that Joseph provides reasonable partial explana- tions for Chomsky’s changing —and increasingly critical— remarks about Saussure over the twenty year period prior to 1986. But there is a simpler generalization available to explain Chomsky’s growing rhetorical stridency in that period: the more Chomsky was focused on syntax, the more critical he was of Saussure, who could not or would not incorporate syntactic theorizing into his theory of langue. The earlier (pre-Aspects) publications were extremely general in their overall content, so general that Saussure was a worthy precursor. Starting with Chomsky’s nearly exclusive focusing on syntax in the mid- 1960s, it was hardly surprising that Saussure would be transformed from revered antecedent to whipping boy. Consider the context of what has probably been Chomsky’s most negative one-sentence evaluation of Saussure, namely that ‘The impoverished and thoroughly inadequate conception of language expressed by Whitney and Saussure and numerous others proved to be entirely appropriate to the current stage of linguistic research’ (Chomsky 1968: 18). Context makes it clear that ‘the current stage’ was one in which syntactic analysis was not on the agenda because ‘There was no clear understanding a century ago as to how one might proceed to construct generative grammars that “make infinite use of finite means…”’ (Chomsky 1968: 18)⁸. Saussure’s conception was hence impoverished and thoroughly inadequate because it lacked a syntactic component. I am sure that it is true that even the most intellectually creative individuals are pleased to point to the giants of the past who in some way ‘anticipated’ their ideas. But the bottom line is that, as far as syntax is concerned, Saussure anticipated very little in the way of generative theorizing.

3. Did Chomsky (1986) draw from Anderson (1985)?

I repeat the central Chomsky quote here:

It should be noted that familiar characterizations of ‘language’ as a code or a game point correctly toward I-language, not the artificial construct E-language. A code is not a set of representations but rather a specific system of rules that assigns coded representations to message-representations. Two codes may be different, although externally identical in the message-code pairings that they provide. Similarly, a game is not a set of moves but rather the rule system that underlies them. The Saussurean concept of langue, although far too narrow in conception, might be interpreted as appropriate in this respect. (Chomsky 1986: 31)

Let us take this quote at face value, as Joseph did, and conclude that by 1986 Chomsky was interpreting Saussure’s conception of langue as an I-language-based rule system. The obvious question to ask is why Chomsky would have reevaluated his position. Given that no new relevant source material emanating from Saussure appeared in the decade prior to 1986, I can think of only one possible explanation. The year prior to the publication of Chomsky’s Knowledge of language saw the appearance of Stephen R. Anderson’s book Phonology in the twentieth century (Anderson 1985). Anderson presents an unorthodox and, to my knowledge, unique view of Saussure’s phonological theorizing, in which, for Saussure:

…a grammar provides a system of rules, or principles particular to a particular language, which characterize some of these representations as (potentially) belonging to different signs, and others as (potentially) belonging to the same sign in Saussurean terms. … the rules of language… are particular to that language, and, taken together, they characterize the system by which sound differences correspond to oppositions between signs. Saussure’s point, formulated in these terms, is clear: it is the business of the linguist to study not the nature of (phonetic) representations but the system of rules which underlies the differentiation of signs and thus constitutes a particular language. (Anderson 1985: 34-35; emphasis in original)

Anderson goes on to write:

We must emphasize that, while Saussure had no sympathy for a description of alternations which posited unitary underlying forms and rules altering the character of segments, he certainly considered alternations to be a rulegoverned aspect of sound structure. Rather, he took the rules involved to be ones which directly related one surface form (in a given language) to another, without assigning priority to either (or setting up an indirectly attested third form from which both are derived). As such, all of his rules have the character of ‘lexical redundancy rules’ (in the sense of Jackendoff 1975) or ‘correspondences’ (in the sense of Lopez 1979). (Anderson 1985: 52-53)

Anderson gives a number of illustrative examples of Saussure’s rule- based approach. Perhaps the most interesting is taken from the notes from one of his Greek and Latin phonology courses (see Reichler-Béguelin 1980). As a result of the historical change of rhotacism in Latin, that language displayed [s] – [r] alternations, as in honōs/honōrem. Working through the exposition in Saussure’s notes, Anderson concludes that the ‘alternation between [s] and [r] under determinate conditions… was, for Saussure, reflected as a rule of the grammar of Latin’ (Anderson 1985: 54).

Is it possible that Chomsky had read Anderson’s book, leading him to reevaluate Saussure’s overall approach? The answer is ‘Almost surely, no’. For one thing, Knowledge of language had gone to press before the appearance of Phonology in the twentieth century. Secondly, Anderson informs me (p. c.) that he had not sent Chomsky a pre-publication manuscript of the book, nor did they have any exchange of letters or emails with respect to the book, either before or after publication. And third, if Chomsky had nevertheless had access to the book, it would have been very unlike him to fail to cite one of his own former students on an issue of such magnitude. One must conclude, then, that Anderson’s reevaluation of Saussure played no role in any (putative) reevaluation of Saussure by Chomsky.

4. Understanding Chomsky’s 1986 quote

In my view, Chomsky on page 31 of Knowledge of language does not attribute to Saussure the view that langue was an I-language-based system of rules. The key word in the quotation is ‘interpreted’. The overall context of the citation, as well as Chomsky’s prior rhetorical practice (see below, §5), indicate that what Chomsky really means is ‘reinterpreted’. In other words, he is saying (in effect), ‘Saussure’s grammar was a version of E-language. However, if we want to be charitable, we can think of it as a model of I-language and take off from there’.

The strongest piece of evidence for my interpretation (aside from the fact that it gets Chomsky off the hook for gross inconsistency) is that he overtly follows the same rhetorical strategy with respect to the philosopher Willard Quine in the same section. Quine, Chomsky asserts, is interested only in E-language:

Quine, for example, has argued that it is senseless to take one grammar rather than another to be ‘correct’ if they are extensionally equivalent, characterizing the same E-language, for him a set of expressions (Quine 1972). (Chomsky 1986: 20)

Now consider the last sentence of the Chomsky quote under analysis and the sentence that follows it:

The Saussurean concept of langue, although far too narrow in conception, might be interpreted as appropriate in this respect. The same is true of Quine’s definition of a language as a ‘complex of present dispositions to verbal behavior’ insofar as it focuses on some internal state rather than E-language, although it is unacceptable for other reasons. (Chomsky 1986: 31)

Chomsky, who had just been charitable to Saussure, is offering the same gesture to Quine. He has not suddenly and dramatically changed his opinion of Quine’s theorizing. Rather Chomsky is saying that despite Quine’s own personal goals, there is no impediment to reinterpreting the end product of those goals in a way more congenial to his world view.

In any event, in an email message to Chomsky I presented to him the 1965: 4 quote and the 1986: 31 quote back to back and asked him: ‘Did you reevaluate your interpretation of Saussure between 1965 and 1986, or are people misunderstanding what you wrote later?’ He replied:

There’s no change in my views. The phrase ‘might be interpreted’ refers to a possible sympathetic interpretation, in the specific context under discussion. (Noam Chomsky, personal communication 20 June 2012; emphasis in original)

It is an interesting question why Chomsky might have offered a ‘possible sympathetic interpretation’ of Saussure’s langue in this context. I suspect that it derives from his specific goals in the opening chapters of Knowledge of language. In these chapters, Chomsky is not motivating a particular theory of grammar, but rather a particular approach to grammar, one which situates our grammatical competence in the mind of the individual. He had for over two decades argued for such approach, of course, but nowhere previously had contrasted (what he called in this book for the first time) I-language approaches from E-language approaches. Earlier, for the most part, he had simply accused the bulk of his twentieth century antecedents of proposing empirically incorrect grammatical theories, not with proposing epistemologically incorrect theories. But in chapter 2 of Knowledge of language Chomsky devotes several pages to arguing that Saussure, his structuralist successors, and the philosophers Lewis and Quine had E-language-style grammars in mind, not I-language grammars. Since ‘the concept of E-language, however construed, appears to have no significance’ (1986: 31), he found himself in the awkward position of appearing to dismiss the studies of his predecessors as insignificant, if not utterly worthless. Clearly that has never been his attitude toward such individuals. Hence Chomsky’s noting (in the case of Saussure and Quine, at least) that their work could be evaluated much more favorably if one reinterprets them as sharing his I-language-based goals, however ‘narrow’ and ‘unacceptable’ their reinterpreted theory might remain.

5. Chomsky’s earlier rhetorical reinterpretations

Page 31 of Knowledge of language was not the first instance where we find Chomsky rhetorically attributing goals similar to his own to those whose work he was critiquing. In fact, it had long been his standard practice. For example, he begins his key chapter of Syntactic Structures, ‘On the goals of linguistic theory’, with the claim that ‘a grammar of the language L is essentially a theory of L’ (Chomsky 1957: 49). He goes on to discuss requirements ‘that could be placed on the relation between a theory of linguistic structure and particular grammars’ (50). From strongest to weakest they comprise a ‘discovery procedure’ for the theory, a ‘decision procedure’, and an ‘evaluation procedure’. Chomsky then writes:

As I interpret most of the more careful proposals for the development of linguistic theory [footnote omitted — FJN], they attempt to meet the strongest of these three requirements. That is, they attempt to state methods of analysis that an investigator might actually use, if he had the time, to construct a grammar of a language directly from the raw data. (Chomsky 1957: 52; emphasis added)

Here for the first time we find Chomsky being charitable to his adversaries⁹ by attributing to them the same conception —that of regarding a grammar of a language as a theory of that language— that he himself had. Very few linguists at the time would have described their aims in such a manner, a point driven home by the Voegelins, who remarked that ‘It [the argumentation employed by transformationalgenerative grammarians — FJN] places models of their own making as constructs followed by their predecessors and thereby distorts history’ (Voegelin & Voegelin 1963:22)¹⁰. Be that as it may, Chomsky’s rhetorical maneuver was successful in resetting the ground rules for debate over the adequacy of competing approaches to grammar.

Chomsky later adopted a similar rhetorical strategy with respect to the end product of earlier structuralist theorizing. In fact, much of Chomsky (1964b) in its various versions was devoted to looking at the syntactic and phonological analyses published by (mainly) post-Bloomfieldians and arguing that, as far as syntax was concerned, they could have been produced by a context-free phrase grammar (and were therefore inadequate) and, as far as phonology was concerned, they could have been produced monolevel rules of segmentation (and were therefore inadequate). Chomsky cautioned:

It should be noted, however, that modern generative grammars are typically not conceived as generative grammars, but as descriptive statements about a given corpus (text). Hence the taxonomic model, as described below, is no more than an attempt to formulate a generative grammar which is in the spirit of modern procedural and descriptive approaches. (Chomsky 1964b: 11)

In other words, he is once again reinterpreting the work of his adversaries, so as better to demolish it. Chomsky’s strategy was adopted and greatly elaborated in two book-length studies by Paul Postal, Postal (1964) and Postal (1968), which dealt with syntax and phonology respectively. Even a sympathetic reviewer of the former book could note:

For him [i. e., Postal, and, by parity of argument, Chomsky as well— FJN] the charge that a grammar is a phrase-structure grammar and, therefore, considered as a generative device, demonstrably weaker than a transformational grammar, is the strongest criticism that can be made of it. For this reason he is prepared to ignore all the very great differences in ideas concerning the methods and aims of linguistics that separate him from these linguists, and to interpret their work as if it was intended to fulfill exactly the conditions that he himself would prescribe for linguistic research, in order, so to speak, to set them up for the final knock-out blow. The result is a view of the last thirty years of American linguistics which makes it look as if linguists in this period were groping towards the (in fact, entirely original) position adopted by Chomsky in Syntactic Structures. It is hardly surprising, therefore, that, viewed from this standpoint the most striking characteristics of their work should appear to be its imprecision, inconsistency, and an amazing tolerance of anomalies. (Thorne 1965: 75; emphasis added)

By the early 1970s Chomsky and his cothinkers found little reason to continue to train their guns on the Post-Bloomfieldians and other structuralist thinkers outside of the now wellestablished generative tradition. The conflict had become an internal one: Chomsky’s main opposition was now composed of practitioners of the version of transformational generative grammar called ‘generative semantics’. In fact, for almost a decade the developers of this trend were numerically dominant within the world of formal theorizing (see Newmeyer 1986 for discussion). Chomsky’s first assault on this tendency took place during his 1968-1969 class lectures at MIT, which were published as Chomsky (1971). Here again Chomsky attributes to his opponents propositions that they do not themselves accept, so as better to demolish them. He writes:

Thus the deep structures, in this theory, are held to meet several conditions. First, they determine semantic representation. Second, they are mapped into well-formed surface structures by grammatical transformations (without any subsequent insertion of lexical items). Third, they satisfy the set of formal conditions defined by base rules;… I will refer to any elaboration of this theory of grammar as a ‘standard theory’, merely for convenience of discussion and with no intention of implying that it has some unique or conceptual or empirical status. Several such elaborations have been proposed and investigated in the past few years. (Chomsky 1971:185)

Chomsky goes on to argue that several proposals in the nascent genera- tive semantic approach (McCawley 1968a; 1968b; 1970; Lakoff 1968; 1970b) are examples of such ‘elaborations’ of the standard theory, which, examined closely, turn out to be no more than ‘notational variants’ of analyses formulable in that theory and hence incorrect¹¹. Yet manifestly, McCawley and Lakoff in those papers had rejected all three of the conditions comprising ‘a standard theory’¹². In other words, once again Chomsky was interpreting the work of others in ways that they would not accept (and did not accept — see Lakoff 1971) as a rhetorical strategy for combatting the proposals in this work. In the long run, I would say, this strategy was quite successful (see again Newmeyer 1986).

6. More on Chomsky (1986) and Beyond

As noted above, Joseph (2002) proposes an alternative explanation for Chomsky’s apparently considering Saussure’s views on I-language in 1986 to be highly congenial to his own. Joseph begins cautiously by noting that ‘[w]e are still too close to the changes in question to be certain that we are judging them accurately and impartially’ (Joseph 2002: 153). But he does go on in the following pages to offer a tentative judgment on why Chomsky would have become more ‘Saussurean’ by 1986. Joseph points out that over the previous twenty years Chomsky had progressively de-emphasized the contrast between deep and surface structures and had therefore turned back ‘in the direction of the sort of “flat” structure which once had made Saussure the possessor of an “impoverished and thoroughly inadequate conception of language” …’ (Joseph 2002: 153). Joseph then goes on to note that as syntax has become more ‘minimal’ in recent years, much of the work carried out by the rules of syntax in earlier models is now performed by features of the lexicon. The minimizing of the syntax in favor of the lexicon constitutes, as Joseph notes, a resuscitating of a key element of Saussure’s view of language.

What Joseph writes about the evolution of Chomsky’s ideas is essentially correct as an overall outline. In my opinion, however, it is inadequate as a proffered explanation for Chomsky’s wording on page 31 of Knowledge of language. In a nutshell, 1986 is too early for Chomsky to have been led to the conclusion that his views were (re)converging with those of Saussure’s. The amount of highly abstract syntactic machinery proposed in that book is about as un-Saussurean as any set of proposals for grammatical analysis could possibly be. Just to give a few examples, in Chomsky (1986) we find a distinction proposed between chains and CHAINS, where the latter include chains and expletive-argument pairs (p. 132), clitics moving in the level of logical form (p. 156), and an elevenline licensing condition for binding (p. 171). It was really only in the mid 1990s that Chomsky began to extol the minimization of the work of the (narrowly defined) syntactic component.

I am aware of only two works that refer to Saussure in Chomsky’s post-1986 writings. Both are positive and both seem to reflect the kind of convergence that Joseph (2002) had in mind. In the first, Chomsky notes that since the pared down genetically-determined initial state of the human language faculty ‘permits only a restricted variety of I-languages to develop’ (Chomsky 2000: 27), it follows that ‘variation of I-languages may reduce to Saussurean arbitrariness (an association of concepts with abstract representations of sound) and parts of the sound system, relatively accessible and, hence, “learnable” (to use a term with misleading connotations)’ (p. 27). Later in the book, Chomsky becomes even more explicitly Saussurean, remarking that ‘…language variation appears to reside in the lexicon. One aspect is “Saussurean arbitrariness,” the arbitrary links between concepts and sounds’ (p. 120).

One of Chomsky’s recent goals has been to attribute as much as possible to ‘third-factor explanations’, that is, those not specific to the faculty of language. The idea is that the more that can be attributed to external faculties, the more stripped down and minimalist the language faculty itself. The historically most prominent of such explanatory devices are the various economy principles, which have undergone development since around 1990. Surprisingly perhaps, Chomsky cites Saussure as the inspiration for such principles:

Much work in the structuralist tradition already suggested that the organization of linguistic inventories obeys certain economy principles (see Williams 1997 for a recent discussion in terms of the Blocking Principle of the Saussurean idea that ‘dans la langue il n’y a que des différences’ [in langue there are only differences]). (Chomsky 2002: 31)

The Blocking Principle (whose generative origins are in Aronoff 1976) prohibits certain lexical formations if another of equivalent meaning exists (glory blocks *gloriosity, for example) as well as a wide range of possible anaphoric dependencies. Interestingly, Williams appeals to Saussure’s notion of ‘value’ as the inspiration for the Blocking Principle, which is one of the few instances that this notion has ever been cited by a generative grammarian, and certainly the only highly positive instance.

7. Conclusion

Chomsky in his 1986 book Knowledge of language appears for the first time to attribute to Saussure the idea that langue is a system of internalized rules. However, that was not Chomsky’s intention. The apparent attribution was nothing more than the use of a rhetorical device of a type that Chomsky had availed himself several times in prior publications.

References

Anderson, Stephen R. (1985), Phonology in the twentieth century, Chicago, University of Chicago Press.

Aronoff, Mark (1976), Word formation in generative grammar, Cambridge, MA, MIT Press.

Chomsky, Noam (1957), Syntactic structures, The Hague, Mouton.

Chomsky, Noam (1962), A transformational approach to syntax. In A. HILL (Ed.), Proceedings of the Third Texas Conference on Problems of Linguistic Analysis in English (p. 124-158). Austin, University of Texas Press.

Chomsky, Noam (1963), Formal properties of grammars. In R. D. LUCE, R. R. Bush and E. GALANTER (Eds.), Readings in mathematical psychology, Volume II (p. 323- 418). New York, John Wiley & Sons.

Chomsky, Noam (1964a), Current issues in linguistic theory. In J. A. FODOR and J. KATZ (Eds.), The structure of language: readings in the philosophy of language (p. 50- 118). Englewood Cliffs, NJ, Prentice-Hall.

Chomsky, Noam (1964b), Current issues in linguistic theory, The Hague, Mouton.

Chomsky, Noam (1964c), The logical basis of linguistic theory. In H. G. LUNT (Ed.), Proceedings of the Ninth International Congress of Linguists (p. 914-977). The Hague, Mouton.

Chomsky, Noam (1965), Aspects of the theory of syntax, Cambridge, MA, MIT Press.

Chomsky, Noam (1966), Topics in the theory of generative grammar. In T. A. SEBEOK (Ed.), Current trends in linguistics, Volume III: Theoretical foundations (p. 1-60). The Hague, Mouton.

Chomsky, Noam (1968), Language and mind, New York, Harcourt, Brace & World.

Chomsky, Noam (1969), Linguistics and philosophy. In S. HOOK (Ed.), Language and philosophy (p. 51-94). New York, New York University Press.

Chomsky, Noam (1971), Deep structure, surface structure, and semantic interpretation. In D. STEINBERG and L. JAKOBOVITS (Eds.), Semantics: An interdisciplinary reader in philosophy, linguistics, and psychology (p. 183-216). Cambridge, Cambridge University Press.

Chomsky, Noam (1972), Language and mind: Enlarged edition, New York, Harcourt Brace Jovanovich.

Chomsky, Noam (1986), Knowledge of language: Its nature, origin, and use, New York, Praeger.

Chomsky, Noam (2000), New horizons in the study of language and mind, Cambridge, Cambridge University Press.

Chomsky, Noam (2002), On nature and language, Cambridge, Cambridge University Press.

De Mauro, Tullio (1972), Saussure et Chomsky, in Cours de linguistique générale by F. de Saussure, édition critique préparée par Tullio de Mauro (p. 400-404). Paris, Payot.

Godel, Robert (1957), Les sources manuscrites du Cours de linguistique générale de F. de Saussure, Paris, Minard.

Godel, Robert (1966), F. de Saussure’s theory of language. In T. A. SEBEOK (Ed.), Current trends in linguistics, volume 3: Theoretical foundations (p. 479-493). The Hague, Mouton.

Godel, Robert (1970), Théorie de la phrase, La sintassi. Atti del III convegno della Società di Linguistica Italiana, p. 11-41.

Hill, Archibald A. (ed.) (1962). Proceedings of the Third Texas Conference on Problems of Linguistic Analysis in English. Austin, University of Texas Press.

Hymes, Dell, and John FOUGHT (1981), American structuralism, The Hague, Mouton.

Jackendoff, Ray (1975), Morphological and semantic regularities in the lexicon, Language 51, p. 639-71.

Joseph, John E. (2002), From Whitney to Chomsky: Essays in the history of American linguistics, Amsterdam, John Benjamins.

Joseph, John E. (2012), Saussure, Oxford, Oxford University Press.

Koerner, E. F. K. (1973), Ferdinand de Saussure: Origin and development of his linguistic thought in western studies of language. A contribution to the history and theory of linguistics, Braunschweig, Vieweg.

Koerner, E. F. K. (2002), Toward a history of American linguistics, London, Routledge.

Lakoff, George (1968), Instrumental adverbs and the concept of deep structure, Foundations of Language 4, p. 4-29.

Lakoff, George (1969), On derivational constraints, Chicago Linguistic Society 5, p. 117-139.

Lakoff, George (1970a), Global rules, Language 46, p. 627-39.

Lakoff, George (1970b), Pronominalization, negation, and the analysis of adverbs. In R. A. JACOBS and P. S. ROSENBAUM (Eds.), Readings in English transformational grammar (p. 145-165). Waltham, MA, Ginn.

Lakoff, George (1971), On generative semantics. In D. STEINBERG and L. JAKOBOVITS (Eds.), Semantics: An interdisciplinary reader in philosophy, linguistics, and psychology (p. 232-296). New York, Cambridge University Press.

Lopez, Barbara (1979), The sound pattern of Brazilian Portuguese, Unpublished Ph. D. dissertation, UCLA.

Mccawley, James D. (1968a), Lexical insertion in a transformational grammar without deep structure, Chicago Linguistic Society 4, p. 71- 80. Reprinted in Grammar and meaning, by James D. MCCAWLEY (1976), p. 155- 166. New York, Academic Press.

Mccawley, James D. (1968b), The role of semantics in grammar. In E. BACH and R. HARMS (Eds.), Universals in linguistic theory (p. 125-170). New York, Holt, Rinehart and Winston.

Mccawley, James D. (1970), Where do noun phrases come from?. In R. JACOBS and P. ROSENBAUM (Eds.), Readings in English transformational grammar (p. 166-183). Waltham, MA, Ginn.

Newmeyer, Frederick J. (1986). Linguistic theory in America: Second edition. New York, Academic Press.

Postal, Paul M. (1964), Constituent structure: A study of contemporary models of syntactic description, Bloomington, IN, Indiana University Research Center in Anthropology, Folklore, and Linguistics.

Postal, Paul M. (1968), Aspects of phonological theory, New York, Harper and Row.

Postal, Paul M. (1970), On the surface verb ‘remind’, Linguistic Inquiry 1, p. 37-120.

Quine, Willard V. (1972), Methodological reflections on current linguistic theory. In D. DAVIDSON and G. HARMAN (Eds.), The semantics of natural language (p. 442- 454). Dordrecht, Reidel.

Reichler-Béguelin, Marie-José (1980), Le consonantisme grec et latin selon de Saussure: Le cours de phonétique professé en 1909-1910, Cahiers Ferdinand de Saussure 34, p. 17-97.

Ross, John R. (1970), On declarative sentences. In R. JACOBS and P. ROSEN- BAUM (Eds.), Readings in English transformational grammar (p. 222-272). Waltham, MA, Ginn.

Saussure, Ferdinand de (1916/1966). Course in general linguistics. New York, McGraw- Hill. [Translation of Cours de linguistique générale. Paris, Payot, 1916].

Saussure, Ferdinand de (1916/2005). Cours de linguistique générale. Publié par Charles Bally et Albert Séchehaye avec la collaboration de Albert Riedlinger. Edition critique préparée par Tullio DE MAURO. Paris, Payot.

Thorne, James P. (1965), Review of Constituent structure by Paul Postal, Journal of Linguistics 1, p. 73-76.

Voegelin, C. F., and F. M. VOEGELIN (1963), On the history of structuralizing in 20th century America, Anthropological Linguistics 5, p. 12- 35.

Williams, Edwin (1997), Blocking and anaphora, Linguistic Inquiry 28, p. 577-628.

¹ Artículo publicado en Anderson, J. Moeschler, F. Reboul (eds.), L’Interface Langage-Cognition: Actes du 19e Congrès International des Linguistes, Genève: Librairie Droz 2013, pp. 233-252.

² For a recent state-of-the-art documentation of the evolution of Saussure’s ideas, see Joseph (2012).

³ I-language and ‘(grammatical) competence’ are roughly equivalent notions. However, E-language should not be confused with ‘performance’ or with Saussurean parole. E-language is a way of conceptualizing grammar, while performance and parole pertain to the actual use of language.

⁴ Some commentators, however, reject the idea that Saussure conceived of langue as merely a systematic inventory. For Robert Godel, for example, the difference between langue and competence ‘est surtout terminologique’ (Godel 1970: 35). For a brief discussion of Saussure’s view of syntax, see footnote 6 below.

⁵ See also Chomsky (1964a; 1964b; 1964c), which differ from each other in certain aspects, but are practically identical with respect to their opinions of Saussure. Space limitations prevent me from supplying all of Chomsky’s published references to Saussure.

⁶ De Mauro, however, lists a number of ways in which he believes that Chomsky’s ideas differ from Saussure’s: Saussure is more interested than Chomsky in research methodology, in the ethnographic side of languages, in semiological links between language and other systems, in semantics, in working out a theory of parole / performance, and in explaining the relation between the biological unity of the human species and the historical diversity of languages.

⁷ It is not easy to determine where syntax fit in for Saussure. One could not ask for a more explicit attribution of syntax to parole than the following: ‘La phrase est le type par excellence du syntagme. Mais elle appartient à la parole, non à la langue …’ (Saussure 1916/2005: 172). But Saussure hedges a bit in the following paragraph. For example, ‘les locutions toutes faites’ like à quoi bon and prendre la mouche belong to langue, presumably because, being memorized as wholes, they mimic fixed elements of morphology and lexicon. But then, more problematically, Saussure writes that ‘il faut attribuer à la langue, non à la parole, tous les types de syntagmes construits sur des forms regulières’ (p. 173). Under one interpretation of that position, all rule-governed syntax could be encompassed under langue. He concludes the section with an agnostic: ‘Mais il faut reconnaître que dans le domaine du syntagme il n’y a pas de limite tranchée entre le fait de langue, marque de l’usage collectif, et le fait de parole, qui depend de la liberté individuelle’ (p. 173).

⁸ One should not draw the conclusion that Saussure opposed formalism in principle. For example, in his notes we find the comment that ‘Il n’y a pas, il ne peut y avoir d’expressions simples pour les notions linguistiques. L’expression simple sera algébrique ou elle ne sera pas’ (Godel 1957: 49). According to Robert Godel, Saussure’s ‘dearest concern was to cast the theory of language into the rigid mold of a mathematical treatise’ (Godel 1966. 481).

⁹ Or uncharitable, depending on one’s point of view.

¹⁰ For a somewhat amusing consequence of the lack of shared theoretical goals between the early Chomsky and more established linguists, it is instructive to read the exchange between him and Archibald Hill after his presentation of the paper ‘A transformation- al approach to syntax’ (Chomsky 1962) (see especially pages 158- 160 of Hill 1962). Hymes & Fought (1981) has a useful overview of issues and debates surrounding the cited quote from Syntactic Structures, though I disagree with its interpretation on many points.

¹¹ ‘Hence incorrect’ because later in that paper Chomsky elaborates a number of serious difficulties with the standard theory.

¹² Other papers published by generative semanticists before 1971 were even clearer in their rejection of Chomsky’s three conditions: Lakoff (1969; 1970a); Postal (1970); Ross (1970). Some of these might have appeared in the journals after Chomsky’s 1971 paper went to the printers, but Chomsky was well aware of their contents, since they were presented at lectures at MIT and Harvard as early as 1968.