Pre-structuralist semiology: materiality of language in Ferdinand de Saussure

1 Preliminary remarks

The name of Ferdinand de Saussure is traditionally associated with two theoretical paradigms in twentieth-century Western thought: semiotics and structuralism. However, in researching the role of materiality in Saussure’s project for a science of language, the present study will terminologically move away from both. In his writings, Saussure never used the term semiotics, and introduced as a science of semiosis a field he called semiology. In the subsequent tradition, those two terms are often used interchangeably since Saussure’s semiology is mostly identified as a linguistic counterpart of what emerged about the same time as semiotics in the context of American pragmatism (mainly due to the work of Charles Sanders Peirce). A more careful reading of both Saussure and Peirce, however, shows the fundamentally different nature of their projects. Semiotics and semiology existed for a while as concurrent or synonymous terms in the Continental context of the 1960s, with semiotics gaining the upper hand after the work of Roman Jakobson and Umberto Eco. In the present study, we choose to use systematically the terms semiology and semiological as more adequately rendering Saussure’s original conceptual framework about human signifying activity.

The second terminological preliminary point concerns the term structuralism. The problem of Saussure’s generally acknowledged involvement in structuralism deserves detailed research, which cannot be undertaken here. We will limit our remarks to already existing research on the topic of the emergence of the method of structuralism in the intellectual circles of the Prague linguistic school at the time of Roman Jakobson’s and Nikolai Trubetzkoy’s presence there.^[1]

Third, we must take into account the specific status of Saussure’s manuscript legacy. Today we know that the first decades in the reception of Saussure’s thought through the version of his Course in General Linguistics edited by his colleagues Charles Bally and Albert Sechehaye and published posthumously in 1916, provided a rather distorted rendering of Saussure’s chaotic and eccentric ideas in linguistics and studies of culture, channeling them back to more classical and even truistic interpretations. The critique of Bally and Sechehaye’s work started with Robert Godel’s (1957) research on the manuscript sources of Saussure’s linguistic conception in the 1950s, and was followed by important contributions like Rudolf Engler’s critical edition of the Course in the 1960s (Saussure 1989), and monographs by Konrad Körner, d’Arco Silvio Avalle, Johannes Fehr, Ludwig Jäger, and Beata Stawarska, among others.^[2]

Thanks to Engler’s critical edition of the Course collating the published version with a huge amount of other manuscript notes, and to the separate editions of Saussure’s three courses of lectures on general linguistics in the 1990s, today we can assess the actual scale of Bally and Sechehaye’s editorial intervention and its effect on the representation of Saussure’s ideas. However, the systematization of Bally and Sechehaye has to a certain extent become traditional and – even at the cost of significant displacements of some of Saussure’s ideas – has the merit of introducing a plausible map of the central questions that structured in one way or another the vast scope of Saussure’s scholarly interests. One of these points is the delineation of the notion of semiology as a separate realm of interest.

2 The notion of semiology and its development in Saussure

The notion of semiology appears for the first time in a published text by Ferdinand de Saussure in his posthumously published Course in General Linguistics (CLG, 1916). The editors, Charles Bally and Albert Sechehaye, placed it in the Introduction to the CLG, in Chapter III, Paragraph 3, entitled “Place de la langue dans les faits humains. Sémiologie” (‘The place of language among human facts. Semiology’; CLG 33, E 270).

In the critical edition of the CLG, edited by Rudolf Engler, semiology is simply defined as une science des signes (‘a science of signs’; E 277, II R 12).^[3] Saussure speaks about it entirely in the future tense, emphasizing that such a science doesn’t exist yet. The question of the limits and characteristics of semiology arises. Saussure refers to a book in epistemology by Adrien Naville, Nouvelle classification des sciences. Naville (1845–1930) was a philosopher and a friend of Saussure’s, his work was focused mainly on religious philosophy, but also on epistemology, and Nouvelle classification des sciences underwent no less than three editions: 1888, 1901, and 1920. Saussure’s reference to Naville and the passages of Nouvelle classification des sciences that discuss semiology have been researched by Tullio De Mauro (Saussure and De Mauro 2005a: 318–319), and especially by Engler (1974: 45–73, 1980: 3–16, esp. 4–6), and later by Joseph (2012: 460–462) who also highlights the differences between Saussure and Naville (mainly with respect to the position of semiology among other human sciences, especially psychology and sociology) while still tracing the genealogy of twentieth-century semiology back to their discussion (Joseph 2012: 536–538).

In the first edition of Nouvelle classification des sciences from 1888, semiology isn’t mentioned at all; neither is it mentioned in the third one from 1920, published more than fifteen years after Saussure’s death. The term semiology appears only in the second edition from 1901, and the footnote in the CLG mentioning Naville (CLG 34, E 297) refers precisely to this edition. Here we find three additional paragraphs on semiology and its place among other sciences. These passages are historically the first instance where the form semiology appears in print. In fact, Naville doesn’t claim to have invented the term – he directly recalls Saussure and in introducing semiology, seems to do some sort of marginal note on an idea of a friend of his:

M. Ferdinand de Saussure insiste sur l’importance d’une science très générale, qu’il appelle sémiologie ^[4] et dont l’objet seraient les lois de la création et de la transformation des signes et de leur sens. La sémiologie est une partie essentielle de la sociologie. Comme le plus important des systèmes de signes c’est le langage conventionnel des hommes, la science sémiologique la plus avancée c’est la linguistique ou science des lois de la vie du langage … La linguistique est, ou du moins tend à devenir de plus en plus, une science de lois; elle se distinguera toujours plus nettement de l’histoire du langage et de la grammaire. (Naville 1901: 104)

(‘Mr. Ferdinand de Saussure insists on the importance of a very general science which he calls semiology and the object of which would be the laws of the creation and transformation of signs and their meaning. Semiology is an essential part of sociology. Since the most important sign system of all is conventional human language, the most advanced semiological science would necessarily be linguistics, or the science of the laws of life of language … Linguistics is, or at least tends to become more and more, a science of laws; it will distinguish itself ever more clearly from language history and grammar.’)^[5]

So, despite the reference to an authority in epistemology (his friend Naville), it was Saussure who invented the notion of semiology. Thus, semiology is defined as “une branche des sciences relevant de la psychologie et de la sociologie” (‘a scientific branch pertaining to psychology and sociology’; E 284, II R 16). And further, “[c]ette psychologie des signes sera une partie de la psychologie sociale” (‘[t]his psychology of signs would be a part of social psychology’; E 285, D 7).

In Saussure’s view, the epistemological value of semiology is identified as pertaining to both psychology and sociology (Joseph 2012: 461).^[6] Furthermore, Saussure seems to be speaking of semiology as a science broader than linguistics^[7] and gives examples with other semiological systems such as rites and ceremonies, the rules of politeness, different writing systems, the alphabet of the deaf and dumb and that of the blind, military or marine signals (most of these examples are subsumed under the heading of “social institutions”). The development of such a science is thus described as necessitated by the existence of a potential scientific object called faits sémiologiques (‘semiological facts’; E 282, D 182, SM III 113, S 2.7).^[8] Saussure assumes that these are all sign systems and consequently, there must be common laws that govern them all, something that is shared by all kinds of sign systems, and of course, present in language as well. That is why, in accordance with Naville’s classification of sciences, semiology must belong to the so-called sciences de lois (‘sciences of laws’), more precisely, to its psychological part (Naville 1901: 23–39). These are exclusively theoretic sciences that focus on principles (theorems). In the case of semiology, these principles disclose themselves as psychological in nature, yet are not related to individual psychology. Semiological facts pertain to social psychology and are governed by socio-psychological laws (Naville 1901: 105–107).

3 Relation of linguistics to semiology

The relation of linguistics to semiology is one of the central problems in research on Saussure. Generally, scholars agree on the fact that the semiological dimension of language constitutes one of the central interests of the Swiss linguist. In a way, Saussure’s general linguistics (or as he preferred to put it, the science of language) may be considered as coinciding with his semiological project.^[9] The topic has been discussed at length by authors such as Avalle (1995: 31–76), Engler (1974, 1980, and Wunderli (1976: 42–44), and is elegantly summed up by Johannes Fehr:

Avec la sémiologie pour projet, avec cette entreprise difficile mais fascinante, d’introduire (à nouveau) le signe dans une linguistique qui lui avait donné congé en raison de sa réputation excessivement philosophique et spéculative, se trouve exprimé ce qui constitue tout à la fois la situation décalée de Saussure au sein de la linguistique et son importance dans l’histoire des idées, importance qui déborde largement son rôle de savant enfermé dans sa discipline. (Fehr 2000 [1995]: 49)

(‘With semiology as a project, with this difficult yet fascinating enterprise to (re)introduce the sign in a linguistic framework that has abandoned it due to its excessively philosophical and speculative reputation, we find expressed the trait that constitutes both Saussure’s transformed situation at the core of linguistics, and his importance in the history of ideas, an importance that goes far beyond his role of a specialist limited to his own discipline.’)

However, it is worth pointing out some thought-provoking details about the relation between (the ideal future) linguistics and the future general science of signs. At first, Saussure claims that semiology is supposed to be “broader than linguistics” (“une science plus large que la linguistique”; E 277, B 8) because it must encompass all other sign systems beyond natural language. This means that for Saussure, human signifying activity goes well beyond language, and to this extent, we can accept the attempts at interpreting Saussurean semiology as an early version of general semiotics. However, right after introducing the notion of semiology in his Second Course, Saussure underscores very carefully the fact that “[d]ans la sémiologie, la langue sera naturellement la science la plus importante” (‘[i]n semiology, [the science of] language will be, of course, the most important science’; E 290, G 1.2b):

Mais d’emblée, il faut dire que la langue occupera le compartiment principal de cette science; <elle en sera le patron général>. Mais ce sera par hasard: théoriquement, elle n’en sera qu’un cas particulier. (E 290, II R 12, SM II, 53)

(‘But from the outset, we must say that language will occupy the central compartment of this science; <it will be the patron of it all>. But this will be by accident: theoretically, it will be just another case.’)

Two central points must be highlighted here: language being the most important of all sign systems means that linguistics should be the model of every semiology. On the other hand, the opposite is also true – language is “just another” sign system; thus, linguistics should probably be only a form of special semiology. This mutual or double dependence between semiology and linguistics constitutes one of the key features of Saussure’s project for semiology, because it would mean that in fact semiology must be structured as linguistics. Why is this so important? We might assume that in developing a form of general linguistics, Saussure eventually developed a semiology. He points out that linguistics would be just a particular case of semiology, but nevertheless, probably, the most important one. Understanding the circular motion between linguistic and semiological speculation in Saussure’s thought (Fehr 2000 [1995]: 105–113)^[10] is essential to the reading of the material found in 1996, as we shall see later.

Manuscript material testifies that Saussure entertained the idea of semiology long before his courses in general linguistics (1906–1907), probably already about 1881 when he composed a sketch on phonetics (Jakobson 1969). The term appears in other manuscript notes such as those on a course in Gothic held at the École des Hautes-Études in Paris in 1885–1886, and in the treatise On the Dual Essence of Language (to be discovered in 1996; Chidichimo 2014: 108–112). Still, it is difficult to assess the exact level of precision of the definition of both general linguistics (science of language), and semiology at this early point in the development of Saussure’s thought. The difficulties in the dating of Saussure’s manuscripts and the hesitations he had about the terms he preferred do not allow us an unequivocal interpretation of the relation between these two fields (to reflect this specificity of Saussurean terminology, Giuseppe Cosenza introduces the notion “semi-manufactured terminology” or terms in process of terminologization, Cosenza 2016a: 128, 2016b: 178; see also Chidichimo 2014: 103; Linda 1998: 223–249). So, one might suppose that the clear delineation of both general linguistics, and semiology, might have come only later, by the time of the three courses (after 1906), or even doubt that there is any clear delineation.

In his First Course in general linguistics (1907), Saussure barely mentions semiology, only twice in passing, in the context of the relation between the spoken and the written language, and under the form of the adjective sémiologique – speaking about “two semiological axes” (the system of the spoken language and that of the written language; Saussure 1996: 10–11). This could mean that the notion was still not very well defined for himself, either. A more detailed discussion on the notion of semiology then follows in the Second Course (1908–1909), where Saussure introduces the term systematically in the beginning of the course, defining semiology as “la science des signes qui nous fait connaître en quoi peuvent consister les signes, leur lois, etc.,” (‘the science of signs that tells us what signs, their laws, etc., can consist of’; Saussure 1997: 113) and linking it directly to the notion of value.^[11] This is also visible in the development of the argument in the second part of the Third Course (1910–1911), where – only after discussing both linguistic geography and history – Saussure turns to basic theoretical questions, such as the nature of the linguistic sign (mentioning the notion semiology; Saussure and Constantin 2005). A very clear bipartition is visible in the structure of this course: a polemic text sequence refuting language as nomenclature, along with the whole conceptual apparatus that supports it;^[12] and then – an affirmative section developing the theory of linguistic value. The marginal mentions of the notion semiology seem to confirm the statements about the relation of linguistics and semiology from the Second Course, namely, that as a science, semiology should be modeled after linguistics: “Aucune série de signes n’aura une importance plus considérable dans cette science [sc. la sémiologie] que celles des faits linguistiques” (‘No other series of signs will have a more considerable importance in this science [semiology] than those of linguistic facts’; E 282, C III 274). It becomes clear that linguistic facts are considered subsumable under the idea of a sign, and then elevated to the series with the most important role within the realm of human signifying activity. Thus, when speaking of science of language, Saussure actually constructs the model of semiology: “La sémiologie aura à voir si elle doit s’occuper des signes arbitraires ou des autres; son domaine sera plutôt celui des systèmes de signes arbitraires dont la langue est le principal exemple” (‘Semiology should consider whether it must focus on arbitrary signs or on some other kind; its realm would be rather that of the systems of arbitrary signs, of which language is the principal example’; C III 281).^[13]

4 On the Dual Essence of Language and its importance

In 1996, a discovery was made: a folder containing until then unknown manuscripts by Ferdinand de Saussure was found at the Hôtel de Tertasse, one of the mansions of the Saussure family in Geneva. The exact circumstances of this discovery are described by Vincent Barras who was the first to gain access to the archives of Raymond de Saussure (Amacker 2011: 9–10). The section of the archive that belonged to Ferdinand de Saussure included a number of separate notes developed in the manner typical of the Swiss linguist – a piling up of unsystematic thoughts and observations.^[14] Since most of the folders in the batch bear the title De l’essence double du langage, or a version of this phrase, some scholars assumed that these writings might have been meant for publication or even a preparation of an actual treatise – more systematic than almost all the rest of Saussure’s written texts. The notes have been dated to the early 1890s (probably between 1891 and 1894, i.e., to a very early point in Saussure’s career as a linguist) and, together with some additional notes, were published for the first time in 2002, edited by Simon Bouquet, Rudolf Engler, and Antoinette Weil, under the title De l’essence double du langage, in the volume Écrits de linguistique générale, where the editors make an attempt to reconstruct a possible systematic treatise.^[15] The new material was swiftly integrated in research on Saussure’s legacy, and a critical edition of the materials was prepared by René Amacker and published in 2011 under the title De la double essence du langage with the more general designation Science du langage (Saussure 2011). Amacker provides detailed information about the condition of the notes and the general composition of the folders and files in the archive (Amacker 2011: 17–26) and rearranges the fragments with respect to Saussure’s own, albeit sparse, indications (Amacker 2011: 31–32). However, it is important to note that neither Engler’s, nor Amacker’s arrangement could be accepted as the original order intended by Saussure, nor that we are in fact presented with the text of an unfinished treatise. In both cases, the order is provided by the editors while the condition of the folders makes it practically impossible to give an exact assessment of the original order or of the exact significance of the manuscript within the Saussurean corpus. A substantial contribution to the reception of the presumed treatise was the commented German edition of 2003 by Ludwig Jäger (Saussure and Jäger 2003), as well as the commented Italian translation of 2005 by Tullio De Mauro (Saussure and De Mauro 2005b).

Especially interesting here is a volume of 2016, edited by François Rastier and entitled “De l’essence double du langage” et le renouveau du saussurisme, which makes an overall review of the materials discovered so far and offers a detailed reassessment of the whole Saussurean doctrine in the light of the latest findings. While reassessing the status and the importance of the manuscript entitled De l’essence double du langage (ED), Rastier claims that this text presents “un traité de linguistique qui dans une rédaction assez continue expose des principes théoriques et méthodologiques ordonnés par un projet unitaire” (‘a treatise in linguistics that develops in a fairly continuous manner some theoretical and methodological principles intended to serve a unitary project’; Rastier 2016: 15), a statement that could hardly be supported by the archive materials. Yet, Rastier successfully highlights some of the important aspects of Saussure’s linguistic and semiological thought:

Post-structuraliste à sa manière, Saussure élabore le projet rationnel de la linguistique mais le fonde ainsi dans une herméneutique de la complexité … Cette gnoséologie est fondée d’une part sur une préconception non substantialiste du réel. (Rastier 2016: 18)^[16]

(‘A poststructuralist of his own kind, Saussure develops a rational project of linguistics, but anchoring it in a hermeneutics of complexity … This gnoseology is based partially on a non-substantialist conception of the real.’)

The anti-essentialist (or, as Rastier puts it, anti-ontological) charge of Saussure’s thought is pointed out here with the strongest possible emphasis. Similar features in Saussure’s thought have been noted and made use of in numerous studies on the linguist’s legacy, but he has never been so explicitly defined as a poststructuralist.^[17] In the present study, we would prefer to use the term “pre-structuralist” as a suitable substitute for Rastier’s phrase “post-structuraliste à sa manière.” The connotations of the adjective pre-structuralist are to be understood along the lines of an approach that precedes structuralism and is in this sense non-structuralist or even counter-structuralist. So, the concept pre-structuralist is not intended to convey the meaning of something that anticipates structuralism.^[18] At the same time, it is to be thought of as related to actual structuralism in a manner very similar to the one in which poststructuralist tendencies are related to structuralism – not as a simple opposition and negation, but as a further development of structuralism, taking a next step in the same direction and accepting the inevitable consequences in an Aufhebung-type of movement. Pre-structuralism should be then a form of reversed or countertemporal Aufhebung of structuralism. From that perspective, we should posit the question of those elements of Saussure’s linguistic-semiological doctrine that ED could reveal as new to the contemporary reader; and next, how they could change our perception of human signifying activity in general.

The manuscript presents us with a theory of the nature of the linguistic sign, that we do not encounter in Saussure’s later manuscripts or in the three courses in general linguistics (1906–1911). Although some of its aspects could be reconstructed from other extant sources, mainly from Saussure’s notes and writings in general linguistics and on the Germanic legends,^[19] the version here includes some hitherto unknown moments that still remain not fully appreciated despite the increasing scholarly interest in the text. Weighing up the data from this manuscript against the backdrop of the already researched sign theory from the CLG and other manuscript notes, that came chronologically later, we can notice some important elements that could help us in a threefold way. First, the earlier version gives us the opportunity to develop an alternative theoretical model of semiosis – and especially of verbal semiosis, the semiosis in natural languages. Basically, this will help us provide answers to the seminal semiotic question of how meaning is produced that cannot be inferred from the courses in general linguistics. Second, as a consequence of the first point, it would procure a different understanding of hermeneutics, since knowing how signification comes about entails a certain reverse procedure of reading and interpreting meaning.^[20] And third, such an enterprise is a step towards restoring the importance of natural language and language theory in the field of general semiotics. These questions are going to be kept henceforth on the systematic level of the analysis, since the uncertainty of the chronology of the archive materials, as well as Saussure’s own doubts and hesitations that are reflected in the state of his manuscript materials do not allow us to formulate a conclusive statement about Saussure’s own final attitude towards these seminal semiological questions. The conceptual divergence of the earlier writings as compared to the later notes and the courses might be due to an inner development of Saussure’s thought or to the requirements of an introductory course for students – at this point, it is difficult to decide. Nevertheless, the conceptual elaboration of problems of general semiology in their version from On the Dual Essence could give way to alternative readings on the level of systematic semiology that still can contribute to a better understanding of the meaning-producing processes.

Assuming that for Saussure, a future linguistic theory (the theory of language as it should be, not as it was practiced at his time) is supposed to be interchangeable with general semiology (De Angelis 2022b), we shall address the most central of all linguistic questions according to Saussure: the question of the actual linguistic units. This is in fact an epistemological (or as Rastier prefers to call it, gnoseological) question and concerns the object of linguistic science. As Saussure notices at a very early point in his academic activity, the difference between linguistics and other sciences is that the objects of linguistics do not preexist the activity of the linguist. We are unable to speak of linguistic facts as existing objectively in the world (like geological, biological, or chemical facts), the research of which would form the scientist’s task. The objects of linguistics are practically created by the very activity of the linguist who is approaching them. Around the year 1900, this radically constructivist thesis was in fact something unusual and original. This strong anti-essentialist foundation of Saussure’s initial position towards the linguistic object could be detected throughout his research and teaching, but remained obscured by the edition of 1916 of the CLG by Bally and Sechehaye.^[21]

Saussure’s point of departure in the treatise On the Dual Essence is the claim that it is the gaze of the linguist that creates linguistic objects, and not vice versa. He introduces his central idea about the point of view as a dimension of particular importance.^[22] From this starting point, Saussure posits as his basic claim the following hypothesis: linguistic entities exist, and they can be scientifically determined.^[23] The delimitation and determination of what is to be perceived as a linguistic fact is now envisioned as the first and central objective of any serious theory of language: “ce qui est essentiel, c’est le problème des unités” (‘the essential thing is the problem of units’; SM 30). Many scholars have drawn attention to the fact that Saussure regarded the existence and the definition of linguistic units as a fundamental, but very uncertain, element of linguistics. Simon Bouquet (2016: 82–84), for example, supports the idea of identifying language units, linguistic objects, and semiological units. Thus, a single movement of delimitation would enable us to discover all three of them: the elements of language, the scientific object of linguistics, and the definition of the object of the new discipline of semiology. We read in the Second Course:

La linguistique aurait pour tâche de délimiter quelles sont <réellement> ces unités valables de tout genre. <On ne peut pas dire qu’elle s’en soit rendu compte, car elle n’a fait que discuter sur des unités mal définies.> Non seulement cette détermination des unités qu’elle manie sera la tâche la plus pressante de la linguistique, mais ce faisant elle aura rempli sa tâche tout entière. (Saussure 1997: 20)

(‘The task of linguistics should be to determine what its legitimate units of every type really are. We cannot say that it has ever taken this into account, because it has always discussed only units that have been wrongly defined. Not only is this determination of units which it performs the most urgent task of linguistics, but in doing so, linguistics would practically accomplish its entire task.’)

Bouquet identifies these units with semiological units (signs) based on Saussure’s claim that “[l]a signification seule permet de délimiter les unités!” (‘[o]nly signification allows us to delimit the units!’; E 1802, B 25). Moreover, the signifying language unit should in fact be defined not as a unity of an acoustic image and mental concept,^[24] as we are usually told, not even as a twofold entity composed of signifier and signified, but as a value (Bouquet 2016: 84–85). Language should be thus interpreted as a system of values that determine each other mutually.

Although we are not going to discuss Saussure’s notion of value in detail here, we will accept it as a central element of the theory of language as a signifying system, as a pinnacle of Saussure’s discussion on the tasks of the science de langage. In the Second and the Third Course he establishes that “je ne fais pas de différence fondamentale entre: une valeur, une identité, une unité, une réalité (sens linguistique), un élément concret linguistique” (‘I don’t make an essential difference between: a value, an identity, a unit, a reality [in a linguistic sense], a concrete linguistic element’; E 1803, B 31). Although these notions must be carefully distinguished from one another, they are established as the basic series, as the terminological inventions Saussure regarded as his actual contribution to general linguistics.

4.3 The self-mediating system

On pages 64–65 of the manuscript (Bibliothèque de Genève, Arch. de Saussure 372/2–8) Saussure revisits the problem of the relation of the signifier and the signified. He comes back to the four-sided entity of the quaternion:

Comment saisir l’extrême malentendu qui domine les raisonnements sur le langage ? On pose qu’il existe des termes doubles comportant une forme, un corps, un être phonétique – et une signification, une idée, une chose spirituelle. Nous disons d’abord que la forme est la même chose que la signification, et que cet être-là est quadruple. (SL 77: 148; parallel with modifications in ELG 7: 42)

(‘How can we grasp this extreme misunderstanding that dominates the reflections on language? It is usually stated that there are double terms connecting a form, a body, a phonetic entity – and a signification, an idea, a mental thing. We say, first, that form is the same thing as signification, and that the whole entity is in fact quadruple.’)

This extremely intriguing statement which explicitly identifies form and idea is followed by a diagram contrasting the traditional view of the sign as a two-faced entity, and the new conception proposed by Saussure (Figures 1–3).

Figure 1:

Saussure’s five-sided concept of sign as edited by Engler (ELG 7: 42).

Figure 2:

Saussure’s five-sided concept of sign as edited by Amacker (SL 77: 148–149).

Figure 3:

Bibliothèque de Genève, Arch. de Saussure 372/2–8: 65, facsimile.

Apart from rejecting the traditional idea of the sign as a two-faced entity, this diagram displays an alternative formulation of the basic constellation (three irreducible relations, four irreducible terms). What appeared in the earlier formulation^[40] as paratactic relations (sign-to-sign and signification-to-signification) is here reformulated as “general difference between forms (signs)” and “general difference between significations.” Thus, the series of forms and significations that we were presented with, seem now to be summarized under the heading of difference. Furthermore, there are two sets of differences that are related to each other: the general difference of forms and the general difference of significations.

We can also see that Amacker’s rendering of the scheme (Figure 2) is much closer to what is to be found in the manuscript than Engler’s representation (Figure 1). In the manuscript and in Amacker’s version, the lines that indicate the nature of the relations between the elements in the quaternion (column I) give the impression of an all-sided connection, each element being equally related to every other, while in Engler’s presentation column I looks rather similar to a group of two fractions with an indefinite relation to each other. It appears that Saussure had first drawn something like a coordinate system, organizing the position of each of the four irreducible terms, and then tried to cross out one of its sides – the line dividing une forme and une signification (this moment is indicated by Amacker in a critical note, SL 77: 149, n. 8). So, there is a relation between form and signification, signifier and signified – but in another sense, there is not. Whether there is such a relation or not is undecidable, and as it was pointed out earlier, “la forme est la même chose que la signification” – meaning that they inhabit the same topological space.

As for the general differences between forms and between significations, now we see that what appeared as limited series of significations or forms (abc, AHZ; un certain nombre de…) now seems to have been reformulated so that it could encompass the whole of the given significations or the given forms in a certain language – it is no less than a general difference. We might even say that it is equal to the principle of difference, the very fact that there are different forms or meanings in a linguistic system.

How should we read this strange quasi-coordinate system that regulates the all-sided reference of irreducible terms within the set of three irreducible relations? If the singular signifier and signified (une forme, une signification) are both aporetically identified and incapable of entering into a direct relation with each other, the link is constituted by the two sets of differences. Since the only direct relation of a singular form is that to the series of different significations, and the only possible relation of the singular signification is the one to the series of forms, we shall define the mediating element between them as the whole of the system. In a self-mediating system, each single signifying element in the system (the language) is determined as such by all other elements and by all existing relations between them. This indicates a recursive structure of the system. A similar interpretation is proposed by Tomaso Russo Cardona who prioritizes the notion recursivité on the basis of mathematical considerations (2008: 88, 95–99).

To give an example of a self-mediating system, let us imagine a very simple system of three elements: A, B and C. If the element A is determined by the elements B and C, but, in its turn, takes part in the determination of both B and C (the same scheme applying to both B and C), this would mean that A is already included in the external elements that determine it, and thus – determines itself from the outside. It is both internal and external to itself. This situation could be interpreted as recursion.^[41] Moreover, when we say determined by, we actually mean being different from, since the difference is the principle of mutual determination of elements in a semiological system.^[42] The meaning of the notion of difference in Saussure’s usage implies both the principle of relation and the complex of all relata. Difference is the basis of all unités linguistiques (‘linguistic elements’) and its importance could be inferred from many sections of the CLG. However, in On the Dual Essence, Saussure not only emphatically insists on the role of difference in language systems but describes it as: (1) fundamentally negative; (2) the element that enables signifiers and signified to enter into a mutual relation:

Principe fondamental de la sémiologie, ou de la « langue » envisagée régulièrement comme langue et non pas comme résultat d’états précédents Il n’y a dans la langue ni signes, ni significations, mais de DIFFÉRENCES de signes et des DIFFÉRENCES de signification; lesquelles 1^o n’existent pas les unes absolument que par les autres (dans les deux sens) et sont donc inséparables et solidaires; mais 2^o n’arrivent jamais à se correspondre directement. D’où l’on peut immédiatement conclure: que tout, et dans les deux domaines (non séparables d’ailleurs), est NÉGATIF dans la langue – repose sur une opposition compliquée, mais uniquement sur une opposition, sans intervention nécessaire d’aucune espèce de donnée positive. (ELG 22b: 70; SL 100: 181–182)

(‘Fundamental principle of semiology, or of “language” understood regularly as language and not as a result of previous states In language, there are neither signs, nor significations, but DIFFERENCES of signs and DIFFERENCES of signification; which, first, have no absolute existence except through each other (in both senses) and are therefore inseparable and solidary; but, second, they never come to correspond to each other directly. Hence, we could conclude: that everything, in both areas (not separable otherwise), is NEGATIVE in language – it rests on a complicated opposition, but only on an opposition, without any need for an intervention of a positive given.’)

On the one hand, we have this insistence on negativity and the purely differential character of the elements at all linguistic levels, but on the other – the idea that in this differentiating process, sound and meaning actually start relating to each other. So, without this all-permeating negative differentiality, no semiological process, no meaning production would ever come into being:

Ainsi, non seulement il n’y aura pas de termes positifs, mais des différences ; mais deuxièmement, ces différences résultent d’une combinaison de la forme et du sens perçu. (ELG 20b: 66; SL 51: 109^[43])

(‘Thus, not only aren’t there going to be any positive terms, but only differences; but, second, these differences are the result of a combination of the form with the <perceived?> meaning.’)

5 The material dimension of language

If all that matters in language as a system are negative differences, what is the role of the material aspect? One of the orthodox structuralist linguistic theses that have their origin in the CLG is the idea that both sides of the sign are of an ideal order – they are both of a psychological nature. A similar statement appears in ED:

[L]es deux éléments du mot sont réciproquement dans l’ordre spirituel ; notre point de vue constant sera de dire que non seulement la signification, mais aussi le signe est un fait de conscience pur. (ELG 2a: 19; SL 53: 116)

(‘[T]he two elements of the word are mutually within the mental order; our stable point of view would be to say that not only the signification but also the sign is a pure fact of consciousness.’)

This means that there is nothing material in language – even when speaking of its phonetics. The vocal aspect of language would then be only a mental representation of sounds, not the actual sounds we produce and perceive. It is important to stress here that this predicament concerns the phonological aspect of the sign as a sign. Yet, if we come back to the final quaternion and the four-sided diagram, we will notice that apart from the final quaternion (the actual semiological entity), there is a fifth element: column II, designated as “Figure vocale (servant de forme ou de plusieurs formes dans I)” (‘Vocal figure (serving as a form or as multiple forms to an entity in column I)’; see Figures 1–3).

What is the vocal figure? This is the level of sound regardless of its signifying function. This is the phonetic dimension outside the four-sided signifying constellation. The vocal figure pertains to the order of phonetics (or etymology). That is why Saussure prefers to define the vocal figure starting from the definition of form rather than proceeding in the opposite direction, defining the form on the basis of the vocal figure, because “il faut partir de la donnée semiologique” (‘we should be starting from the semiological fact’; ELG 6e: 37; SL 64: 132). So, although serving as a form, the vocal figure does not take part in the semiological dimension of language:

Une figure vocale devient une forme depuis l’instant crucial où on l’introduit dans le jeu de signes appelé langue, de la même façon qu’un morceau d’étoffe dormant à fond de cale devient un signal à l’instant où il est hissé 1^o parmi d’autres signes hissés au même moment et concourant à une signification ; 2^o entre cent autres qui auraient pu être hissés, et dont le souvenir ne concourt pas moins à [ ] (ELG 6e: 38; SL 65: 132)

(‘A vocal figure becomes a form at the crucial moment when it is introduced into the interplay of signs called language, just as a piece of fabric lying at the bottom of the ship’s hold becomes a signal at the moment it is hoisted 1. along with other signs hoisted at the same moment and conveying a signification; 2. from among a hundred others that could have been hoisted and the memory of which does not conform less to [ ]’)

Entering the semiological system, the purely material element starts playing the role of a signifying form. The problem that remains is the necessity of this new division – between form and vocal figure. If the form is of a psychological (or ideal) order (just as the signified notion) and the vocal figure is purely material, and if the one could be transformed into the other, this means that the purely material elements become in a sense im-materialized upon entering language.^[44] Moreover, this would mean that the signifying aspect of language is split into two – it could exist as pure matter, without relation to semiosis,^[45] and it could exist as a part of language, but then – it is no longer material.

The aporetic mode of this element is reflected in a note that isn’t a part of the treatise On the Dual Essence of Language but was found in the same folder with it in 1996. In Engler’s edition, these notes are designated as “Nouveaux Item” and placed at the end of the edition, while Amacker inserts them as preliminary notes at the beginning as “Nouvelles notes Item.” This preliminary position makes the reading of the notes in the framework of ED more difficult and encourages a reverse procedure – to read the treatise itself in the light of the Item. In one of these notes, Saussure introduces the notion of kénôme (ELG p. 93, SL 5: 56–57) – from the Greek κένωμα, meaning ‘that which is empty,’ an empty shell. This is the vocal figure. It is the material receptacle that could be inhabited by the living being of a sign (and here by sign I mean the semiological relation of relations in the quaternion). The note on the kenome discloses clearly the difference between form and vocal figure:

On commet cette erreur de croire [qu’il y a] 1. un mot comme par exemple voir existent en soi, 2. une signification, qui est la chose associée à ce mot. Or [ ], c’est-à-dire que c’est l’association même qui fait le mot, et que hors d’elle il n’y a plus rien.

La meilleure preuve est que vwar dans une autre langue aurait un autre sens. Mais, cela vu, il est donc bien clair que vous n’avez plus le droit de diviser, et d’admettre d’un côté le mot, de l’autre sa signification. Cela fait tout un. Vous pouvez seulement constater le kénôme et le sème associatif (ELG p. 93, SL 5: 56–57)

(‘One tends to make the mistake of believing [that there is] 1. a word, like for example voir [to see], existing in itself; 2. a signification which is the thing associated with that word. Now [ ], that is to say that it is the association itself that creates the word and that outside of it, there is nothing. The best proof is that in another language, vwar [the phonetic transcription of voir] would have another meaning. But once this has become obvious, it is now clear that you no longer have the right to divide and to allow, on the one hand, the word, and on the other – its signification. It would be all one. You can only affirm the kenome and the associative seme ’)

However fleeting it may be, the notion of kenome demonstrates that the key semiological question of how signifier and signified are related to each other becomes now slightly, yet crucially, displaced: it is transformed into a question about how the mental entity of the semiological relation of relations (in this case, the associative seme) could become flesh, could be incarnated in a kenome. In fact, if we come back to the distinctive fragment defining the actual meaning of linguistic dualism, we would notice that this is the very core of the language duality: “the duality of the vocal phenomenon AS SUCH and the vocal phenomenon AS A SIGN – of the physical (objective) fact and the physical-mental (subjective) fact, and in no way [in the duality] of the ‘physical’ fact of sound as opposed to the ‘mental’ fact of signification” (ELG 2d: 20–21; SL 37: 87).

As we indicated earlier, this passage contains Saussure’s most direct statement on linguistic dualism. Therefore, it is somewhat strange that it hasn’t attracted enough attention from scholars. The most profound analysis of this fragment belongs to Kazuhiro Matsuzawa who notices a peculiar inner contradiction in it, pointing out the actual perspectives that the fragment opens up onto the interpretation of Saussure’s ideas:

[O]n peut admettre un nouveau dualisme que Saussure formule ici contre le dualisme courant. Ce passage est particulièrement important, car on est souvent tenté de lier fort rapidement la dualité du langage à la bi-facialité du signe linguistique … La dualité est dans ce cas justement entre la figure vocale et le signe linguistique dans sa totalité … La formulation de Saussure reste équivoque … Ce flottement résulte du fait que la dualité de la figure vocale et du signe linguistique se répète en abyme dans le signe linguistique lui-même, comme une autre dualité, celle du signifiant et du signifié. Si Saussure accorde une attention toute particulière, non sans équivoque, à une telle dualité, c’est parce qu’il est impossible de concevoir le langage séparé complètement de la matérialité physique de la figure vocale. (Matsuzawa 2016: 110–111)

(‘[W]e can assume a new dualism that Saussure formulates here against the traditional one. This passage is particularly important because we are often tempted to connect too quickly the duality of language to the two-faced nature of the linguistic sign … In this case, the duality is precisely between the vocal figure and the linguistic sign in its totality … Saussure’s formulation remains equivocal … This uncertainty results from the fact that the duality of the vocal figure and the linguistic sign repeats itself to infinity within the linguistic sign itself as another duality, that of the signifier and the signified. If Saussure pays particular attention, not unequivocally, to such a duality, it is because it is impossible to consider language as completely separate from the physical materiality of the vocal figure.’)

This observation is crucial to understanding both the nature of the signifying relation according to Saussure, and the acknowledgement of the function of materiality in language. What is then the meaning of this duality of language? At first it seems to be the duality of signifier and signified, or mental and physical aspects of language – in fact, this conception is trivial and was already widespread among linguists and philosophers of language before Saussure. Here we see something else: Saussure rejects this traditional position (it is all the more inexplicable that he has gone down in history as its chief proponent) and replaces it by a more subtle analysis of the semiological relations in language.

The actual opposition in language is, thus, not the one between physical and mental, between le son et le sens (‘sound and meaning,’ as Roman Jakobson would put it a few decades later), but between physical and physical-mental, between two functions, two roles of the vocal phenomenon. Once, taken in its pure form, and a second time, taken as a semiological fact, in the role of what would be designated in the Third course as signifier. We see that this second, subjective, side is not only “mental” – it is “physical-mental.” This would mean that materiality is already included within the signifying relation, and appears there, as Matsuzawa puts it, en abyme. The quaternion is already of this hybrid, physical-mental nature.

The importance of the material dimension is explicitly visible in another series of notes that have been published by Rudolf Engler in volume IV of the critical edition of the CLG – these are the Item notes (E 3306–3324). Without going into detail about the significance of this set of notes in the Saussurean corpus (see Gandon 1995; Jäger 2010: 134–163), we will only mention that here, Saussure explores another terminological framework (again an experiment in the search of a better substitute for the ill-suited sound-concept opposition). It consists of the terms sème and sôme understood respectively as the inner semiological relation (the element that we encountered as “associative seme”) and its outer shell, its ‘body’ (the kenome). This sôme is defined as cadavre – in this case a ‘material body’ rather than a ‘dead body,’ and envelope – the ‘envelope’ or ‘wrapping’ of the semic element. One note in this series puts a special emphasis on the material element in the sign constellation:

<En me promenant>, je fais <sans rien dire> une encoche sur un arbre, comme par plaisir. La personne qui m’accompagne garde l’idée de cette encoche, et il est incontestable qu’elle associe deux ou trois idées à cette encoche dès ce moment, alors que je n’avais pas moi-même d’autre idée que de la mystifier ou de m’amuser. – Toute chose matérielle est déjà pour nous signe: c’est-à-dire impression que nous associons à d’autres, <mais la chose matérielle parait indispensable>. La seule particularité du signe linguistique est du produire une association plus précise que toute autre, et peut-être verra-t-on que c’est là la forme la plus parfaite d’associations d’idées, ne pouvant être réalisé que sur un sôme conventionnel. (E 3320.4, t. IV: 40)

(‘<While taking a stroll> I am making, <without saying anything,> a mark on a tree, just for fun. The person who is accompanying me notices the idea of this mark, and it is incontestable that from this moment on she associates two or three ideas with this mark although I have simply had the intention to joke with her or to amuse myself. – Every material thing is for us already a sign: that is, an impression that we associate with other impressions, <but the material element seems indispensable>. The only characteristic trait of the linguistic sign is that it produces an association more precise than any other, and maybe one day we will notice that it is here that the most perfect association is to be found, unable to be brought about otherwise than in a conventional sôme [body].’)

Statements like “Every material thing is for us already a sign” and “the material element seems indispensable” reveal the problematic and contradictive status of Saussure’s idea that “not only the signification but also the sign is a pure fact of consciousness.” It seems now that only the form within the quaternion could be perceived as “a pure fact of consciousness,” but the form in the role of a vocal figure still must be material. Nevertheless, they might be perceived as one and the same thing (just analyzed in different aspects). The exclusively fragmentary nature of Saussure’s manuscript legacy and the lack of almost any authorized publications prevents scholarship from extracting a stable, contradiction-free “doctrine,” and thus precludes any possible “Saussurean orthodoxy” like the one Bally and Sechehaye tried to establish. That is why it is more productive to understand the framework of a question posed by Saussure than to strive for an unequivocal answer. Therefore, we shall conclude the argument by stating that it is in the relation of the semiological unity to materiality, to the kenome of pure sound, that Saussure finds the neuralgic point of semiological reasoning. This is what he designates as duality. Paradoxically, and despite the insistence on the “mental character” of both signifier and signified, materiality emerges on both sides of the linguistic difference: it is a kind of “dead matter” as a piece of fabric sleeping (dormant) at the bottom of the ship’s hold, and a second time, as an animated, hoisted flag, it is already included in the four-sided relation of the semiological system, not on the level of the single form, but on the level of negative differences that recursively construct the system. In fact, Saussure doesn’t employ here the metaphors of living and dead; he would rather insist on the exchange of values according to the model of economic circulation where material entities acquire value not because of their singular inner nature, but insofar as they are involved in the system of exchange of values. Thus, the central contribution of Saussure’s revision of traditional linguistics concerns the re-tracing of this most important interior border of language as a signifying system, the one that separates the physical from the mental.

Abbreviations

B

François Bouchardy’s notes

C

Émile Constantin’s notes

CLG

Cours de linguistique générale. 1916, éd. Bally et Sechehaye

D

Georges Dégallier’s notes

E

Cours de linguistique générale. 1989 [1968], éd. critique Engler

ED

De l’essence double de langage

ELG

Écrits de linguistique générale. 2002, éd. Engler, Bouquet, Weil

G

Léopold Gautier’s notes

R

Albert Riedlinger’s notes

S

Mme A. Sechehaye’s notes

SL

Science du langage. De la double essence du langage. 2011, éd. Amacker

SM

Les sources manuscrites du Cours de linguistique générale de F. de Saussure. 1957, éd. Godel