2. Ferdinand de Saussure¹

2.1 Ferdinand de Saussure's Cours de linguistique generale (Course in General Linguistics) is a peculiar book, not merely published but in part composed after the author's death. Since he ‘destroyed the rough drafts of the outlines used for his lectures’, the editors, Charles Bally, Albert Sechehaye, and Albert Riedlinger, used ‘the notes collected by students’ in order to ‘attempt a reconstruction, a synthesis’, and to ‘recreate F. de Saussure's thought’ (CG xviiif). To ‘draw together an organic whole’, the editors tried to ‘weed out variations and irregularities characteristic of oral delivery’, and to ‘omit nothing that might contribute to the overall impression’ (CG xix). Thus, the ‘Saussure’ of the Cours is a composite voice, speaking from a lecture platform between 1897 and 1911 and passing through the notebooks of followers who confess that ‘the master’ ‘probably would not have authorized the publication of these pages’ (CG xvii, 38, xviiif). Many problems with its formulation and interpretation may reflect the difficulties of its composition.

2.2 Saussure -- or ‘Saussure’, as I should write perhaps -- seems fully conscious of his role as founder of a ‘science’. He constantly searches for generalities, high-level abstractions, and fundamental definitions. Over and over, he states what is ‘always’ or ‘never’ the case, what applies in ‘each’ or ‘every’ instance, what are the ‘only’ relevant aspects, and so on. At times, these universalizing assertions may go beyond what can be demonstrated, or conflict with each other in puzzling ways.² Formulating the common denominators of Saussurian ‘thought’ can thus be quite challenging.

2.3 His ‘hesitation to undertake the radical revision which he felt was necessary’ in linguistics seems to have deterred him from writing a general book; in fact ‘he could not bring himself to publish the slightest note if he was not assured first of the fundamental foundations’ (Benveniste 1971: 33). In a letter to Antoine Meillet dated 4 January 1894 he proclaimed himself ‘disgusted’ ‘with the difficulty’ of ‘writing ten lines concerning the facts of language which have any common sense’, and with ‘the very great vanity of everything that can ultimately be done in linguistics’ (ibid., 33f). He lamented ‘the absolute ineptness of current terminology, the necessity to reform it, and, in order to do that, to show what sort of subject language in general is’. In the Cours, he still finds ‘current terminology’ ‘imperfect or incorrect at many points’, and its components ‘all more or less illogical’ (CG 44). Still, he often proposes and defends terms with bravura, and many of these have become standard. And he ‘does not hesitate to use’ ‘the expressions condemned’ by ‘the new school’ he envisions (CG 5n) (cf. 2.30).³

2.4 Like most of the theorists in my survey, Saussure was highly discontent with the state of the discipline (cf. 13.3). He charged that ‘no other field’ was so beset by ‘mistakes’, ‘aberrations’, ‘absurd notions, prejudices, mirages, and fictions’ (CG 7, 3f, 97, 215). He deplored ‘the confusion’ ‘in linguistic research’ as well as the ‘absurdities of reasoning’, and the ‘erroneous and insufficient notions’ created by his predecessors (CG 99, 4f) (cf. 2.10). The intent to found a new direction can sharpen such polemics, especially when established ‘schools’ ‘watch the progress of the new science suspiciously’ and each ‘mistrusts the other’ (cf. CG 3).

2.5 ‘Before finding its true and unique object’, ‘the science that has been developed around the facts of language passed through three stages’ (CG 1) (cf. 4.4ff; 8.6-9, 15; 12.22-26; 13.4-8).⁴ First, the ‘study’ of ‘“grammar”‘ was ‘based on logic’, but ‘lacked a scientific approach and was detached from language itself’. Preoccupied with ‘rules for distinguishing between correct and incorrect forms’, grammar ‘was a normative discipline, far removed from actual observation’. Second, ‘classical philology’ was devoted to ‘comparing texts of different periods, determining the language peculiar to each author, or deciphering and explaining inscriptions’ (CG 3, 1). This approach ‘followed written language too slavishly’, ‘neglected the living language’, and focused on ‘Greek and Latin antiquity’ (CG 1f). Third, ‘comparative philology’ explored the relatedness of many languages, but ‘did not succeed in setting up the true science of linguistics’, because it ‘failed to seek out the nature of its object of study’ (CG 2f). Also, ‘the exaggerated and almost exclusive role’ ‘given to Sanskrit’ was a ‘glaring mistake’ (CG 215) (cf. 4,4, 40; 8.4f, 74, 8⁶; 12.20f).

2.6 Although (or because) he owed so much to it,⁵ Saussure was especially critical of ‘philology’, the historical study of language. Because ‘modern linguistics’ ‘has been completely absorbed in diachrony’ (i.e., issues of ‘evolution’), its ‘conception of language is therefore hybrid and hesitating’; this ‘linguistics’ ‘has no clear-cut objective’ and fails ‘to make a sharp distinction between states and successions’ (CG 81f). In contrast, ‘the “grammarians” inspired by traditional methods’ at least tried to ‘describe language-states’. Though ‘traditional grammar neglects whole parts of language’, does not ‘record facts’, and ‘lacks overall perspective’, ‘the method was correct’: however ‘unscientific’, ‘classical grammar’ is judged ‘less open to criticism’ than ‘philology’ (cf. 13.4). Now, ‘linguistics, having accorded too large a place to history, will turn back to the static viewpoint of traditional grammar, but in a new spirit and with other procedures, and the historical method will have contributed to this rejuvenation’ (CG 82f) (cf. 2.15; 6.49; 7.4; 8.38; 12.41, 88; 13.7). In effect, ‘general linguistics’ would become a ‘true science’ by supplying the theoretical and methodological framework absent from earlier approaches, while drawing freely on their findings and examples.

2.7 Saussure envisioned ‘linguistics’ taking its place among ‘other sciences that sometimes borrow from its data, sometimes supply it with data’ -- e.g., ‘political history’, ‘psychology’, ‘anthropology’, ‘sociology’, ‘ethnography’, ‘prehistory’, and ‘palaeontology’ (CG 102f, 147, 9, 6, 224) (cf. 13.9-20). Yet ‘linguistics must be carefully distinguished’ from such sciences, which can contribute only to ‘external linguistics’, concerning ‘everything that is outside’ the ‘system’ of ‘language’ (CG 6, 9, 20f) (cf. 2.9; 13.9). In return, ‘we can draw no accurate conclusions outside the domain of linguistics proper’ (CG 228).

2.8 On a grand scale, Saussure foresaw ‘a science that studies the life of signs within society’, and ‘called it semiology’ (CG 16). ‘Linguistics is only a part of that general science’ and is charged with ‘finding out what makes language a special system within the mass of semiological data’. ‘If we are to discover the true nature of language, we must learn what it has in common with all other semiological systems’ (CG 17) (cf. 6.50-56; 12.9f). For Saussure, ‘language, the most complex and universal of all systems of expression, is also the most characteristic; in this sense linguistics can become the master-pattern for all branches of semiology’ (CG 68) (cf. 6.53; 13.18, 21f). Though he didn't elaborate on this future science in detail, he predicted it would establish ‘laws’, ‘rules’, and ‘constant principles’ (CG 16f, 88).

2.9 To explain why ‘semiology’ had ‘not been recognized as an independent science with its own object’, Saussure contends that ‘heretofore language has almost always been studied in connection with something else, from other viewpoints’ (CG 16) (cf. 6.5ff; 9.2). He now announces, in a much-quoted aphorism at the close of the book, that ‘the true and unique object of linguistics is language studied in and for itself’ (CG 232) (cf. 6.64; 13.35). Against Dwight Whitney, he demurs that ‘language is not similar in all respects to other social institutions’ (CG 10). Also, ‘other sciences work with objects that are given in advance’, whereas in ‘linguistics’, ‘it would seem that it is the viewpoint that creates the object’ (CG 8) (cf. 13.58).

2.10 In Saussure's estimate, ‘all idioms embody certain fixed principles that the linguist meets again and again in passing from one to another’ (CG 99). Hence, ‘the linguist is obliged to acquaint himself with the greatest possible number of languages in order to determine what is universal in them by observing and comparing them’ (CG 23) (cf. 6.57; 13.18, 49, 12⁴). ‘But it is very difficult to command scientifically such different languages’, and ‘each idiom is a closed system’, so ‘each language in practice forms a unit of study’ (CG 99). In this connection, Saussure concedes that ‘the ideal, theoretical form of a science is not always the one imposed upon it by the exigencies of practice; in linguistics, these exigencies are more imperious than anywhere else; they account to some extent for the confusion that now predominates in linguistic research’ (cf. 2.4).

2.11 ‘Language’ constitutes a ‘linguistic fact’ that, Saussure hopes, can ‘be pictured in its totality’ (CG 112). To do so, ‘we must call in a new type of facts to illuminate the special nature of language’; and must ‘throw new light on the facts’, whether ‘static’ or ‘evolutionary’ (CG 16f, 189f) (cf. 2.6, 36). For instance, ‘concepts’ are ‘mental facts’; ‘analogy’ is ‘a universal fact’; ‘a phonological system’ is a ‘set of facts’; and so on (CG 11, 176, 171, 34).

2.12 But dealing with ‘facts’ may be quite problematic, since ‘nothing tells us in advance that one way of considering the fact in question takes precedence over any other’ (CG 8). We may have to ‘sift the facts’ ‘many times to bring to light’ the essentials (cf. CG 202). ‘The most serious mistake in method’ is to suppose that ‘the facts embraced’ by a ‘law’ ‘exist once and for all instead of being born and dying within a span of time’ (CG 146). Even where the ‘facts’ may suggest otherwise, ‘we must defend our principle: there are no unchangeable characteristics’ (CG 230f). ‘Permanence results from sheer luck’.

2.13 The range or extension of a fact is also a problem. On the one hand, ‘it is a serious mistake to consider dissimilar facts as a single phenomenon’ (CG 146). Against the neo-grammarians and philologists, who tried to show how ‘a set of facts apparently obeys the same law’, Saussure argues that such ‘facts’ are ‘isolated’ and ‘accidental’; and that ‘regardless of the number of instances where a phonetic law holds, all the facts embraced by it are but multiple manifestations of a single particular fact’ (CG 93f).⁶ He suggests that ‘the term “law”‘ might ‘be used in language as in the physical and natural sciences’, but only from a timeless ‘panchronic viewpoint’ he opposes (CG 95) (cf. 13.11). All the same, he refers to ‘laws that govern the combining of phonemes’, the ‘evolution’ of a ‘word’, the ‘accentuation’ of ‘syllables’, or the status of ‘initial consonants’ and ‘vowels’ (CG 51, 86, 30f).

2.14 Evidently, Saussure couldn't quite decide whether ‘the facts of language’ are ‘governed by laws’ (CG 91) (cf. 12.22). ‘The laws of language’ differ from ‘every social law’, which is ‘imperative’ (‘comes in by force’) and ‘general’ (‘covers all cases’). ‘Like everything that pertains to the linguistic system’, a ‘law’ ‘is an arrangement of terms, a fortuitous, involuntary result of evolution’ (CG 86). ‘And the arrangement that the law defines is precarious precisely because it is not imperative’ (CG 92). Moreover, ‘laws’ such as those governing ‘alternation’ may be ‘only a fortuitous result of underlying’ ‘facts’ (CG 159). In sum, ‘speaking of linguistic law in general is like trying to pin down a [phantom]’ (CG 91).

2.15 Saussure's deliberations already raise the persistent problem in modern linguistics of how to decide what is ‘real’ (1.12f; 13.24f, 57). At times he seems confident: ‘when we examine “abstractions” more closely, we see what part of reality they actually stand for, and a simple corrective measure suffices to give an exact and justifiable meaning to the expedients of the grammarian’ (CG 184) (cf. 13.57). He chides other schools for ‘notions’ with ‘no basis in reality’, though he himself is forced on occasion (e.g., when considering ‘geographical diversity’, which disrupts his conception of the closed system) into a ‘schematic simplification’ that ‘seems to go against reality’ (CG 4, 196).

2.16 At any rate, ‘the concrete entities of language are not directly accessible’ (CG 110). So he would justify the thesis that ‘language is concrete’ with the mentalistic premise that ‘associations which bear the stamp of collective approval’ ‘are realities that have their seat in the brain’ (CG 15) (cf. 2.31, 66, 83; 13.10). ‘The concrete object of linguistic science is the social product deposited in the brain of each individual’ (CG 23). When ‘sound and thought combine’, they ‘produce a form, not a substance’; ‘all the mistakes in our terminology, all our incorrect ways of naming things that pertain to language, stem from the involuntary supposition that the linguistic phenomenon must have substance’ (CG 113, 122) (cf. 6.28-31; 12.89, 11¹⁴).

2.17 Still, ‘to base the classifications’ ‘for arranging all the facts’ ‘on anything except concrete entities’ ‘is to forget that there are no linguistic facts apart from the phonic substance cut into significant elements’ (CG 110) (cf. 3.18; 13.26). Hence, in order to show that ‘abstract entities are always based, in the last analysis, on concrete entities’, Saussure invokes the ‘series of material elements’ (CG 138). ‘Thought’ follows ‘the material state of signs’ (CG 228) (3.10; cf. 13.84). ‘Syntax’ resides inside ‘material units distributed in space’; and ‘words’ are situated in ‘the substance that constitutes sentences’ (CG 139, 172) (cf. 13.33). Despite such jarring passages, Saussure emphasizes that ‘language exists independently’ of ‘the material substance of words’; that ‘the word-unit’ is ‘constituted’ ‘by characteristics other than its material quality’; and that ‘a material sign is not necessary for the expression of an idea’ (CG 18, 94, 86). ‘A material unit exists only through its meaning and function’, just as these two require ‘the support of some material form’ (CG 139).

2.18 Considerations like these made Saussure uneasy about ‘calling the word a concrete linguistic object’ (CG 8) (cf. 3.31; 4.54, 60; 5.53; 6.23; 7.70, 7¹⁹ 8.54; 12.69, 71, 77; 13.29). ‘There has been much disagreement about the nature of the word’; ‘the usual meaning of the term is incompatible with the notion of a concrete unit’ (CG 105).⁷ Nevertheless, ‘being unable to seize the concrete entities or units of language directly, we shall work with words’ (CG 113). Insofar as ‘the word’ ‘at least bears a rough resemblance’ to ‘the linguistic unit’ and ‘has the advantage of being concrete’, ‘we shall use words as specimens equivalent to real terms in a synchronic system, and the principles that we evolve with respect to words will be valid for entities in general’ (CG 113f). After all, ‘the word is a unit that strikes the mind, something central in the mechanism of language’; so ‘everything said about words applies to any term of language’ (CG 111, 116) (cf. 13.54).

2.19 Saussure is determined to view ‘language’ as ‘a well-defined object in the heterogeneous mass of speech facts’ (CG 14). ‘We must put both feet on the ground of language and use language as the norm of all other manifestations of speech’ (CG 9, i.r.) (cf. 2.8). But to do so, he drastically limits the object of study: ‘the science of language is possible only if’ ‘the other elements of speech’ ‘are excluded’ (CG 15) (cf. 2.7, 9). He draws a firm dichotomy between ‘language [langue]’ and ‘human speech [langage]’, making the former ‘only a definite part’ of the latter and oddly arguing that ‘language’ ‘can be classified among human phenomena, whereas speech cannot’ (CG 9, 15). ‘We cannot put’ ‘speech’ ‘in any category of human facts, for we cannot discover its unity’; only ‘language gives unity to speech’ (CG 9, 11) (13.39). ‘Speech cannot be studied, for it is not homogeneous’ (CG 19). Nonetheless, we are counselled to ‘set up the science of language within the overall study of speech’, and told that ‘the subject matter of linguistics comprises all manifestations of human speech’ (CG 17 6).

2.20 To further limit ‘the rational form linguistic study should take’, Saussure makes a dichotomy between ‘language’ [langue] and ‘speaking’ [parole] (CG 98, 13, i.a.) (cf. 13.36). ‘The two objects are closely connected’ and ‘interdependent’, yet are ‘two absolutely distinct things’ (CG 18f). ‘Speaking is necessary for the establishment of language, and historically, its actuality always comes first’ (CG 18). But ‘language’ is ‘passive’, ‘receptive’, ‘collective’, and ‘homogeneous’, while ‘speaking’ is ‘active’, ‘executive’, ‘individual’, and ‘heterogeneous’ (CG 13, 15). Unlike ‘language’, ‘speaking is not a collective instrument; its manifestations are individual and momentary’, and ‘depend on the will of speakers’ (CG 19). Saussure vowed to ‘deal only with linguistics of language’; even if he ‘uses material belonging to speaking to illustrate a point’, he ‘tries never to erase the boundaries that separate the two domains’ (CG 19f). Though according to his editors, he ‘promised to the students’ a ‘linguistics of speaking’ he did not live to present, the Cours indicates that such a ‘science’ wouldn't belong to ‘linguistics proper’; ‘the activity of the speaker should be studied in a number of disciplines which have no place in linguistics except through their relation to language’ (CG xix, 20, 18) (cf. 2.7; 9.6).

2.21 In yet another trend-setting dichotomy, Saussure claimed that ‘language and writing are two distinct systems of signs; the second exists for the sole purpose of representing the first’ (CG 23). ‘The linguistic object is not both the written and the spoken forms of words; the spoken forms alone constitute the object’ (CG 23f) (cf. 4.37-44; 6.50; 8.72ff; 9.42f; 12.83; 13.33). Like Bloomfield, he seems indignant about ‘the tyranny’ whereby ‘writing’ ‘usurps the main role’ (CG 31, 24). ‘Grammarians’ are chided for ‘drawing attention to the written form’, ‘sanctioning the abuse’ with ‘free use’ of ‘pronunciation’, and ‘reversing the real, legitimate relationship between writing and language’ (CG 30) (cf. 9.42f). For Saussure, ‘writing obscures language; it is not a guise for language but a disguise [or travesty]’. At least, writing seems to confuse him: he calls it ‘stable’, then ‘unstable’; he rejects ‘the notion that an idiom changes more rapidly when writing does not exist’ yet grants that ‘spelling always lags behind pronunciation’ (CG 27, 29, 24, 28).

2.22 ‘Spelling’ annoys him particularly, being replete with ‘inconsistencies’, ‘aberrations’, ‘irrational’ or ‘illegitimate’ forms, and ‘absurdities’ that ‘cannot be excused’ (CG 28f) (cf. 4.38; 8.73ff). ‘By imposing itself upon the masses, spelling influences and modifies language’ (CG 31). ‘Visual images lead to wrong pronunciations’, ‘pathological’ ‘mistakes’, ‘monstrosities’, and ‘deformations’ -- fit for ‘teratological’ inquiry (i.e. ‘“the study of monsters”’) (CG 31f, 22). This irritation may have been fuelled by his native French -- in contrast, say, to the ‘ingenious’ and ‘remarkable analysis’ displayed by ‘the Greek alphabet’, ‘realizing almost completely’ ‘a one-to-one ratio between sounds and graphs’ (CG 53, 39). Also, he saw things with the eyes of a phonetician and a historian: ‘the pronunciation of a word is determined, not by its spelling, but by its history’, whereas ‘spelling’ does not follow ‘etymology’ (CG 31, 28).⁸ Nonetheless, he makes no strong case for ‘spelling reform’ and ‘hopes only that the most flagrant absurdities’ ‘will be eliminated’ (CG 34; cf. 2.69; 8.74).

2.23 Ultimately, he relents about writing: since ‘the linguist’ ‘is often unable to observe speech directly, he must consider written texts’ and ‘pass’ through ‘the written form’ ‘to reach language’ (CG 6, 34) (cf. 4.43f; 12.82). ‘The prop provided by writing, though deceptive, is still preferable’ (CG 32). So, ‘far from discarding the distinctions sanctioned by spelling’, Saussure ‘carefully preserves them’, e.g., because ‘the opposition between implosives and explosives is crystal clear in writing’ (CG 53, 62) (cf. 2.72).

2.24 A kindred reservation is raised against ‘literary language’, being here ‘any kind of cultivated language, official or otherwise, that serves the whole community’ (CG 195) (cf. 4.41; 6.4; 12⁴). Though this reservation is maintained more consistently than that against writing, the motives offered for it -- aside from the hardly contestable provision that ‘the linguist must consider not only correct speech and flowery language, but all other forms of expression’ (CG 6) -- are rather obscure and contradictory. For example: ‘the privileged dialect, once it has been promoted to the rank of official or standard, seldom remains the same’; yet ‘literary language, once it has been formed, generally remains fairly stable’, and ‘its dependency on writing gives it a special guarantee of preservation’ (CG 195, 140). Or: ‘literary language’ ‘breaks away from’ ‘spoken language’ and ‘adds to the undeserved importance of writing’, yet does not ‘necessarily imply the use of writing’ (CG 21, 25, 196). Or again: ‘when a natural idiom is influenced by literary language’, ‘linguistic unity may be destroyed’; yet ‘given free reign, a language has only dialects’ and ‘habitually splinters’ (CG 195). Whatever his motives, Saussure did set a countertrend to traditional grammar by marginalizing literary examples (cf. 3.4; 4.41; 6.4).⁹

2.25 Saussure proposed to ‘localize’ his restricted notion of ‘language’ ‘in the limited segment of the speaking-circuit where an auditory image becomes associated with a concept’ (CG 14). ‘Language’ is ‘organized thought coupled with sound’; and ‘each linguistic term is a member, an “articulus” in which an idea is fixed in a sound and a sound becomes the sign of an idea’ (CG 111, 113) (cf. 3.11, 32, 35; 6.30f; 12.17, 22, 47, 61). This viewpoint led to the famous thesis that the ‘sign’ ‘results from associating’ a ‘signified’ with a ‘signifier’ (CG 67) (cf. 8.20; 11.85; 12.11, 47).¹⁰ ‘The linguistic entity exists only through’ this ‘associating’; ‘whenever only one element is retained, the entity vanishes’ (CG 101f).

2.26 ‘Language’ is thus a ‘self-contained whole and a principle of classification’ by virtue of being ‘a system of distinct signs corresponding to’ or ‘expressing’ ‘distinct ideas’ (CG 9f, 16) (cf. 3.40; 12.58). ‘As in any semiological system, whatever distinguishes one sign from the others constitutes it’ (CG 121). Therefore, ‘language is characterized as a system based entirely on the opposition of its concrete units’ or ‘on the mental opposition of auditory impressions’ (CG 107, 33). ‘The general fact’ is ‘the functioning of linguistic oppositions’ (CG 122). Saussure's most extreme formulation is also the most frequently quoted: ‘in language there are only differences without positive terms’ (CG 120). ‘Language’ is ‘organized thought coupled with sound’; and ‘each linguistic term is a member, an “articulus” in which an idea is fixed in a sound and a sound becomes the sign of an idea’ (CG 111, 113) His crucial reservation, however, is seldom quoted and reinvokes the dual nature of the sign: ‘the statement that everything in language is negative is true only if the signified and the signifier are considered separately’ -- whereby, we just saw, ‘the entity vanishes’. ‘The sign in its totality’ of two entities ‘is positive in its own class’. ‘Their combination is a positive fact; it is even the sole type of facts that language has, for maintaining the parallelism between two classes of differences is the distinctive function of the linguistic institution’ (CG 120f) (cf. 4.26).

2.27 Each of the ‘two elements’, ‘the idea and the sound’, ‘functions’ in ways which ‘prove that language is only a system of pure values’ (CG 111). To some extent, the two sides control each other. ‘The source material of language’ is ‘pictured’ as ‘two parallel chains, one of concepts and the other of sound-images’ (CG 104) (cf. 6.41; 9.3; 12.43, 69). ‘In an accurate delimitation, the division’ of the two ‘chains’ ‘will correspond’. Moreover, in ‘countless instances’, ‘the alteration of the signifier occasions a conceptual change’, and ‘it is obvious that the sum of ideas distinguished corresponds in principle to the sum of the distinctive signs’ (CG 121) (but cf. 2.29; 5.64, 67, 75ff; 7.82; 11.36; 12.93; 13.59). ‘Any nascent difference will tend invariably to become significant’; reciprocally, ‘any conceptual difference perceived by the mind seeks to find expression through a distinct signifier, and two ideas that are no longer distinct in the mind tend to merge into the same signifier’. ‘Thought’ may be ‘forced’ ‘into the special way that the material state of signs opens to it’ (CG 228).

2.28 But the two sides do not control each other to the extent that ‘the bond between the signifier and the signified is arbitrary’ (CG 67) (cf. 3.3; 4.27; 9.13, 32; 11.86).¹¹ If the ‘sign’ ‘results from’ that bond, Saussure ‘can simply say: the linguistic sign is arbitrary’, i.e., ‘it is unmotivated’, as shown by words for the same thing (“tree”) in different languages (CG 66f, 69) (cf. 4.27; 9¹⁷). This ‘principle’ ‘dominates all the linguistics of language’ (CG 68). Among its ‘numberless’ ‘consequences’, I mention three I think essential to Saussurian argument. First, ‘the arbitrary nature of the sign explains’ ‘why the social fact alone can create a system’ (CG 113). Second, `arbitrary and differential are two correlative qualities’: ‘a segment of language can never in the final analysis be based on anything but its noncoincidence with the rest’ (CG 118). Third, ‘in linguistics to explain a word is to relate it to other words, for there are no necessary relations between sound and meaning’ (CG 189). If ‘the choice of a given slice of sound to name a given idea’ were not ‘completely arbitrary’, ‘the notion of value would be compromised, for it would include an externally imposed element’ (CG 113).

2.29 Although ‘no one disputes the principle of the arbitrary nature of the sign’, and ‘wholly arbitrary’ ‘signs’ ‘realize better than others the ideal of the semiological process’ (CG 68), Saussure betrays some uneasiness. At one point he calls ‘arbitrariness’ an ‘irrational principle’ ‘which would lead to the worst sort of complication if applied without restriction’ (CG 133). He is accordingly ‘convinced’: ‘everything that relates to language as a system’ serves ‘the limiting of arbitrariness’ (cf. 2.56). The linguist must ‘study’ ‘language’ ‘as it limits arbitrariness’. Various ‘degrees’ may range ‘between the two extremes -- a minimum of organization and a minimum of arbitrariness’ (CG 131, 133). Such ‘proportions’ might ‘help in classifying’ ‘diverse languages’; those ‘in which there is least motivation are more lexicological, and those in which it is greatest are more grammatical’ (CG 133; cf. CG 161) (cf. 13.59). Or, ‘within a given language’, we might consider how ‘all evolutionary movement may be characterized by continual passage from motivation to arbitrariness’ and vice versa (CG 134). Or again, we might examine how ‘motivation varies, being always proportional to the ease of syntagmatic analysis and the obviousness of the meaning of the subunits present’ (CG 132). ‘At any rate, even in the most favourable cases motivation is never absolute; not only are the elements of a motivated sign themselves unmotivated’, ‘but the value of the whole term is never equal to the sum of the value of its parts’ (cf. 2.27; 5.29, 67; 12.93; 13.59).

2.30 Alongside ‘motivated’, ‘natural’ is treated as a converse of ‘arbitrary’ (CG 69), and here too, Saussure is not fully consistent. He vows that ‘natural data have no place in linguistics’ (CG 80). Similarly, ‘the traditional divisions of grammar’ ‘do not correspond to natural distinctions’ (CG 136) (cf. 3.23; 4.71). And ‘the false notion’ of ‘language’ as ‘a natural kingdom’ leads to ‘absurdities’ (CG 4).¹² Even if ‘semiology’ ‘welcomes’ the ‘natural sign, such as pantomime’, ‘its main concern will still be the whole group of systems grounded on the arbitrariness of the sign’ (CG 132).¹³ Despite this, Saussure invokes ‘natural dialectal features’, ‘the natural fact’ of ‘geographical diversity’, the ‘two natural coordinates’ of ‘associative’ and ‘syntagmatic’, and the ‘natural organic growth of an idiom’ (CG 201, 196, 203, 137, 21). In his view, by ‘giving language first place among the facts of speech, we introduce a natural order’ (CG 9). With a comparable inconsistency, he condemns the designation of language as an ‘organism’ or an ‘organic’ entity, but frequently applies these terms himself (CG 5, 231, 21f, 69, 153, 193).

2.31 The division of the sign into signified and signifier is not the same as the division of the ‘speaking-circuit’ into the ‘psychological parts (word-images and concepts)’ and the ‘physiological (phonation and audition)’ (CG 12). ‘Speaking’ involves the ‘physiological’, whereas ‘language’ ‘is exclusively psychological’ (CG 18; cf. CG 8, 12ff; 13.14). So ‘both terms involved in the linguistic sign’, the signified and the signifier, ‘are psychological and are united in the brain by an associative bond’ (CG 65f) (2.16). Even the ‘material and mechanical manifestations’ are ‘psychological’; ‘the psychophysical mechanism’ is significant only for ‘exteriorizing’ the ‘combinations’ that ‘express’ ‘thought’ (CG 6, 14). The ‘sound-image’ ‘is not the material sound, but the psychological imprint of the sound, the impression that it makes on our senses’ (CG 66). Still, we have seen Saussure invoking the ‘material’ aspect to suggest the ‘concreteness’ of language (cf. 2.17, 27).¹⁴

2.32 In exuberant moments, Saussure pictures language as a fortunate development for the human mind, agreeing this time with traditional ‘philosophers and linguists’: ‘without language, thought is a vague, uncharted nebula’ (CG 112) (cf. 3.3; 6.2, 31; 12.17). ‘There are no pre-existing ideas, and nothing is distinct before the appearance of language’. ‘Psychologically, our thought -- apart from its expression in words -- is only a shapeless and indistinct mass’, a ‘floating realm’ (CG 111f). ‘Thought, chaotic by nature, has to become ordered in the process of its decomposition; language takes shape between two shapeless masses’, namely, ‘the indefinite plane of jumbled ideas and the equally vague plane of sounds’. ‘Without the help of signs we would be unable to make a clear-cut consistent distinction between two ideas’. At such moments, Saussure downplays the influence of ‘arbitrariness’, against which ‘the mind contrives to introduce a principle of order and regularity into certain parts of the mass of signs’ (CG 133). ‘The mechanism of language is but a partial correction of a system that is by nature chaotic’. He even says, contravening his own conception of system, that ‘language’ ‘is a confused mass, and only attentiveness and [de]familiarization will reveal its particular elements’ (CG 104) (cf. 13.42).

2.33 Signalling a mentalist orientation linguistics would later reject (4.8, 13.4f, 10f), Saussure invokes ‘the sum of word-images stored in the minds of all individuals’, where ‘forms’ are ‘associated’ ‘through their meanings’, as the basis for ‘the social bond that constitutes language’ (CG 13, 165). In this sense, ‘linguistics has only the perspective of speakers’ (CG 212).¹⁵ But he concedes that ‘we never know exactly whether or not the awareness of speakers goes as far as the analyses of grammarians’ (CG 138). ‘Doubtless speakers are unaware of the practical difficulties of delimiting units’; ‘in the matter of language, people have always been satisfied with ill-defined units’ (CG 106, 111) (cf. 13.7, 59).

2.34 All the same, Saussure set yet another trend for linguistics by typically implying that the categories and notions he proposes are shared by the minds of ‘speakers’ (e.g. CG 138, 160, 185, 192) (cf. 13.49). He depicts the ‘objective analysis based on history’ and done by ‘the grammarian’ as ‘but a modified form’ of the ‘subjective analysis’ ‘speakers constantly make’ (CG 183) (cf. 13.58). ‘Both analyses are justifiable, and each retains its value’, even if he can find ‘no common yardstick for both the analysis of speakers and the analysis of the historian’. ‘In the last resort, however, only the speakers’ analysis matters, for it is based directly upon the facts of language’. Fair enough, but he stressed that the ‘facts’ can be elusive, even for experts (2.12ff).

2.35 A compromise would be to assign the knowledge of speakers to a level of which they are not ‘conscious’.¹⁶ In accounting for ‘analogy’, for example, ‘no complicated operation such as the grammarian's conscious analysis is presumed on the part of the speaker’; ‘the sum of the conscious and methodological classifications made by the grammarian’ ‘must coincide with the associations, conscious or not, that are set up in speaking’ (CG 167, 137f). Or, if ‘language is not complete in any speaker’ and ‘exists perfectly only within a collectivity’, we might assign the knowledge to ‘the collective mind of speakers’, wherein ‘logical and psychological relations’ ‘form a system’ (CG 14, 99f). This designation would be appropriate for ‘synchronic linguistics’, whereas ‘diachronic linguistics’ would ‘study relations that bind together successive terms not perceived by the collective mind but substituted for each other without forming a system’ (CG 99f, i.r.).

2.36 The quest for the locus of language thus leads to Saussure's ‘radical distinction between diachrony and synchrony’ (CG 184). Though ‘very few linguists suspect’ it, ‘the intervention of the factor of time creates difficulties peculiar to linguistics and opens to their science two completely divergent paths’ (CG 79). Saussure now calls for ‘two sciences of language’, one ‘static’ or ‘synchronic’, and the other ‘evolutionary’ or ‘diachronic’ (CG 81). Because ‘static linguistics’ was not yet established and seemed ‘generally much more difficult’ (CG 101), Saussure favoured it in his own theorizing. For him, ‘language is a system whose parts can and must be considered in their synchronic solidarity’ (CG 87). ‘Language is a system of pure values determined by nothing except the momentary arrangement of terms’; and ‘a system of interdependent terms in which the value of each term results solely from the simultaneous presence of others’ (CG 80, 114). Only ‘synchronic facts’ ‘affect the system as a whole’ and are therefore ‘always significant’ (CG 85, 87; cf. CG 95). ‘In analysis, then, we can set up a method and formulate definitions only after adopting a synchronic viewpoint’ (CG 185). Besides, ‘the synchronic viewpoint’ ‘is the true and only reality to the community of speakers’ (CG 90, 212).

2.37 In contrast, ‘the diachronic perspective deals with phenomena that are unrelated to systems’, and with ‘partial facts’ (CG 85, 87). ‘The diachronic phenomenon’ is ‘the evolution of the system’ through ‘a shift in the relationship between the signifier and the signified’ (CG 181) (cf. 2.48). ‘In a diachronic succession, the elements are not delimited for once and for all’; they ‘are distributed differently from one moment to the next’ (CG 179). Hence, ‘the units delimited in diachrony would not necessarily correspond to those delimited in synchrony’ (CG 181). Moreover, ‘the synchronic fact’ ‘calls forth two simultaneous terms’, whereas ‘the diachronic fact’ ‘involves’ ‘only one term’: ‘for the new one to appear’, ‘the old one’ ‘must first give way to it’ (CG 85). These theses complicate ‘the problem of the diachronic unit’ and ‘the essence’ of ‘evolution’ (CG 181). ‘An element taken from one period’ qualifies as ‘the same’ as ‘an element taken from another period’ only if ‘regular sound changes’ intervene and if the ‘speaker passes from one form to the other without there being a break in their common bond’ (CG 181f) (cf. 2.73).

2.38 For Saussure, ‘the opposition between the two viewpoints, the synchronic and the diachronic, is absolute and allows no compromise’ (CG 83). ‘The more rigidly they are kept apart, the better it will be’, and their respective ‘“phenomena”‘ ‘have nothing in common’ (CG 22, 91). ‘The synchronic law is general but not imperative’ and merely ‘reports a state of affairs’ (CG 92). In ‘diachrony, on the contrary’, we find ‘imperativeness’ that ‘is not sufficient to warrant applying the concept of law to evolutionary facts’, which, ‘in spite of certain appearances’, are ‘always accidental and particular’ (CG 93; cf. 2.47, 55; 4.75). Hence, a synchronic approach fits better the standard notion of how science works.

2.39 Elsewhere, however, he concedes that ‘the system and its history’ ‘are so closely related that we can scarcely keep them apart’ (CG 8). ‘Synchronic truth is so similar to diachronic truth that people confuse the two or think it superfluous to separate them’ (CG 96). ‘In fact, linguistics has confused them for decades without realizing that such a method is worthless’ (CG 97). The ‘force of circumstances’ is blamed for ‘inducing’ us to ‘consider’ ‘each language’ ‘alternately from the historical and static viewpoints’ (CG 99). Yet Saussure insists it is ‘absolutely impossible to study simultaneously relations in time and relations within the system’ (CG 81). ‘We must put each fact in its own class and not confuse the two methods’..

2.40 He accordingly finds it ‘obvious that the diachronic facts are not related to the static facts they produced’ (CG 83). ‘A diachronic fact is an independent event; the synchronic consequences that stem from it are wholly unrelated to it’ (CG 84). Hence, ‘the linguist who wishes to understand a state must discard all knowledge of everything that produced it’ (CG 81). ‘He can enter the mind of speakers only by completely suppressing the past; the intervention of history can only falsify his judgment’ (CG 81; cf. CG 160). Admittedly, ‘the forces that have shaped the state illuminate its true nature, and knowing them protects us against certain illusions’; ‘but this only goes to prove clearly that diachronic linguistics is not an end in itself’ (CG 90).

2.41 This line of argument implies we can assign events or causes to a separate science or domain from their results or effects (cf. 2.74). Apparently, the key factor is that ‘diachronic facts are not’ ‘directed toward changing the system’; ‘only certain elements are altered without regard to the solidarity that binds them to the whole’ (CG 84). Even when ‘a shift in a system is brought about’ or a ‘change was enough to give rise to another system’, the ‘events’ responsible are ‘outside the system’ ‘and form no system among themselves’ (CG 95, 85). ‘In the science of language, all we need do is to observe the transformations of sounds and to calculate their effects’; ‘determining the causes’ is not ‘essential’ (CG 18).

2.42 A further problem is how to gather data ‘outside the system’. If ‘the linguist’ ‘takes the diachronic perspective, he no longer observes language, but rather a series of events that modify it’ (CG 90). ‘The causes of continuity are a priori within the scope of the observer, but the causes of change in time are not’ (CG 77). Still, ‘evolutionary facts are more concrete and striking’ than ‘static’ ones; the ‘observable relations tie together successive terms that are easily grasped’ (CG 101). We are confronted with ‘observable modifications’; ‘innovations’ ‘enter into our field of observation’ when ‘the community of speakers has adopted them’ (CG 88, 98). This ‘community’ cannot however be the only point of reference, since the ‘succession in time’ of ‘the facts of language’ ‘does not exist insofar as the speaker is concerned’ (CG 81). ‘The linguist’ needs ‘external evidence’, such as ‘contemporary descriptions’ (which ‘lack scientific precision’), ‘spelling’, ‘poetic texts’, ‘loanwords’, ‘puns’, and ‘stories’ (CG 35f, i.r.).

2.43 Similarly agile manoeuvring is performed to deal with ‘geographical’ ‘diversity’, which gets relegated to ‘external linguistics’, presumably for not applying to closed or uniform systems: ‘it is impossible, even in our hypothetical examples, to set up boundaries between the dialects’ (CG 191, 204). 'Dialectal differences’ appear when an ‘innovation’ ‘affects only a part of the territory’ (CG 200). In another about-face, Saussure says ‘divergences in time escape the observer, but divergences in space immediately force themselves upon him’ (CG 191). ‘Geographical diversity was, then, the first observation made in linguistics and determined the initial form of scientific research in language’. But to preserve the ‘profound unity’ postulated in his synchronic approach and ‘hidden’ by ‘the diversity of idioms’, Saussure maintains that ‘time’ ‘is actually the basic cause of linguistic differentiation’: ‘by itself, space cannot influence language’ (CG 99, 198). If ‘change itself’ and ‘the instability of language stem from time alone’, ‘geographical diversity should be called temporal diversity’ (CG 198f, i.r.).¹⁷ Here, effects get fully referred back to the events that caused them -- just the reverse of the argument for keeping diachrony separated (2.40f).

2.44 Saussure implies that language changes all by itself: ‘language is not controlled directly by the mind of speakers’, and the ‘sign’ and ‘language’ ‘always elude the individual or social will’ (only ‘speaking’ is ‘wilful’) (CG, 228 19, 17, 14). ‘Speakers do not wish’ or ‘try to change systems’, but ‘pass from one to the other, so to speak, without having a hand in it’ (CG 84ff). But it's hard to see how language can change at all if ‘the signifier’ ‘is fixed, not free, with respect to the linguistic community’ (CG 71). ‘The masses have no voice in the matter, and the signifier chosen by language could be replaced with no other’. ‘No individual, even if he willed it, could modify in any way at all the choice’; and ‘the community itself cannot control so much as a single word’. ‘We can conceive of a change only through the intervention of specialists, grammarians, logicians, etc.; but experience shows us that all such meddlings have failed’ (CG 73) (cf. 4²⁷).

2.45 Notwithstanding, ordinary speakers do change language. ‘An evolutionary fact is always preceded’ by ‘a multitude of similar facts in the sphere of speaking’ (CG 98). ‘Nothing enters language without having been tested in speaking, and every evolutionary phenomenon has its roots in the individual’ (CG 168) (cf. 3.57; 4.81, 4⁶). ‘One speaker had to coin the new word, then others had to imitate and repeat it until it forced itself into standard usage’. So ‘in the history of any innovation there are always two distinct moments: (1) when it sprang up in individual usage; and (2) when it became a fact of language, outwardly identical but adopted by the community’ (CG 165). The ‘distinction made’ between ‘diachrony’ and ‘synchrony’ is ‘in no way invalidated’ by this process. But we would need, it seems to me, a reliable way, short of interviewing the entire community, for telling just when an innovation passes from one ‘moment’ to the other, and for (here again) separating causes from effects (cf. 4.77f).

2.46 His idea that ‘evolution in time takes the form of successive and precise innovations’ (CG 200) must have made Saussure uneasy about inexact improvisations of speakers. When he deals with ‘folk etymology’, he does just what he scolds ‘grammarians’ for: ‘thinking that spontaneous analyses of language are “wrong”‘ (CG 183) (and cf. 2.49). ‘This phenomenon called folk etymology’ ‘works somewhat haphazardly and results only in absurdities’, ‘mistakes’, and ‘deformations’ (CG 173ff). During ‘crude attempts to explain refractory words’, the words get ‘misunderstood’, ‘corrupted’, and ‘mangled’. The grounds for castigating folk etymology can only be that it is not sanctioned by historical knowledge -- which, Saussure grants, ordinary speakers do not have (cf. CG 81, 90, 100, 160, 212; 2.35, 40, 64).

2.47 One of the rare occasions¹⁸ when his reverent editors reassure us that Saussure is not ‘being illogical or paradoxical’ is when he ‘speaks of both the immutability and the mutability of the sign’ -- and assigns both phenomena the same cause, namely, ‘the arbitrary nature of the sign’ (CG 73ff). On the one hand, ‘language always appears as a heritage of the preceding period’; ‘no society’ ‘has ever known language other than as a product inherited from preceding generations’ (CG 71). ‘The arbitrary nature of the sign’ ‘protects language from any attempt to modify it’; ‘the sign’ ‘follows no law other than that of tradition’ (CG 73f). Also, ‘society’, ‘inert by nature, is a prime conservative force’. ‘Generations’ ‘fuse and interpenetrate’; ‘speakers are largely unconscious of the laws of language’; and ‘even their awareness would seldom lead to criticism, for people are generally satisfied with the language they have received’ (CG 72). Such arguments are thought to show that ‘the question of the origin of speech’ ‘is not even worth asking’; ‘the only real object of linguistics is the normal, regular life of an existing idiom’ (CG 71f) (cf. 3.11, 18; 8.6, 17).

2.48 On the other hand, ‘time changes all things; there is no reason why language should escape this universal law’ (CG 77). ‘The arbitrary nature of the sign’ is now deployed to explain why ‘language is radically powerless against the forces which from one moment to the next are shifting the relationship between the signified and the signifier’ (CG 75) (cf. 2.37). Those forces ‘loosen’ ‘the bond between the idea and the sign’. A further paradox arises: ‘the sign is exposed to alteration because it perpetuates itself’ (CG 74).

2.49 Perhaps because such shifts disrupt Saussure's search for order, he tends to misprize language change as a matter of ‘deteriorations’, ‘vicissitudes’, ‘damage’, ‘disturbance’, ‘breaking’, and ‘effacement’, ‘in spite of’ which ‘language continues to function’ (CG 87, 152ff, 161). He envisions ‘a blind force against the organization of a system of signs’, and a ‘great mass of forces that constantly threaten the analysis adopted in a particular language-state’ (CG 89, 169) (cf. 3.13, 16). He admonishes that ‘the total number of forms is uselessly increased’; ‘the linguistic mechanism is obscured and complicated’; and ‘phonetic evolution first obscures analysis, then makes it completely impossible’ (CG 161, 155). Like those brought against folk etymology, such condemnations disrupt the descriptive, non-evaluative methodology whereby Saussure wants to overcome ‘the illusion’ of ‘the first linguists’, for whom ‘everything that deviated from the original state’ was ‘a distortion of an ideal form’ (CG 163) (cf. 4.80).

2.50 ‘Fortunately, analogy counterbalances the effects’ of ‘transformations’ (CG 161). ‘An analogical form is a form made on the model of one or more other forms in accordance with a definite rule’ (i.r.). It can ‘offset’ ‘change’, ‘restore regularity’, and ‘unify’, ‘preserve’, or ‘renew’ ‘forms’ (CG 171f). Thus, ‘analogy always plays an important role’ in the ‘preservation’ or ‘redistribution of linguistic material’ (CG 173). ‘The most obvious and important effect of analogy’ is to ‘substitute more regular forms composed of living elements for older irregular and obsolescent forms’ (CG 171). To the extent that it ‘always uses old material for its innovations’, ‘analogy’ ‘is remarkably conservative’ (CG 172). Since it ‘constantly renews’ ‘forms’, ‘analogy’ is claimed to ‘intervene’ even ‘when forms remain unchanged’. ‘A form may be preserved’ either by ‘complete isolation or complete integration in a system that has kept the basic parts of the word intact and that always comes to the rescue’ (CG 173). If change appears harmful, the system appears beneficial.

2.51 To be sure, ‘analogy’ ‘is capricious’ (CG 162) and alters as well as preserves. It ‘collaborates efficiently with all the forces that constantly modify the architecture of an idiom’ (CG 171). It ‘reflects the changes that have affected the functioning of language and sanctions them through new combinations’. ‘To analogy are due all normal nonphonetic¹⁹ modifications of the external side of words’ (CG 161). However, ‘imperfect analyses sometimes lead to muddled analogical creations’ (CG 171).

2.52 By showing how ‘language never stops interpreting and decomposing its units’, ‘analogy’ is a good illustration of ‘the principle of linguistic creativity’, and a ‘manifestation of the general activity that singles out units for subsequent use’ (CG 169, 165f). ‘Any creation must be preceded by an unconscious comparison of the material deposited in the storehouse of language, where productive forms are arranged according to their’ ‘relations’ (CG 165) (cf. 5.47; 7.76; 8.58). Hence, ‘analogy’ presupposes ‘awareness and understanding of a relation between forms’. This ‘awareness’ leads to ‘the chance product’: ‘the form improvised’ by ‘the isolated speaker’ ‘to express his thought’ (CG 165f). ‘It is wrong to suppose that the production process is at work only when the new formation actually occurs: the elements were already there’ to guarantee the ‘potential existence in language’ of any ‘newly formed word’ (cf. 6.23f). ‘The final step of realizing it in speaking is a small matter in comparison to the build-up of forces that makes it possible’. This time, effects get downplayed in favour of causes in order to give us ‘one more lesson in separating language from speaking’ (CG 165).

2.53 ‘Analogy is therefore proof positive that a formative element exists at a given moment as a significant unit’ (CG 170). ‘Every possibility of effective talk’ has the same source as ‘every possibility of analogical formations’: the way ‘speech is continually engaged in decomposing its units’ (CG 166).²⁰ ‘If living units perceived by speakers at a particular moment can by themselves give birth to analogical formations, every definite redistribution of units also implies a possible expansion of their use’ (CG 170). ‘All such innovations are perfectly regular; they are explained in the same way as those that language has accepted’ (CG 168f). ‘Decomposable’ ‘words can be rated for capacity to engender other words’ (CG 166) (cf. 5.47).²¹

2.54 The question of when two forms are the same now receives a different treatment than it did in the discussion of the ‘diachronic unit’ (2.37). Here, ‘analogical innovation and the elimination of the older forms are two distinct things’ (CG 164). ‘Analogical change’ is an ‘illusion’; ‘nowhere do we come upon a transformation’ of an element. The reasoning behind this claim must be Saussure's belief that ‘analogy is psychological’, ‘grammatical, and synchronic’, rather than ‘phonetic’ (CG 165f, 161). So ‘analogy by itself could not be a force in evolution’, nor be ‘an evolutionary fact’, even though ‘the constant substitution of new forms for old ones is one of the most striking features in the transformation of language’ (CG 169, 171). ‘Enough’ ‘creations of speakers’ endure ‘to change completely the appearance of its vocabulary and grammar’ (CG 169) (2.49).

2.55 This tricky reasoning reflects the intent to place ‘grammar’ mainly on the ‘synchronic’ side; ‘since no system straddles several periods, there is no such thing as historical grammar’ (CG 134) (cf. 2.74).²² Thus, ‘all grammatical laws’ ‘are synchronic’, even though ‘grammatical classes evolve’ (CG 159, 141). It would be ‘radically impossible’ that ‘a phonetic phenomenon would mingle with the synchronic fact’ in ‘grammar’ (CG 152) (cf. 2.51, 54). In return, ‘morphology, syntax, and lexicology interpenetrate because every synchronic fact is identical’ (CG 136). Correspondingly, ‘morphology has no real, autonomous object’; ‘it cannot form a discipline distinct from syntax’ (CG 135) (cf 3.26, 34f; 4.61f, 65; 5.51, 53f; 6.45, 49; 7.75f; 8.57; 9.31, 34, 75, 9⁵, 9¹⁵; 11.35, 40; 12.71, 75, 77; 13.28). Also, ‘the lexical and the syntactic blend’; ‘there is basically no distinction between’ ‘a phrase’ and a ‘word that is not a simple, irreducible unit’ -- ‘the arrangement’ of ‘groups of words in phrases’ ‘follows the same fundamental principles’ as does that of ‘subunits of the word’ (CG 135f) (cf. 3.26, 34f; 4.61; 5.53f; 8.57; 9.75; 11.40; 12.75).

2.56 To capture ‘synchronic facts’, we should recognize that ‘in language everything boils down’ not only ‘to differences, but also to groupings’ (CG 136, 128) (cf. 8.51, 78-82; 9.75-81). To study groupings, yet another major dichotomy is proposed: we should ‘gather together all that makes up a language state and fit this into a theory of syntagms and a theory of associations’ (CG 136). ‘Each fact should’ ‘be fitted into its syntagmatic and associative class’ (CG 137). ‘Only the distinction’ ‘between syntagmatic and associative relations can provide a classification that is not imposed from the outside’ (CG 136). ‘The groupings in both classes are for the most part fixed by language; this set of common relations constitutes language and governs its functioning’ (CG 127). Moreover, ‘syntagmatic’ and ‘associative’ ‘solidarities’ ‘are what limits arbitrariness’ and supplies ‘motivation’: ‘(1) analysis of a given term, hence a syntagmatic relation; and (2) the summoning of one or more other terms, hence an associative relation’ (CG 132f) (cf. 2.29). So ‘the whole subject matter of grammar should be arranged along its two natural coordinates’; and ‘almost any point of grammar will’ ‘prove the necessity of the dual approach’ (CG 137).

2.57 This fresh dichotomy is predictably propounded in mentalistic terms. ‘Our memory holds in reserve all the more or less complex types of syntagms, regardless of their class or length, and we bring in the associative groups to fix our choice when the time for using them arrives’ (CG 130). ‘Every’ ‘unit is chosen after a dual mental opposition’ (CG 131). For instance, ‘the isolated sound’ ‘stands in syntagmatic opposition to its environing sounds and in associative opposition to all other sounds that may come to mind’. Or, the ‘parts’ of ‘syntagms’, such as the ‘subunits’ of ‘words’, can be ‘analysed’ because they can be ‘placed in opposition’ (CG 129). Similarly, ‘from the synchronic viewpoint’, each word ‘stands in opposition to every word that might be associated with it’ (CG 95) (cf. 2.26).

2.58 These assertions fit ‘the only definition’ Saussure ‘can give’ for ‘the unit’ of ‘language’: ‘a slice of sound which to the exclusion of everything that precedes and follows it in the spoken chain is the signifier of a certain concept’ (CG 104, i.r.). In fact, ‘almost all units of language depend on what surrounds them in the spoken chain, or on their successive parts’ (CG 127). The principle is therefore that ‘in the syntagm a term has value because it stands in opposition to everything that precedes or follows it, or to both’ (CG 123) (12.51, 81), but gives no demonstrations for stretches of real discourse. He is merely extending his concept of ‘opposition’ to cover the problem of ‘the unit’ in the sequence, and thereby altering the concept. In the abstract system (e.g., of phonemes or morphemes), elements must be similar in many respects (e.g. class or category) in order to give full value to their opposition. No such principle is required for successive elements in a chain or syntagm; they could differ in all manner of diverse ways that contribute less to their value than does their manner of combination (cf. 12.50f, 56, 70).

2.59 In a Saussurian perspective, the production of discourse could be a process of ‘thinking unconsciously of diverse groups of associations’ and ‘mentally eliminating everything that does not help to bring out the desired differentiation at the desired point’ (CG 130). He envisions language units ‘calling up’ or ‘recalling’ others (CG 130, 134, 164) (cf. 11.69, 87). Such claims are unproblematic if ‘the sum of the conscious and methodical classifications made by the grammarian who studies a language-state without bringing in history must coincide with the associations, conscious or not, that are set up in speaking’ (CG 137f) (cf. 2.33f; 13.49). But the situation is more precarious if ‘it is by a purely arbitrary act that the grammarian groups’ ‘words’ ‘in one way rather than in another’ (CG 127) (cf. 3.13; 4.82; 5.30; 6.15; 9.3; 13.27). For instance, ‘in the mind of speakers the nominative case is by no means the first one in the declension, and the order in which terms are called depends on circumstances’.

2.60 By definition, ‘syntagms’ are ‘combinations supported by linearity’ and ‘always composed of two or more consecutive units’ (CG 123). ‘Syntagmatic groupings mutually condition each other’ (CG 128). Indeed, ‘syntagmatic solidarities’ are ‘what is most striking in the organization of language’ (CG 127, i.r.). Such images as ‘spatial co-ordinations’, ‘two units distributed in space’, or ‘a horizontal ribbon that corresponds to the spoken chain’ (CG 128, 136f) indicate that Saussure was influenced by the appearance of alphabetic written language (cf. 2.5, 17, 21, 23; 13.33). When ‘auditory signifiers’ are ‘represented in writing’, ‘the spatial line of graphic marks is substituted for succession in time’ (CG 70) (cf. 2.72).

2.61 One problem with ‘the syntagm’ is clearly recognized: ‘there is no clear-cut boundary between the language fact, which is the sign of collective usage, and the fact that belongs to speaking and depends on individual freedom’ (CG 125) (but cf. 2.20, 33, 44f). ‘In a great number of instances’, ‘both forces have combined in producing’ ‘a combination of units’ and have done so ‘in indeterminate proportions’. Therefore, ‘not every syntagmatic fact is classed as syntactical’ (and pertaining to the language system), ‘but every syntactical fact belongs to the syntagmatic class’ (CG 137). ‘The sentence is the ideal type of syntagm, but it belongs to speaking, not to language’ (CG 124) (cf. 13.54). Only ‘pat phrases in which any change is prohibited by usage’ ‘belong to language’ (cf. 4.60; 5.32, 54; 7³⁴; 9.93; 13.28).

2.62 Also, ‘to language rather than to speaking belong the syntagmatic types that are built upon regular forms’ (CG 125). One example is ‘word-parts: prefixes, roots, radicals, suffixes, and inflectional endings’ (CG 185). ‘The root [racine] is the irreducible element common to all words of same family’ -- ‘the element in which the meaning common to all related words reaches the highest degree of abstraction and generality’ (CG 186). The ‘radical [radical]’ ‘is generally the common element’ in ‘a series of related words’ and ‘conveys the idea common to every word’ (CG 185). Despite the similar definitions (and common etymology), the ‘radical’ differs from the ‘root’ -- ‘even when phonetically identical to it’ -- in being ‘reducible’, longer, and less ‘general’ and ‘abstract’ (CG 186). ‘The prefix goes before the part of the word that is recognized as the radical’; ‘the suffix is the element added to the root to make a radical’ (CG 187). ‘The prefix also differs from the suffix’ in being ‘more sharply delimited, for it is easier to separate from the word as a whole’ (CG 188). ‘A complete word usually remains after the prefix is removed’, but not after ‘the suffix’ is.

2.63 Word-parts can be a problem in gathering data. Saussure enlists cases where ‘the division’ ‘is self-evident’; where the ‘radical emerges spontaneously when we compare’; or where ‘the speaker knows, before he has made any comparison with other forms, where to draw the line between the prefix and what follows it’ (CG 185, 188).²³ Moreover, ‘the root’ ‘is a reality in the mind of speakers’, though they ‘do not always single it out with equal precision’ (CG 186). ‘In certain idioms’, in ‘German, for instance’, ‘definite characteristics call the root to the attention of speakers’; but ‘the feeling for roots scarcely exists in French’ (CG 186f). Still, ‘structural rules’, ‘regular alternations’, and ‘possible oppositions’ that ‘single out the subunits’ ‘which language recognizes and the values which it attaches to them’ (CG 187ff) might be grasped only by specialists.

2.64 Word-parts can start out as separate ‘elements’ and then get ‘welded’ by ‘agglutination’ ‘into one unit’ ‘which is absolute or hard to analyse’ (CG 169, 176). ‘The mind gives up analysis -- it takes a short-cut’ -- and ‘the whole cluster of signs’ ‘becomes a simple unit’ (CG 177). ‘The phenomenon’ has ‘three phases: (1) the combining of several terms in a syntagm’; (2) ‘the synthesizing of the elements into a new unit’; and ‘(3) every other change necessary to make the old cluster of signs more like a simple word’, e.g., ‘unification of accent’. Saussure finds a ‘striking’ ‘contrast’: ‘agglutination’ ‘blends’ ‘units’ and ‘works only in syntagms’, whereas ‘analogy’ ‘builds’ ‘units’ and ‘calls forth associative series as well as syntagms’ (CG 177f). ‘Agglutination is neither wilful nor active’ and its ‘elements’ are ‘slowly set’; ‘analogy’ ‘requires analyses and combinations, intelligent action, and intention’, and makes ‘arrangements’ ‘in one swoop’. However, Saussure admits, ‘often it is difficult to say whether an analysable form arose through agglutination or as an analogical construction’ -- ‘only history can enlighten us’ (CG 178f), and, we were told, ordinary speakers do not perceive diachronically (cf. 2.35, 40, 46).²⁴

2.65 The counterpart of ‘syntagmatic’ is, as we saw, ‘associative’, a domain that would later be called ‘paradigmatic’ (4.57f; 5.74; 6.34; 8.32; 9.3; 12.71).²⁵ ‘Whereas a syntagm immediately suggests an order of succession and a fixed number of elements, terms in an associative family occur neither in fixed numbers nor in a definite order’ (CG 126). In the latter ‘family’, then, ‘a particular word is like the centre of a constellation’, or ‘the point of convergence of an indefinite number of co-ordinated terms’ that ‘float around’ within ‘one or more associative series’ (CG 126, 129). ‘Large’ ‘associations’ ‘fix the notion of parts of speech’ by ‘combining all substantives, adjectives, etc.’ (CG 138). However, ‘the traditional divisions of grammar’ ‘do not correspond to natural distinctions’ (CG 136) (cf. 3.23; 4.55; 5.73; 6.49; 8.43; 13.24). ‘The mind creates as many associative series as there are diverse relations’, though Saussure's editors suggest that ‘the mind naturally discards associations that becloud the intelligibility of discourse’ (CG 125ff).

2.66 The mentalist outlook is crucial here because ‘co-ordinations formed outside discourse’ ‘are not supported by linearity’ or by ‘the theory of syntagms’ (CG 123, 136). ‘Their seat is in the brain; they are part of the inner storehouse that makes up the language of each speaker’ (CG 123) (2.16). ‘These associations fix word-families, inflectional paradigms, and formative elements (radicals, suffixes, inflectional endings, etc.) in our minds’ (CG 138). Perhaps Saussure's inclination toward mentalism on this point reflects his determination to keep his ‘science’ clear of ‘speaking’, the domain which, as we shall see with Bloomfield, Pike, and Firth, best supports a non-mentalist orientation. All the same, the ‘functioning of the dual system’ Saussure depicts must be inferred from actual ‘discourse’ (CG 129) before it can be projected into ‘our minds’ (cf. 13.1).²⁶

2.67 Such problems are conspicuously less acute in respect to the sounds of language, the area which Saussure, like many of our theorists, considered most basic (cf. 2.17, 70f; 3.18, 58f; 4.30, 79; 5.42, 5¹²; 7.20, 72; 8.66f; 12.80, 82; 13.27). In ‘the domain of phonetics’, the ‘absolute distinction between diachrony and synchrony’ is easiest to ‘maintain’ (CG 141). The same point could, I think, be made for the division between ‘language’ and ‘speaking’, or between ‘social’ and ‘individual’: the sounds of language possess a more reliable identity apart from any one set of occurrences than do, say, meanings (cf. 2.85; 13.27).

2.68 Saussure's assessment differs from the one favoured in later linguistics when he uses ‘phonetics’ for ‘the study of the evolution of sounds’, and ‘phonology’ for ‘the physiology of sounds’ (CG 33) (cf. 4.30; 6.43; 8.70; 12.80). In addition, he avers that ‘phonetics is a basic part of the science of language; phonology’ ‘is only an auxiliary discipline and belongs exclusively to speaking’. He repeatedly warns against ‘lumping together’ the two ‘absolutely distinct disciplines’. The ‘principles of phonology’ are concerned with ‘the phonational mechanism’ and ‘mechanical’ ‘elements’ (CG 48, 51, 38). But ‘phonational movements do not constitute language’; ‘explaining all the movements of the vocal apparatus’ ‘in no way illuminates the problem of language’ (CG 33) (cf. 2.71, 77; 3.17; 4.29; 6.7).

2.69 Elsewhere, though, he uses the term ‘phonology’ in its later standard sense: ‘the description of the sounds of a language-state’ (CG 140). ‘We must draw up for each language studied a phonological system’ comprising ‘a fixed number of well-differentiated phonemes’ (CG 34) (cf. 4.29f, 33f, 45; 5.42f; 6.43; 8³⁵; 12.80, 89; 13.26). ‘This system’ is declared ‘the only set of facts that interests the linguist’; ‘graphic symbols bear but a faint resemblance to it’ (CG 34f). ‘Modern linguists have finally seen the light’ and ‘freed’ ‘linguistics’ ‘from the written word’ (CG 32f), although his own exclusion of writing was not maintained (2.21ff). His ‘rational method’ for ‘dealing with a living language’ includes both ‘(a) setting up the system of sounds revealed by direct observation, and (b) observing the system of signs used to represent -- imperfectly -- these sounds’ (CG 37). ‘Phonology’ can ‘provide precautionary measures for dealing with the written form’ (CG 34). He even concedes that ‘the perceptible image of the written word’ keeps us from ‘perceiving only a shapeless and unmanageable mass’; ‘apart from their graphic symbols, sounds are only vague notions’ (CG 32) (cf. 2.23; 4.40; 6⁸; 8.71, 8³³). Surprisingly though, he recommends that ‘a phonological alphabet’, with ‘one symbol for each element’, be reserved for ‘linguists only’ (CG 33f) (cf. 8.75). ‘A page of phonological writing’ would present a ‘distressing appearance’ and be ‘weighed down by diacritical marks’. ‘Phonological exactitude is not very desirable outside science’ (cf. 4.32).

2.70 Saussure advocates a ‘science that uses binary combinations and sequences of phonemes as a point of departure’ (CG 50) (cf. 5.21, 40). This ‘science would treat articulatory moves like algebraic equations: a binary combination implies a certain number of mechanical and acoustical elements that mutually condition each other’ (CG 51) (cf. 2.60; 13.15). ‘In a phonational act’, i.e., ‘the production of sound by the vocal organs’, the ‘universal’ aspect transcending ‘all the local differences of its phonemes’ is ‘the mechanical regularity of the articulatory movements’ (CG 38, 51) (cf. 3.14, 21; 4.29; 5.42; 6.43; 7.20; 8.66, 70; 12.80; 13.26). In these ‘movements’, ‘a given sound obviously corresponds to a given act’ (CG 40). ‘All species of phonemes will be determined when all phonational acts are identified’ (CG 43). Accordingly, ‘the phonologist’ should ‘analyse a sufficient number of spoken chains from different languages’ in order to ‘identify and classify the elements’, ‘ignoring acoustically unimportant variations’ (CG 40).

2.71 A ‘natural point of departure for phonology’ is to ‘divide ‘the sound chain’ ‘into homogeneous’ ‘beats’, ‘each beat’ ‘corresponding’ to a ‘concrete irreducible unit’ and ‘characterized by unity of impression’ (CG 38, 53) (cf. 9¹⁶). ‘A phoneme is the sum of the auditory impressions and articulatory movements, the unit heard and the unit spoken, each conditioning the other’ (CG 40) (cf. 3.17f; 4.28ff; 5.43; 12.80f). ‘The auditory beat’ matches the ‘articulatory beat’. In fact, ‘auditory impressions exist unconsciously before phonological units are studied’ and enable ‘the observer’ to ‘single out subdivisions in the series of articulatory movements’ (CG 38). So ‘auditory impression’ is ‘the basis for any theory’ and ‘comes to us just as directly as the image of the moving vocal organs’ (CG 38). But Saussure offers a ‘classification of sounds according to their oral articulation’, even though these ‘movements do not constitute language’ (CG 44f, 33) (cf. 2.68, 77; 3.21; 4.34).²⁷

2.72 ‘The signifier, being auditory, is unfolded solely in time’ and ‘represents a span’ ‘measurable in a single dimension’ (CG 70) (cf. cf. 2.60; 13.33). In any ‘grouping’, a given ‘sound’ ‘stands in syntagmatic opposition to its environing sounds and in associative opposition to all other sounds that may come to mind’ (CG 131) (cf. 2.57). ‘Phonologists too often forget that language is made up’ ‘of expanses of spoken sounds’, whose ‘reciprocal relations’ merit ‘attention’ (CG 49f) (cf. 4.35; 8.65). Indeed, ‘the science of sounds becomes invaluable only when two or more elements are involved in a relationship based on their inner dependence’. Here, ‘combinatory phonology’ can ‘define the constant relations of interdependent phonemes’, such as that between ‘implosion and explosion’ (CG 51f) (cf. 2.23). ‘Freedom in linking phonological species is checked by the possibility of linking articulatory movements’.

2.73 Whereas ‘phonology is outside time, for the articulatory mechanism never changes’, ‘phonetics is a historical science, analysing events and changes and moving through time’, and therefore ‘the prime object of diachronic linguistics’ (CG 33, 140) (cf. 2.37). Though he believes ‘phonetic evolution is a disturbing force’, Saussure says ‘phonetic changes are absolutely regular’ in the sense that they ‘result in the identical alteration of all words containing the same phoneme’ (CG 161 153, 143; cf. CG 35). However, ‘absolute changes are extremely rare’; more often, ‘what is transformed’ is ‘the phoneme as it occurs under certain conditions -- its environment, accentuation, etc’. (CG 144). Saussure distinguishes ‘spontaneous and combinatory phonetic phenomena’, the former having an ‘internal’ ‘cause’, and the latter ‘resulting from the presence of one or more other phonemes’.²⁸

2.74 ‘Phonetic changes’ may seem ‘unlimited and incalculable’ (CG 151), but some limits are postulated. For example, ‘phonetic evolution cannot create two forms to replace one’ (CG 155). So ‘phonetic doublets do not exist; the evolution of sounds only emphasizes previous differences’ (CG 157). ‘The same unit cannot be subjected at the same time and in the same place to two different transformations’ (CG 155f). Every ‘duality’ or ‘alternation’ thus gets classified as ‘grammatical and synchronic’, ‘absolutely unrelated to phonetic changes’ (CG 156ff). Here, ‘the diachronic character of phonetics fits in very well with the principle that anything which is phonetic is neither significant nor grammatical’ (CG 141) (cf. 2.54f). If ‘phonetic changes attack only the material substance of words’, ‘in studying the history of the sounds in a word we may ignore meaning’ and ‘consider only the material envelope of a word’ (CG 18, 141). Nonetheless, when ‘phonetic modifications’ ‘result in alternations’ or ‘oppositions’, ‘the mind seizes upon the material difference, gives it significance, and makes it the carrier of conceptual difference’ or ‘attaches grammatical values’ (CG 159, 231). Once again, causes and effects get put into different theoretical domains (cf. 2.41).

2.75 By dividing things up this way, Saussure provides no proper home for ‘etymology’, the history of both forms and their meanings: it is ‘neither a distinct discipline nor a division of evolutionary linguistics’ (CG 189). ‘It is only a special application of principles that relate to synchronic and diachronic facts’. ‘Analogy’ is called upon to ‘show’ that ‘the synchronic relation of several different terms’ ‘is the most important part of etymological research’. ‘Etymology is then mainly the explaining of words through the historical study of their relations with other words’. Its ‘description’ of ‘facts’ ‘is not methodical, for it’ ‘borrows its data alternately from phonetics, morphology, semantics, etc.’ (CG 190). It ‘uses every means placed at its disposal by linguistics, but it is not concerned with the nature of the operations it is obliged to perform’. Besides, ‘etymology’ is fraught with ‘uncertainty’: ‘words with well-established origins’ are ‘rare’, and ‘scholars’ may be led into ‘rashness’ (CG 225).

2.76 To seek ‘the causes of phonetic changes’ is to confront ‘one of the most difficult problems in linguistics’ (CG 147) (cf. 3.54-60; 4.75). Some possibilities are rejected: ‘racial predispositions’, ‘soil and climate’, and ‘changes in fashion’ (CG 147, 151) (cf. 3²; 4.80). Others are provisionally accepted, though not as complete or conclusive causes: ‘the law of least effort’; ‘phonetic education during childhood’; ‘political instability’ of a ‘nation’; and the ‘linguistic substratum’ of an ‘indigenous population’ ‘absorbed’ by ‘newcomers’ (CG 148-51). Yet if we believe ‘a historical event must have some determining cause’ (CG 150), we will be hard put to explain why certain changes and no others occurred at just the times they did.

2.77 The programme for the study of sounds outlined by Saussure has remained a fundamental part of linguistics, though the emphasis on sound changes has receded. Having a mentalist orientation, he wanted a theory that would not depend on ‘material’ aspects (2.16f) and insisted ‘the movements of the vocal apparatus’ do not ‘illuminate the problem of language’ (2.68, 72). But his ultimate recourse was a ‘classification’ based on ‘oral articulation’ (2.71; 13.26).

2.78 Perhaps to offset the abstractness of language, Saussure, like many linguists, draws comparisons with more tangible entities. Though he misprizes the ‘illogical metaphors’ of rival ‘schools’, he admits that ‘certain metaphors are indispensable’ (CG 5). Some of his own are fairly proximate, e.g., when he compares ‘language’ to ‘a dictionary of which identical copies have been distributed to each individual’; or ‘the social side of speech’ to ‘a contract signed by the members of a community’; or ‘the vocal organs’ to ‘electrical devices used in transmitting the Morse code’ (CG 19, 14, 18).

2.79 Other metaphors are more remote, e.g., when the language system is pictured in terms of a ‘theatre’, ‘a symphony’, ‘a tapestry’, ‘a garment covered with patches cut from its own cloth’, or ‘the planets that revolve around the sun’ (CG 179, 18, 33, 172, 84f). A ‘system of phonemes’ is said to work like a ‘piano’ (CG 94). Studying ‘the evolution of language’ is compared to ‘sketching a panorama of the Alps’ and ‘moving’ ‘from one peak of the Jura to another’ (CG 82). ‘The autonomy’ of ‘synchrony’ is analogous to ‘the projection of an object on a plane surface’ or to ‘the stem of a plant’ ‘cut transversely’ (CG 87). ‘The word is like a house in which the arrangement and function of different rooms has been changed’; or like a ‘five-franc piece’ that ‘can be exchanged for a fixed quantity’ or ‘compared with similar values’ (CG 183, 115). ‘A linguistic unit is like the fixed part of a building, e.g. a column’ (CG 123f). ‘Thought’ and ‘sound’ resemble ‘the air in contact with a sheet of water’, or the ‘front’ and ‘back’ of ‘a sheet of paper’ (CG 112f, 115).²⁹ ‘Trains’ and ‘streets’ are enlisted to expound the interplay of ‘differences and identities’ (CG 108f). The ‘analogical fact’ is portrayed as ‘a play with a cast of three’ -- the ‘legitimate heir’, ‘the rival’, and ‘a collective character’ (CG 163). ‘The description of a language state’ is modelled after the ‘grammar of the Stock Exchange’, which suggests a more everyday sense for the term ‘values’ (CG 134) (cf. 2.27ff, 36, 58).

2.80 Saussure's ‘most fruitful’ ‘comparison’ is ‘drawn between the functioning of language and a game of chess’ (CG 88f, 22f, 95, 107, 110) (cf. 6.51; 9⁴⁹; 11.4). ‘The respective value of the pieces depends on their position on the chessboard just as each linguistic term derives its value from its opposition to all the other terms’ (CG 88) (cf. 2.26, 57f, 72). Though ‘the system’ ‘varies from one position to the next’, ‘the set of rules’ ‘persists’ and ‘outlives all events’ (CG 88, 95). The ‘material make-up’ of the pieces has no ‘effect on the “grammar” of the game’ (CG 110, 23). However, ‘chess’ is ‘artificial’, whereas ‘language’ is ‘natural’; and ‘the chessplayer intends’ to ‘exert an action on the system, whereas language premeditates nothing’: ‘the pieces’ are ‘modified spontaneously and fortuitously’ (CG 88f) (cf. 2.52, 63, 73).

2.81 Such metaphors relieve Saussure's abstract vision of language by introducing objects or events that could be seen or felt, and whose reality admits little doubt. Yet even the most complex metaphor, the chess game, falls far short of the complexity of language. The rules and pieces of chess are known to anyone who plays the game, and disputes about them are unlikely to arise. The rules and units of language are so numerous, diffuse, and adaptable that even experts seldom agree on any large number of them. A ‘linguistic term’ rarely stands in such a clear and stable ‘opposition to all the others’ as a bishop or a knight differs from all other chess pieces.

2.82 The abstractness of language can also be offset by comparing linguistics to other ‘sciences’ like ‘geology’, ‘zoology’, ‘astronomy’, and ‘chemistry’ (CG 213, 53, 106f) (cf. 13.11). These sciences have reasonably concrete object domains; but Saussure's favoured model was mathematics, which does not (cf. 3.73; 4.21; 13.15). ‘Language’ can be conceived as ‘a type of algebra consisting solely of complex terms’ (CG 122) (cf. 3.72f; 5.27, 86; 6.8, 29, 51, 60; 7.40, 7¹⁸). ‘Relations’ should be ‘expressed’ by ‘algebraic formulas’, ‘proportions’, and ‘equations’, though Saussure does not expect a ‘formula’ to ‘explain the phenomenon’ (CG 122, 164f, 166f, 169). Moreover, ‘studying a language-state means in practice disregarding changes of little importance, just as mathematicians disregard infinitesimal quantities in certain calculations’ (CG 102).

2.83 And building a science of language was Saussure's ultimate aspiration. Presumably, the reason why he ‘probably would not have authorized the publication of these pages’ (2.1) was that his own conceptions seemed too unstable and unsatisfactory to fit his ideals of science. He firmly asserted categorical dichotomies, but could not always maintain them himself, e.g. ‘synchronic’ versus ‘diachronic’ (2.38f), or ‘collective’ versus ‘individual’ (2.20, 61). He emphasized that language is social and psychological, yet wanted linguistics cleanly separated from sociology and psychology (2.7, 16, 28, 31ff, 35, 78; cf. 13.14).³⁰ He situated language in the minds of speakers, but could not decide how far the speaker's knowledge of a language is comparable to the categorical framework of linguistics (cf. 2.33f, 36, 42, 45, 59, 63, 63).³¹ He vacillated between mentalism and mechanism in appealing to notions like ‘brain’, ‘mind’, and ‘thought’ (2.16f, 18, 27, 31ff, 35, 40, 52, 57, 63, 65f, 74; 13.10), yet repeatedly referring to language itself as a ‘mechanism’ (CG 87, 103, 108, 111, 121, 133, 161, 165). Perhaps he wanted to deflect the issues of intention and will (cf. 2.20, 47, 64, 80).

2.84 Of course, the nature of language is so intricate and multiplex that its descriptions often entail inconsistencies, and a pioneering disquisition like the Cours is liable to be full of them. It both asserts and doubts that linguistics should involve a study of speech, pay attention to writing, and accept the word as a basic unit (2.19, 21, 23, 60, 18). Inconsistencies also beset the views that traditional grammar was a mistaken enterprise (2.5f); that grammar has a historical aspect (2.55); and that language is essentially arbitrary (2.28f). Some of these vacillations may be due to the improviso circumstances of its composition, or to the carelessness or exaggerated reverence of the editors, who do not comment upon them. But more importantly, language seems to have been resisting Saussure's determined campaign to make it hold still, to be as static, orderly and precisely circumscribed as he wanted it to be (cf. 13.52).

2.85 Some of the abstractions and dichotomies he deployed in this campaign tended to disperse the very factors that might have assisted him (cf. 13.55). His dismissal of ‘speaking’ and thus of actual discourse led him to inflate ‘the arbitrary nature of the sign’ (2.28f, 47), to fall back on ‘association’ and ‘opposition’ (2.57ff), and to neglect methods of data-gathering (cf. 13.27, 45). His turn against ‘diachrony’ left him deeply perplexed about ‘time’ and ‘history’ (cf. 2.12, 20, 34, 36, 39f, 42f, 45f, 55, 59f, 72f, 75f). Arguing from the neat oppositions of phonemic systems clashed sharply with the elusive, often metaphoric handling of semantics in terms of ‘concepts’, ‘ideas’, ‘thoughts’, and ‘signifieds’ (2.17, 25-28, 31f, 48, 52, 58, 62).

2.86 Many of Saussure's successors have underestimated the intricacies and qualifications within his arguments. Some of his terms, concepts, and dichotomies have been taken at face value, oversimplified, or treated as absolutes for the theory, doctrine, and organization of linguistics. This premature and selective orthodoxy has not merely misrepresented Saussure's intent to raise issues and problems rather than to resolve them, but has impeded comprehensive solutions. The reach of his vision is best revealed in the way that the same perplexities and dilemmas both explicit and implicit in his book have persisted in linguistics ever since. We are still uncertain about how a language is related to the multitude of speech events in the experience of language users, including linguists or grammarians (13.26, 49). We are still without an account of time and space in language (13.33). Disputes still rage over the status of rules or laws applying to all languages, and over the nature of linguistic units, especially in semantics (13.26ff, 60). Written language still dominates the representational methods of theories ostensibly concerned with spoken language (13.33). And little headway has been made in determining what sort of causalities apply in language, and how.

2.87 Thus, Saussure's deliberations deserve their place at the outset of ‘modern linguistics’ by virtue of their problematic nature as well as their monumental scope. He thought it ‘evident’ ‘that linguistic questions interest all who work with texts’ (CG 7). Consequently, ‘that linguistics should continue to be the prerogative of a few specialists would be unthinkable -- everyone is concerned with it in one way or another’ (cf. 3.2). To be sure, Saussure's own work was a major contributor to the specializing of language models. But if read with the care they deserve, his inaugural deliberations provide both an inspiring and a sobering impetus for reconsidering how to stake out possible topographies of the discipline.

NOTES ON SAUSSURE

¹ The Course in General Linguistics, translated by Wade Baskin, is cited as CG. I occasionally use square brackets to give the original French or to emend the English translation, which I found generally reliable. For a thorough exegesis of Saussure's ‘manuscript sources’ see Godel (1957).

² On inconsistencies, see 2.83f, and Notes 3, 10, 12, 14, 16, 17, 20, 22, 23, 29, and 32.

³ He expressly points out such dictions as ‘language does this or that’ or the ‘life of language’ (CG 5n). Further terms he both condemned and used include ‘material’, ‘natural’, ‘organism’, and ‘mechanism’ (2.17, 30, 83).

⁴ Scholars cited here include: for the first stage none, for the second stage Friedrich August Wolf, and for the third stage Franz Bopp (1816), William Jones, Jacob Grimm (1822-36), August Friedrich Pott, Adalbert Kuhn, Theodor Benfey, Theodor Aufrecht, Max Muller (1861), Georg Curtius (1879), and August Schleicher (1861) (CG 1-5) (see my References for presumable source-works). Among those who ‘brought linguistics nearer its true object’, mention is made of Friedrich Christian Diez (1836-38), Dwight Whitney (1875), and ‘the neogrammarians (Junggrammatiker) [Friedrich] K[arl] Brugmann, H[elmut] Ostoff, the Germanic scholars W[ilhelm] Braune, E[duard] Sievers, and H[ermann] Paul, and the Slavic scholar [August] Leskien’ (CG 5) (again, see References). The works receive only cursory, mainly negative commentary, or none, except for Bopp's. A fuller coverage is given by Bloomfield. Firth declared homage to very early grammarians, but only to English ones (cf. 8.15, 8¹²).

⁵ Saussure's first major work was his Memoire sur le systeme primitif des voyelles dans les langues indo-europeenes, published in 1878 and still regarded today as a milestone in philology.

⁶ This abstruse argument reflects Saussure's belief that historical changes have a less systematic organization than the language at any single point in time (2.14, 19f). But elsewhere he says it is ‘a serious mistake to consider dissimilar facts as a single phenomenon’ (CG 146).

⁷ Nor is the word a reliable tool for theory: ‘starting from words in defining things is a bad procedure’, and ‘all definitions of words are made in vain’ (CG 14).

⁸ The study of etymology falls between the cracks in Saussure's scheme, since it is historical and yet not limited to word-sounds. See 2.75.

⁹ In denying ‘phonetic doublets’, Saussure dispatches one case because one of two forms ‘is only a learned borrowing’; two more cases are passed over as ‘literary French’ (CG 156) (cf. 4.83). However, Saussure is not terribly well-disposed toward dialects either: because they conflict with his vision of the unified, closed system, he makes a shaky argument that diversity in space is ‘actually’ diversity in time and thus would fall under his exclusion of diachrony (cf 2.43, 2¹⁷).

¹⁰ In his last lectures (‘from May to July 1911), de Saussure used interchangeably the old terminology (“idea” and “sign”) and the new (“signifier” and “signified”)’ (CG 75, translator's note). The new became standard, especially among semioticians.

¹¹ Whitney is credited with ‘insisting upon the arbitrary nature of the sign’, though he ‘did not follow through’ by making it a defining trait of language (CG 76).

¹² Comparisons between ‘language’ and a ‘plant’ are also decried, though Saussure later compares the static and evolutionary versions of linguistics to cutting a plant ‘transversely’ or ‘longitudinally’ (CG 4, 87f) (2.79).

¹³ ‘Two objections’ are met by arguing that ‘onomatopoeia’ and ‘interjections’ are either outside the system, or if they do enter they become ‘unmotivated’ (CG 69). Compare 3⁴.

¹⁴ ‘The sound-image is sensory, and if I happen to call it “material”, it is only in that sense, and by way of opposing it’ to ‘the concept’ (CG 66).

¹⁵ Saussure argues from here that ‘linguistics’ has ‘only one method’ (CG 212), namely the ‘synchronic’ one he favours (CG 212) (2.36). He did not consider that isolating an ideal, static state of the language may increase rather than reduce the number of possible methods, as the subsequent development of linguistics showed (cf. Ch. 7). On the question of whether linguistic constructs match the knowledge of speakers, compare 2.40, 42, 44, 47, 53, 59, 63, 83; 3.11, 19, 57, 2²², and the passages cited in 13.49.

¹⁶ See for instance CG 38, 47, 72, 118, 138, 165; and compare the passages cited in 13.49. In one passage where the translation has ‘speakers are not conscious’, the original French has ‘the language [langue] is not conscious’ (CG 47).

¹⁷ In one section, ‘geographical isolation’ is waved aside as an ‘unsatisfactory and superficial explanation; differentiation can always be explained without it’ (CG 210). But in another, ‘geographical separation’ is judged ‘the most general force in linguistic diversity’ (CG 193).

¹⁸ The other occasion is ‘Saussure's treatment of holds’, i.e., ‘intermediate stretches’ in ‘spoken chains’, as both ‘mechanical and acoustic entities’ (CG 52).

¹⁹ ‘Non-phonetic’ because Saussure decides to make ‘analogy’ ‘grammatical’ (2.54). In return, ‘phonetics’ is made the centre of ‘diachronic linguistics’ (2.73), befitting the preoccupation of philology with sound changes.

²⁰ In another formulation, ‘language’ is claimed to do the same thing (CG 169) (2.52). The discussion of how innovations occur (cf. 2.45) suggests that ‘speech’ is probably the better term.

²¹ Languages in which ‘most words are not decomposable’ are termed ‘lexicological’, the others being ‘grammatical’ (CG 166) (cf. 2.29; 3.53). Making words out of decomposable units helps ‘limit arbitrariness’ (CG 133) (2.29). This notion of ‘lexicological’ seems to befit the idea of the lexicon being a listing of irregularities (13.59).

²² Elsewhere, however, Saussure concedes that ‘grammatical classes evolve’, thus putting in question ‘the absolute distinction between diachrony and synchrony’ (CG 141). ‘Once the phonetic force is eliminated, we find a residue that seems to justify the idea of a “history of grammar”‘; but ‘the distinction between diachrony and synchrony’ is still judged ‘indispensable’ for reasons ‘calling for detailed explanations outside the scope of this course’ (CG 143).

²³ But compare some formulations suggesting diversity instead: ‘speakers often single out several kinds’ or ‘grades of radicals in the same family of words’; ‘delimitations will vary according to the nature of the terms compared’; and ‘the speaker may make every imaginable division’ (CG 185, 188).

²⁴ The editors comment: ‘the two phenomena act jointly in the history of language, but agglutination always comes first’ and ‘furnishes models for analogy’ (CG 178). If not followed up by ‘analogy’, ‘agglutination’ ‘produces only unanalysable or unproductive words’.

²⁵ The term ‘paradigmatic’ prevents confusion with the different kind of ‘association’ Saussure postulates between sound-image and concept, or between signifier and signified (e.g. CG 14f, 18f, 65f, 76, 102) (2.25).

²⁶ An exception might be where ‘an identical function’ among various forms ‘creates the association in absence of any material support’ (CG 138). A case in point would be the ‘zero sign’ (CG 86, 186) (cf. 4³; 5.56; 6¹⁶; 7.75, 90).

²⁷ The editors ‘supplement’ ‘Saussure's brief description’ with ‘material’ from Otto Jespersen, but claim to be ‘merely carrying out de Saussure's intent’ (CG 41).

²⁸ ‘But a spontaneous fact’ ‘may be conditioned negatively by the absence of certain forces of change’ -- an odd stipulation in a conception devised for ‘the classing of changes’ (CG 144f).