The Need for a Linguistics of “la parole”*

Vladimir Skalička

In recent times, linguistics has considerably broadened its horizons. The older linguistics, which followed up the admirable but one-sided conquests of the school of Junggrammatiker, was devoted mainly to phonetics and morphology, considered historically. Today, the whole range of the new schools is looking for new tasks, and still new tasks, for linguistics.

So-called structuralism, one of these schools, avails itself of various divisions of language phenomena, including that of de Saussure into la langue and la parole. In de Saussure’s time such an antithesis was badly needed. It was de Saussure who established once more the principles of a linguistics of langue, in his book, Cours de linguistique générale, Paris, 1922,¹ now a classic. The editors of this Cours, Ch. Bally and A. Séche- haye, point out that it would be possible to criticize De Saussure’s book for its neglect of a linguistics of la parole.² Structuralism, which possessed great advantages for the investigation of language, continued on de Saussure’s lines. In its phonology, so-called, it had analysed the phonetic systems of world languages and had laid the foundations of a newly conceived literary science.

But there are facts which compel us to reflect anew on the problems of la parole. And that is what we wish to do in this article. We have not the remotest idea, of course, of surveying the whole theory of parole. We want only to examine its problems.

First of all, we want to look at the historical conditions of the antithesis between langue and parole. The older linguistics, in the 20 to 25 centuries preceding the nineteenth, was entirely under the influence of langue, as though it were a catch-word. The broad problem which we have inherited from those times, be it be the problem of the origin of language or of the development of one language from another, of accuracy of language, analogy and anomaly, interlinguistics, descriptive grammar, orthography, etc. etc., the structure of all these problems rests upon a basis of la langue. Only in the 19th century was this bias corrected and under various catchwords the linguistics of parole came into its own. It was formulated most explicitly at the beginning of the 19th century, by W. v. Humboldt.³ Starting from Aristotle’s antithesis between Ιργον and ένέργεια, he declares that language is not Ιργον, something ready-made, already created, but ένέργεία, something that is always being created. Historical grammar, so brilliantly worked out in the 19th century, is dominated by la parole, as by a slogan. De Saussure establishes a link with the old concept of langue, and structuralism makes still greater use of his concept, so that the linguistics of parole remains in the background.

Structuralism is not, however, the only school which is considered valid today. As structuralism is founded first of all on langue, so the so-called “idealistic neophilology” is founded on parole. As proof, let us quote the words of the foremost representative of this school, K. Vossler: „Denn die Sprache als Funktion ist ein Abstraktum, ein entleerter Begriff, der den Fortschritt und das Leben niemals wird fassen können, der auch dadurch nicht voller und beweglicher wird, dass man ihm das vollste und beweglichste, nämlich den Begriff des Lebens selbst, an die Seite gibt. Der Begriff des Lebens schleppt den seine Funktion wie seinen eigenen Leichnam hinter sich her. Soll dieser Leichnam wieder lebendig und fortschrittsfähig werden, so muss das Leben in ihn hineinfahren, so muss die Funktion als Handlung, nicht als Geschehen, als Energeia, nicht als Ergon, und auch nicht als blinde Handlung und Energie, sondern als bewusste scharfäugige, geistige Tätigkeit gedacht werden.”⁴

Evident from the words quoted is first of all a connection with Humboldt’s ideas (cf. the antithesis ‘έργον — ένέργεια ). However, there is clearly a discrepancy between this and the ideas of current historical grammar.

Current historical grammar inquires only into the sequence of phonetic and morphological changes, whereas the idealistic neophilology wishes to see in the development of a language a life, a spiritual activity. And so this school sees in the. individual utterance the expression of a personality⁵ and, in the individual language, the expression of a characteristic culture.⁶

How does this help us today, concerning ourselves with the idea of la parole? Today it is already old-fashioned to talk about “scharfäugige, geistige Tätigkeit”, about “das vollste und Beweglichste, nämlich dem Begriff des Lebens selbst”. Idealistic chatter as a whole is now decidedly out of date. On the other hand, we cannot afford to ignore the labors of Vossler and his colleagues. They tried to investigate in detail the connection between an individual and his speech, a nation and its language. If one tries to establish such a link, one must turn to the idealistic neophilology, as, for example, does Volosinov.⁷

The second group of facts that oblige us to reckon with the problem of la parole lies outside linguistics itself. Outside linguistics, there is linguistic theory, occupying itself with much that linguistics proper ignores; that is to say, with la parole too.

There is, firstly, simple unscientific language, which is theory about what surrounds us and also about itself, i. ?., about language. So there are many words in a language which clearly form a theory of parole, cf., for example, to speak, speech, to stammer, to stutter, to ask, to answer, to speak to the point, to speak at random, intelligible, comprehensible, ambiguous, etc., etc. All are expressions to which the scientific terms of linguistics do not apply. These expressions, however, point to realities of a linguistic nature.

We are aware, moreover, of other branches of science with sub-divisions which are occupied — of course, in a non-linguistic way — with clearly linguistic realities that belong unquestionably to the sphere of la parole.

There is first of all psychiatry, with its teachings on aphasia; it even bases its diagnosis on speech utterances. Psychology, which is based on the most various observations, has something to say about speech utterances too. Psychological typology deals with the ways in which individual types express themselves in speech.

Thirdly, there are still other facts which oblige us to examine the problems of parole again. In various authors certain statements are to be found, for example, thoughts which come near to the idea of parole but do not recognise it, or at least have not been related to and harmonised with it.

Some studies approach the concept of parole very closely; for example, passages from the work of J. Mukarovsky, who otherwise devotes to the idea of parole on!y a few hasty remarks.⁸ The problem of dialogue and monologue⁹ comes, for example, very close to problems of parole. But most obviously concerning itself with the question of parole is the study “Genetika smyslu v Máchovê poesii”.¹⁰ Mukarovsky’s older conception of “deformation” too belongs clearly to the domain of parole.”¹¹

Similarly belong there the reflections of B. Havránek on automatism and actualism.¹²

Finally, we must remind ourselves that some linguists have used a recognised antithesis in an attempt to do justice to the antithesis between langue and parole. Thus A. H. Gardiner in his work The Theory of Speech and Language discovers the antithesis between langue and parole in that between word and sentence.¹³ This is a wrong line of thought and so we cannot proceed further from such a notion. We also cannot agree with the use of a single term for spéech and utterance, like the Russian “ vy skazy vani j e “.¹⁴

We have cited an array of facts which oblige us to consider again the problem of la parole. And now we can proceed to our own thesis.

What is parole in fact? The answer appears to be quite simple. Parole consists of the individual utterances by which langue is made real. But we cannot solve everything thus. There remains the problem of how the code is illustrated in speech, of how parole changes it, etc. — in short, the whole of parole remains as a problem.

The great merit of K. Bühler’s work is that he pointed out again (for the first time since the Greek philosophers) that there are three agents in people’s speech: the subject (ego), the second subject (tu) and the world apart from both subject (is, id). Words, sentences and statements can exist only in connection with these three elements. To this „Organonmodell der Sprache” of BÜHLER it is necessary to add that the three-sidedness applies only to paroie. Langue is a collection of rules, in which all three persons (ego, tu, id) are present only as images, and not as agents, and therefore ego and tu are here ingredients of id. For example, in terms like ego, me, meus, fero, ferimus, ferebam, and in so-called personal style (though J. Mukarovsky talks about individual speech¹⁵ ) the 1st person is never present as agent but only as image. Opposed to it in parole, the 1st and 2nd persons are present as agents, giving to parole a source and meaning. Here is to be found the basic difference between langue and parole, and we therefore begin our essay with the 1st and 2nd persons.

Firstly, we must point out that parole is a dual formation. Parole is not only conversation, with writing, but also listening with reading. This is made clearly evident in examples where ego and tu are divided from each other by time (written literature) or place (broadcasting) or where there is a quantitative disproportion between ego and tu (literature, broadcasting). With every reader, moreover j as well as listener, the individual utterance is something quite different; everyone makes it concrete in a different way,¹⁶ i.e. conceives it otherwise. Let us give an example of a dissimilar concrétisation: For Solomon’s „Song of Songs” or Homer’s poems we have our current interpretation. Moreover, both these creations are interpreted in an allegorical manner.

We do not wish to say here that speaking, as well as listening, is simply parole, or that concrétisation is parole. The matter is not nearly so simple. An individual utterance always has a stronger or weaker tendency towards general validity, and individual concrétisations fight to be recognised as valid.¹⁷ A typical example of such general validation is the so-called cento , i. e. a poetical work composed of the lines of another work. Thus, for example, the Byzantine Passion Play Χριστός πάσχων, of the 11th or 12th century, is a cento of Euripides’ works. parole and langue are not two independent phenomena which accidentally meet but two aspects of the same phenomenon, and always antithetical.

Now the question arises of the way in which the first two persons (1 and 2) achieve validity in speech. Thus we arrive at the antithesis between automatism and freedom, used by J. van Ginneken.¹⁸

I think we cannot dispense with the antithesis between automatism and freedom. The conception of automatism is necessary for the elucidation of the concept of parole. It is quite obvious that speaking is for the most part done· automatically. It is quite evident that to the question “What is it like outside?” we quite automatically answer, “It is raining”. Automatic, in fact still more automatic, is parole for the listener or reader. We are simply aware of the answer “It is raining” and we act in accordance with it.¹⁹

Of course, we must point out that the automatism of speech is by no means a simple matter. It might appear that a firm unchangable langue achieves validity in automatism as opposed to the freedom of parole. But the matter is not so simple. Automatic speech itself is subject to changes. There are plenty of „errors” in it, because it is not controlled. Let us only listen to common people’s speech! What a lot of automatically occurring slips of the tongue there are, of barbarisms, of solecisms, of anacoluthons ! And it is just the same with the listener. Some listeners automatically comprehend what the speaker wanted to say, others automatically choose what suits them.

In every utterance there is a certain element of automatism, and also a certain degree of freedom. De Saussure drew our attention to this freedom of parole. « Le propre de la parole c’est la liberté des combinaisons».²⁰ It is necessary, however, to point out that this participation of freedom does not concern the speaker only (1st person) — that is self-evident — but also the listener (2nd person), who, as we stated before, „concretizes” every utterance in his own way.

There is now the question as to what this freedom of which we speak is concerned with. B. Havránek opposes automatism and actualisation (animation) to each other.²¹ Certainly it is a just antithesis, but too biassed. Actualisation is only one side of the freedom of parole. The freedom of parole is concerned with the choice of words from a great number of synonyms, with the choice of forms, of constructions, etc. Choice is everywhere at work, and is achieved by contrasting various possibilities. This is done by reflection, experiment, dialogue, etc.

Let us pass to the third person, to “what is spoken about”. This third person is the reality beyond language. This reality does not enter into the language directly, but into its image. As Aristotle says: Περί ερμηνείας, 1: “Εστί μ£ν ουν τ& έν τη φωνη τΟν έν τη ψυχη παθημάτων σύμβολα, καΙ τα γραφόμενα των έν τη φωνη. Kat ωσπερ ούδί: γράμματα πδσι. τα αύτά, ούδ£ φωναί αί αύταΙ, ων μέντοι. ταύτα σημεΓα πρώτως, ταύτα πασι. παθήματα της ψυχής, καί ων ταύτα ομοιώματα, πράγματα ήδη ταύτα.²²

We see that reality does not enter into language directly, but into its image. And even in a double image: firstly, “soul states” are images (ομοιώματα) of things, and secondly, they are symbols (σύμβολα) of soul states.

These images are fixed — more or less — in langue. Langue is then a system of semiological devices, more or less fixed, expressing reality.

In speech new reality attacks Langue. The task of speech is to harmonise this new reality with the old, i.e., with experience. This means: to represent the new reality by more or less fixed images. Speech is thus a synthesis of the old Langue and the new reality.

Of course, both sides change in speech: the new reality is more or less truly represented, or misrepresented, and the system of images, which make up Langue, is somewhat changed. Both changes may be either superficial or deep.

The new reality is expressed in the best and easiest manner if we are speaking of ordinary occurrences: it is raining. In such an utterance, of course, Langue changes least.

On the other hand, it is very difficult to equalise experience with new reality if we have to express our attitude to some scientific question in the words and forms at our disposal. Scientific utterances are meant to express reality far removed from experience, just like philosophical or literary utterances. In such utterances, of course, it is most difficult to grasp the new reality and therefore it is done the least exactly, i. ?., in a changed way. Langue is changed simply in this way, whether imperceptibly or greatly. The change in Langue may be only experimental (cf., for example, the nominal sentences of the Czech poet K. J. Erben), or they may be definitive (cf. the elimination of Greek words in Horace’s satires, in contrast with the older satires).

Among these utterances the simplest case is represented by the scientific utterances which we may call reports. Imagine, for example, a work dealing with some chemical investigation. Such a piece of work changes Langue in the sense of giving to certain ideas a new and richer content. It is more difficult to describe the change which occurs in Langue during some theoretical investigation.. Then a struggle takes place between the new reality and the experience which has been captured in Langue. It is necessary, in approaching reality, to create new concepts, revise old ones, to combine concepts in a new manner, etc. The language of poetry moves along other ways, experiments, actualises and deforms, in order to catch some new shade of reality.

We have outlined the antithesis between Langue and parole in speech, considered generally. But we have not yet finished. In language there are formations, such as word, sentence and utterance. What is the relation between Langue and parole in these? Is it always the same, or different?

In order to elucidate the antithesis between iangue and parole in word, sentence and utterance, let us consider what is the semiological relationship within these formations. In our opinion the semiological relation is different in word, sentence and utterance. The very terms in some languages contradict it. For example, in Czech the word „vyznam” signifies “the meaning of a word”, while the word „ smy sl” usually denotes “the meaning of a sentence”.²³ In a similar way German distinguishes „Bedeutung” and „Sinn”.

Let us look at the meanings of the formations of individual languages. The word (or more exactly, the denomination, because sometimes the denomination occurs as two or more words, e.g., aes alienum, “debt”, ius iurandum, “oath”) is a symbol, as we have seen, of some “soul state”, which is itself the image of some sector of reality. Reality, as far as we know it, has been classified and catalogued in a dictionary in a great number of such images, running into hundreds of thousands (e. g., about 140, 000 ancient Greek words have survived; modern languages are still richer). But the number is not infinite at all. The sum total of the words may be arranged in a book, a dictionary. The number of words is constant, given by convention. The meaning of the words is also constant, given by convention. Of course, it is not so constant as the number of words. Every word has more than one meaning, and the struggle between them is most evident in the language of poetry.

In what way is this demonstrated in parole ? In speaking we use a lexicon, i. ?., a catalogue of words. The sieve of words is so fine that we can easily find old words for new reality. It follows from this that the creation of new words is very rare, that the normal habit is to use words already in existence. De Saussure reminds us: «Quand un mot comme indécorable surgit dans la parole, il suppose un type déterminé, et celui-ci à son tour n’est possible que par le souvenir d’un nombre suffisant de mots semblables appartenant Ճ la langue (impardonnable, intolérable, infatigable, etc.). >>²⁴ It follows that the meaning of words does not fully express what we want to express. The word meanings are only supporting points for the meaning of the sentence. Thus we have with words this picture: a very constant langue, with parole manifesting itself in the choice of words, and eventually in changes in their meaning.

Let us now look at the sentence. Here the situation is fundamentally different and also more complicated. It will be easiest for us to begin with syntax. It is the most passable and therefore the most beaten track to the sentence. On syntax, for example, Ch. Bally²⁵ and Ch. W. Morris ²⁶ build up their considerations.

What is syntax, in fact? It is something similar to the lexicon. There are certain relationships between words, and these relationships can be classified. Thus, for example, the analogous relationship between the words ,,οχ” and „low”, „ass” and „bray”, “wolf” and “howl”, “wolf” and “run”. The same with other syntactic relationships. Thus, as in a lexicon, a catalogue is created, quite small of course, a catalogue of syntactic elements.

Parole in syntax is analogous to parole in words. On meeting with new reality we always use old syntax, and the creation of new syntactic connections is quite rare.

It would, however, be a mistake to judge that syntax is the necessary foundation of a sentence. Utterances without syntax testify that syntax is not necessary to expression. The leader of the Futurists, F. T. Marinetti, says of this: « Ce fut en aéroplane, assis sur le cylindre Ճ essence, le ventre chauffé par la tête de l’aviateur, que je sentis tout Ճ coup l’inanité ridicule de la vieille syntaxe heritée de Homère. Besoin furieux de délivrer les mots en les tirant du cachot de la période latine. Elle a naturellement un ventre, deux jambes et deux pieds plats, mais n’aura jamais deux ailes. De quoi marcher, courir quelques instants et s’arrêter presqu’ aussitôt en soufflant. . .» Then he proclaims: « 1. Il faut détruire la syntaxe en disposant les substantifs en hasard de leur naissance. 2. Il faut employer le verbe Ճ l’infinitif, pour qu’il s’adapte élastiquement au substantif et ne le soumettre pas au moi de l’écrivain qui observe ou imagine. Le verbe à l’infinitif peut seul donner le sens continu de la vie et l’élasticité de l’intuition qui la perçoit. 3. Il faut abolir l’adjectif pour que le substantif nu garde sa couleur essentielle. . . 4. Il faut abolir l’adverbe, vieille agrafe qui tient attachés les mots ensemble. . . 5. Chaque substantif doit avoir son double, c’est-a-dire que le substantif doit être suivi, sans locution conjonctive, du substantif auquel il est lié par analogie. Exemples: homme — torpilleur, femme — rade, fouleressac, place — entonnoir, porte — robinet”.

So Marinetti’s description of a battle takes this form:

« Bataille

“Poids + Odeur

“Midi 3/4 flûtes glapissement embrasement toumbtoumb alarme Gargaresch craquement crépitation march Cliquetis sacs fusils sabots clous canons crinières roues caisons juifs beignets pains-á-huiles cantilènes échoppes bouffés chatoiement chassie puanteur cannelle fadeurs flux reflux poivre rixe vermine tourbillon orangers-en-fleurs filigrane misère dés échecs cartes jasmin + muscade + rose arabesque mosaïque charoque hérissement. »

It is clear from this example that the sentence can exist without syntax. This is especially important for parole of the sentence, and pre-eminently for parole of the speaker. As Mukarovsky says, from the moment when we conceive an initial series of meanings to be a sentence, we strive after the unity of sentence meaning.²⁸ We carry the whole sense, if necessary fore ing it, into any complex of words which is linked by the intonation.²⁹

Marinetti indicates the ends of sentences, as can be seen in the quotation, only by a wider gap between two words. It is true, of course, that his sentence is not so clear as a sentence built with the help of syntax. But we have no reason to deny its existence. Syntax is important and useful for the sentence, but it is not essential.

This is proved by other languages, too; for example, Chinese where syntactic relationships are much simpler and looser than in other languages and where the relationship between two words is often given only by their meaning.³⁰

If syntax is not an integral part of the sentence, what then makes a sentence a sentence? If we want to understand correctly the antithesis between langue and parole in the framework of the sentence, we must clarify the problem of the sentence itself. This does not mean, of course, that we should attempt a so-called definition of the sentence. We only want to discover its chief charac teristics. This is very difficult. We know that authors who attempted to define the sentence or to establish the main principles of the sentence, found it to have different characteristics.

Some scholars have seen in the sentence the outcome of preeminently synthetic, others of analytic, activity. H. Paul defined the sentence as follows: „Der Satz ist der sprachliche Ausdruck, das Symbol dafür, daß sich die Verbindung mehrerer Vorstellungen oder Vorstellungsgruppen in der Seele des Sprechenden vollzogen hat, und das Mittel dazu, die nämliche Verbindung der nämlichen Vorstellungen in der Seele des Hörenden zu erzeugen”.³¹ W. Wundt stated as follows: „Der sprachliche Ausdruck für die willkürliche Gliederung einer Gesamtvorstellung in ihre in logische Beziehungen zueinander gesetzten Bestandteile”.³²

It is clear that Paul’s definition considers the sentence to be the result of synthetic activity (linking up conceptions) while Wundt’s sees it as the result of analytic activity. Synthetic also is the syntactic idea of the sentence, an idea constructed on a basis of ancient Aristotelian logic.

Attention has often been drawn to the extra-syntactical relationships of words in a sentence, to what V. Mathesius calls “the actual dissection of the sentence”.³³ One dissects the sentence, for example, “A million people live in Prague” on the basis of the assertion, that about which something is said (as was once said, the psychological subject) — “in Prague” — and the essential point of the assertion, what is said about it (formerly the psychological predicate) — “a million people live”. This principle is sometimes valid in morphology too. In Japanese the “psychological subject” has a suffix or particle wa, e. g. Ni-hon wa yama ga yoi, “Japan is mountainous”, or literally, “Japan what concerns mountain (nomin.) much”.³⁴ Further, the principle of “actual dissection” is at bottom synthetic, since it is concerned with the associating of words in a certain manner.

J. Mukarovsky presented the synthetic principle in a very expressive way; he considers that one of the principles of the “significance structure of the sentence” is “the accumu!ation of meaning”.³⁵ The principle states that significance units (roughly speaking, words) are perceived in succession and linked up to one another:

Mukarovsky himself reminds us that Mathesius’ concept of actual dissection approaches the accumulation concept which has been demonstrated.³⁶

These were examples of the synthetic concept of the sentence. On the other hand there is the analytic concept, exemplified by W. Wundt. Let us quote some more examples.

F. Slotty considers that the meaning of a sentence, like the meaning of a word, is ambiguous. For example, we see in the street a man riding on a cart. This reality arouses our interest. Somebody says: “He is riding in a cart”, and another: “He is going for a ride by cart. “ That is: the same experience, the same objective content (Sachverhalt), which is the “intention” of the sentence, will be expressed by different people either locally or instrumentally.³⁷ We do not want here to analyse the whole matter in detail, but the analytic method is clear from these few words.

J. Mukarovsky knows three principles of sentence structure: 1. Unity of sentence meaning. 2. Accumulation of meaning. 3. Oscillation between static and dynamic meaning. ³⁸ We have already spoken about the second principle (accumulation of meaning) and have pointed out that it is a synthetic principle. The first principle, arising from the singleness of meaning in the sentence, is without doubt analytical. The third principle “is given by the fact that every unit of meaning in the sentence group (word, phrase) tends on the one hand to link the immediate relations to the referent, and on the other hand is bound by the cohesion of the sentence as a whole and only by means of this whole establishes connection with reality”. The third principle proceeds from word to sentence, sometimes from sentence to word — it is then the interplay of both principles, the analytic and the synthetic.

How does this help us in solving the problem of langue and parole in the sentence? From such considerations we can see first of all that syntax introduces mostly !angue into the sentence. Syntax has a catalogue of constituents which must be used. Langue is very strong in the antithesis psychological subject — psychological predicate, as is evident also from the fact that this antithesis is usually expressed morphologicaUy (see the Japanese example quoted). The relation between langue and parole appears more delicate to us if we consider the sentence as a whole in all its profundity. Langue of the sentence, i. ?., the synthesis of sentence norms, is more complicated than a catalogue of rules of syntax. We know from the secondary school that pupils often cannot decipher a Latin sentence from Caesar of which they know all the words — merely because they are not trained to decipher Latin sentences. A similar but still more difficult situation occurs when we learn Chinese. A knowledge of single words and rules of syntax will not help us — we must have a certain training as well, to be able to divine the sense of a Chinese sentence.

So we come to the substantial difference between word and sentence. The word is an image of some sector of reality — the sentence is the image of some contact with reality. Words can be collected into a dictionary; there would be no sense in doing so with sentences (only the most common sentences are arranged in “conversational handbooks”). A new word is an exception — but a new sentence is a regular occurrence. A word fully expresses the sector of reality which it has to express. But a sentence cannot express the infinite complexity of a single situation — it can only indicate the main points (with the help of words).

It follows from all this that langue in a sentence is much weaker than in a word. But on the other hand parole of the sentence is stronger than parole of the word. Freedom in the sentence is greater than in the word.

What has been said of the sentence applies in greater measure to the longer utterance. The principal difference between sentence and longer utterance is the lack of syntax in the longer utterance. So the participation of langue is markedly reduced, for syntax is the main pillar of langue in the sentence. On the other hand a sentence is not defined by the style — that concerns only the longer utterance.

Style may be defined as the individualising organisation of utterances.³⁹

That is to say, in constructing longer utterances we are always choosing, from possible synonyms and synonymous expressions, what will suit the particular style. There are common utterances, for which we use common words, common syntactic phrases, etc. There are scientific utterances, in which we use special terms, logical arrangements of sentences, etc. Then there are rules about appearance, i. ?., how an utterance, common, scientific, philosophical, etc., should be organised. Even personal style is not exempt from langue. The individual also uses his own definite expressions, which characterise him. In this way he creates his characteristic langue.

But where does parole remain? Does it get lost? Certainly that does not happen. There is a great difference between the parole of style and the parole of syntax;. For example, the parole of style has its rules and requirements, but it does not lay claim to them consistently. Deviation from style is not a mistake (as in syntax) but only a deviation, the reason for which can be looked for. Moreover, it is possible to choose between two styles, but it is not possible to choose between two syntaxes. Still less is it possible to mingle two syntaxes, but on the other hand it is possible to mix two styles (e. g. the conversational and the scientific, in an essay). Even from this it is obvious that the parole of the longer utterance is stronger than that of the sentence.

The parole of the longer utterance is, however, much stronger. Only in the utterance is shown a complete and responsible attitude towards new reality. It is here that the strongest need exists for an independent linguistics of parole.

Notes

1. The first edition apepared in 1915.

2. Cours de linguistique générale, p. 10.

3. W. v. Humboldt, Uber die Kawisprache auf der Insel Jawa, Berlin 1836, p. LV.

4. K. Vossler, Gesammelte Aufsätze zur Sprachphilosophie, Munich 1923, p. 101.

5. K. Vossler, Der Einzelne und die Sprache, Logos VII, 1913, or Gesammelte Aufsätze zur Sprachphilosophie, p. 152 et seq.

6. Cf. the title of the book Frankreichs Kultur im Spiegel seiner Sprachentwicklung, Heidelberg 1913, 1921* .

7. Cf. Marksizm i filosfia jazyka, Leningrad, 1929, p. 61 et seq.

8. Cf. Kapitoly ζ ceské poetiky I, pp. 48, 72.

9. Ibidem, cap. I, p. 145 et seq.

10. In the publication Torso a tajemstvi Mâchova díla, Prague 1938.

11. See the publication Spisovnâ cestina a jazykovâ kultura, Prague 1932, pp. 124 et seq.

12. Ibidem, pp. 52 et seq.

13. This is refuted by V. Mathesius, Slovo a slovesnost I, p. 42.

14. Ref. Volo linov, op. cit. p. 78 et seq., 131 et seq.

15. Cf. Kapitoly z ceské poetiky I, p. 48.

16. Concerning .concrétisation’ see F. Vodicka, Slovo a slovesnost VII, p. 117.

17. For example, the Stoics tried to demonstrate the allegorical interpretation of Homer. It is similarly interpreted by Herakleitos (1st century A. D.), in an essay published under the title Heracliti Quaestiones Homericae, Lipsiae, 1910.

18. J. νειη Gi nne к e η, Principles of Psychological Linguistics, Paris, 1907, p. 241 et seq. He builds here on a psychologi- cal basis, above all on the theory of Pierre Janet.

19. B. Havránek draws attention to this further automatism in Spisovná cestina a jazykovâ kultura, p. 52.

20. Cours de linguistique générale, p. 172.

21. Op>. cit., p. 52.

22. In the Latin translation by Didot it runs like this: Quae igitur sunt in voce, sunt notae passionum, quae sunt in anima; et quae scribuntur, sunt notae eorum quae sunt in voce, atque ut lit- terae non sunt apud omnes eaedem, ita пес voces sunt eaedem, sed passiones animi, quarum haec primum sunt signa, eaedem sunt apud omnes, eaedem sunt etiam res, quarum haec passiones sunt simulacra.

23. V á s a-T r á vni c e k, Slovnik jazyka ceského, 1941 , s. v. „vy’znam” and „smysl”.

24. Cours de linguistique générale, p. 173.

25. Linguistique générale et linguistique française, Paris 1932, p. 31 et seq. ~

26. Foundations of the Theory of Signs, Chicago 1938, p. 13 et seq.

28. Kapitoly ζ ceské poetiky I, p. 128.

29. Cf. S. Karcevskij, Travaux du Cercle Linguistique de Prague IV, p. 188.

30. See the author’s article Sur la typologie de la langue chinoise parlée, Archiv Orientální VI, 1945.

31. H. Paul, Prinzipien der Sprachgeschichte 5, 1920, p. 121.

32. W. Wundt, Völkerpsychologie, Die Sprache, 19 0 42 , II, p. 245.

33. See, for example, Ree a sloh in the publication Ctení o jazyce a poesii I, p. 58 et seq.

34. Lange, Lehrbuch der japanischen Umgangssprache, Berlin, 1922, p. 9.

35. Kapitoly z ceské poetiky I, p. 128.

36. Ibidem, p. 130.

37. Charisteria Guilelmo Mathesio quinquagenario a diseipulis et Circuii linguistici Pragensis so~5alibus oblata, “Prague, 1932, p. 92 et seq.

38. Օք. cit. p. 128.

39. See the author‘s article The Problems of Style, Slovo a slove sno st VII.

*From Recuei! Linguistique de Bratislava, I: 21-38 (1948).