On the Conative Function of Language*

A V. Isacenko

The problem of the basic functions of language was first posed, not by linguists, but by psychologists. In this respect we owe much to Karl Buhler, who formulated his point of view¹ as early as 1927. He later returned, on several occasions, to the question of the different functions of language,² summing up the results of his investigations in his magnum opus Sprach- theorie. ³

Taking Plato’s view as his point of departure, Buhler de- fines language as an “organon” which enables “somebody to tell somebody else something about objects”. In this scheme three poles must be clearly distinguished: “somebody”, i.e., the speaker, “somebody else”, i.e., the listener or the addressee, and the “objects”, i. e., the content of the utterance. A correla- tion of these three poles is given in any linguistic situation and is manifested by means of human speech (“to tell something”).

Further on, Buhler defines the three basic functions of language by stating that in any linguistic manifestation one of the three poles prevails. One may distinguish between expres- sion (Ausdruck), conation (Appell) and reference (Darstellung). Having thus delimited the basic functions of language, Buhler analyses different aspects of the central notion of linguistics, viz., the notion of the linguistic sign. A linguistic sign used in the expressive function is a symptom of something lying outside the linguistic sphere. Thus interjections such as oh, ah are symptoms of a certain psychical or physical state. A linguistic sign used in the conative function acquires the character of a signal. Thus interjections such as hey, hello function in speech as signals. Finally, a linguistic sign used in the referential function is to be interpreted as a symbol of a portion of extra- linguistic reality. A noun such as house is, according to Biihler, a symbol of the corresponding thing.

In the course of a systematic analysis of linguistic structure it soon became evident that only referential utterances permit a maximally instructive insight into the structure of language. On the other hand, in the realm of expressive and conative speech, we are confronted all the time with phenomena which seem to disturb the symmetry of the system and to obscure the regularity of structural linguistic relations. This is why the most convincing regularities of linguistic structure are demon- strated on partly artificial, often even on prepared materials, purged of anything not belonging to the “intellectual”, “neutral” style of speech. Facts not considered as “normal” are exclu- ded as exceptions or vaguely explained away as being “expres- sive”, “affectional” or “emotional”.

Such exceptions or anomalities were observed mainly in the repertory of phonemes. In certain categories, e. g., in onomat- opoeic words, in interjections, in words of command, elements are to be found which do not occur in the ordinary use of speech and which clearly do not enter into the phonemic system of the observed languages. These anomalies concern not only the phonemic repertory, but also the repertory of phonemic com- binations. Thus combinations of prepalatal [ t, d, n, ] + [ u ] in Czech do not occur in neutral utterances and are explained as “expressive”.⁴

The observation of such phonemic anomalies led the Czech linguist J. M. Korînek to postulate the term “inter jectionality” as opposed to “notionality” for such linguistic manifestations. Delimitations of linguistic functions on the basis of phonetic features were similarly outlined by other Czech linguists, es- pecially by V. Mathesius, V. Skalicka, and V. Machek.

From the very beginning, however, it seemed to be clear that any utterance may combine two or even three functions simultaneously, but that in any given instance one of these func- tions is dominant.

Let us now return to Plato’s definition. It follows from his maxim “somebody tells somebody else something about ob- jects” that in those cases where the referential function is stressed, our main attention is centered on the “objects” of the extralinguistic world. “Notional” utterances contrast with utterances in which our attitude toward the “objects” is veiled and in which an emotioned attitude between the participants of the dialogue is present. Such utterances with subdued notion- ality are mostly conceived as belonging to homogeneous “anom- alous” or “expressive” linguistics.

Commenting on the development of structural linguistics, N. S. Troubetzkoy states that the expressive and the conative levels were at first practically ignored by scholars.⁵ According to Troubetzkoy the Hungarian linguist Gy. Laziczius was the first to point out the need for a systematic analysis of the ex- pressive and the conative aspects of language. Laziczius him- self, however, was mainly concerned with symptomatic, i. e., expressive linguistic elements.⁶ He did not distinguish the ex- pressive function proper from the conative function. Laziczius classifies linguistic materials into three groups: (1) phonemes, fulfilling all three functions; (2) emphatemes (appearing only in the expressive function), and (3) variants, possessing only the expressive function.

The conative function proper has been treated, as far as I can see, only by N. S. Troubetzkoy.⁷ His views on this sub- ject may be summoned up as follows: “Phonological conation is the means which serves to evoke certain feelings in the hearer”. These feelings, Troubetzkoy continues, are frequently experienced by the speaker as well. But the essential thing about the conative function is this: the listener is “infected” by these feelings. What, according to Troubetzkoy, is focussed in conation, are not the personal feelings of the speaker but the evocation of corresponding feelings in the mind of the listener. As an example Troubetzkoy quotes the German word schooon ! pronounced with a certain intonation to evoke a certain emotion in the listener. He suggests that, for every given language, the inventory of conative means, i. e. of conventional phonetic means serving to “evoke feelings and emotions”, should be drawn up.

Such an interpretation of the conative function does not, however, cover the conative facts. Troubetzkoy’s definition is not only a tribute to psychologism in linguistics; it also serves to obscure the limits between the expressive and the conative functions. Since both these functions have to do with feelings and emotions, the only difference between them, according to Troubetzkoy, is that conative utterances serve to “evoke” feel- ings and emotions in the listener whereas expressive utterances do not.

A different point of view was formulated by J. M. Korînek, who tried to identify linguistic functions with what he called linguistic “styles”. J. M. Korînek divided linguistic manifesta- tions into three groups, according as to whether they focus “Truth, Beauty and Good”. Accordingly, Korînek recognizes three linguistic styles: (1) the logical style, in which the focus is on Truth and which he identifies with the referential function (2) the aesthetical style, where the focus is on Beauty and is identified with the expressive function, and, (3) the ethical style, with the focus on Good and identified with the conative function.⁸ But this scheme obviously suffers from a-priority and remains unproductive.

In order to avoid psychologism as well as a-priority in the interpretation of linguistic facts it is best to introduce a func- tional point of view. Buhler’s stressing of the importance of commands and exhortations for the definition of conation seems to be closer to an adequate interpretation of our problem. Of course, the conative function of language covers a much larger field.

The fundamental difference between the expressive (in the narrow sense of the word) and the conative function of language is determined by the difference in the general orientation of the utterance. In expressive utterances it is symptomaticity that prevails, the speaker’s personality being “the things talked about”. In conative utterances the orientation toward the listener is decisive. The fined aim of conation is not a prediction about one’s own self or a manifestation of the speaker’s mental state. The aim of conation is to provoke a reaction on the part of the listener. We note that the nature of reactions provoked by cona- tive utterances is different from reactions on the part of the listener, provoked by utterances with some other over-all lin- guistic function. Thus, a referential utterance aims in the first place at clearness of what is uttered in order to let the listener be affected by what has been said about “things, actions and states”. In an expressive utterance orientation toward the inner state of mind of the speaker prevails, the presence of a “second person” not being necessary at all. This is the reason why purely expressive utterances (lyrical poetry, incantations, prayers) are monological. Conative utterances are characterized by the fact that they presuppose the presence of a “second person” as well as immediate reactions on the side of the listener (i.e.reactions of any kind, including linguistic reaction), briefly, they pre- suppose re-actions in the basic sense of the word.

Any command or prohibition is, above all, a manifestation of the conative function. The purpose of commands is to se- cure obeyance, i. e. some kind of action. They either provoke certain manifestations of the listener’s behaviour or merely serve to attract the listener’s attention. If we want to consider as conative any linguistic manifestations which are oriented to- ward an active reaction on the part of the listener, we shall have to regard a question as a conative utterance, since any question presupposes an answer, i. e. a certain (linguistic) re- action.

The formal means by which such reactions are achieved are more often than not identical with the means that are familiar in the neutral plane of language. But frequently the formed means of conative utterances are in contrast with (phonemic) means used by the same language in other functions. In many languages the inventory of conative means contradicts the rules dominating the other planes of the language. Such violations of the “nationality” are usually treated as “expressive” markers. In reality many violations of the over-all rules are due not to expressivity, but to conation.

Deviations from the system of the unmarked (referential) linguistic plane may be demonstrated in many languages. We shall exemplify them by means of materials from a Ukrainian dialect of the Lemky group spoken in Ladomirova in Eastern Slovakia.

In this dialect word stress is not phonological; it is fixed on the penultimate, as in Polish or in Eastern Slovak dialects. Furthermore, the dialect of Ladomirova contains neither long vowels nor consonants. These statements are true as long as the neutral, presentative referential plane of language is being considered, but they do not refer to certain linguistic manifes- tations with a clearly conative function.

There exists in many ( if not all) languages, a system of “interjections” used with animals. In the dialect of Ladomirova we find that these calls may contain certain both long vowels and consonants, e.g. [hvo:]! to stop a horse, [c,ipic,ir: ] to call a chicken.

The dialect of Ladomirova has special conative forms for terms of relationship and for Christian names which are dif- ferent from the corresponding appellative forms:

Similar instances of vowel lengthening in personal names used in the “vocative” form have been reported from Russian dialects by S. P. Obnorskij.⁹

In the dialect of Ladomirova the rule regulating the position of stress on the penultimate syllable of the word has no excep- tions as long as the notional plane of language is concerned. In words based on calls to animals this rule is broken; we find words with stress on the last syllable:

The last syllable is stressed in the “vocative” forms of terms of relationship and Christian names:

For a full characterisation of the phonic means used in dif- ferent linguistic functions, it is necessary to give not only phonemic inventories together with a description of the réalisa- tion of the phonemes, but also an inventory of phonemic com- binations. The importance of the study of phonemic combinations was stressed by N. S. Troubetzkoy. Later this principle was applied by J. M. Korînek and V. Machek in connection with etymological problems. But by ascribing the violation of cer- tain combinatory rules to “expressivity” these authors did not distinguish clearly enough between the conative and the properly expressive functions of the language.

In the dialect of Ladomirova palatalized [s, ] and [z,] occur only before [i] (from East Slavic *ë), being thus variants of the phonemes /!/ and /z/; cf. the dative-locative sing, of the fem- inine nouns in -a: [k’as.i] ‘gruel’, [xiz, i] ‘house’. Except in this position the sounds [s, ] and [z,] normally do not occur. But in conative utterances (in the “vocative” form) palatalized [s,] and [z,] occur before /o/, cf. [m’is, o] ‘Michal’, [j’oz, o] ‘Joseph’, [l’az,o] ‘Ladislav’ etc.

This example illustrates a phenomenon which is possibly not devoid of general interest. The form [l’az,o] occurs in the community of Ladomirova in two functions: as a “vocative” form of the first name ‘Ladislav’ and as a family name. The members of the family are called [laz, ‘ovi], a female member of the family [laz, ‘ova]• On becoming a family name the word [1’az, o] changes its function. It has ceased to be only a “vocative” form of a Christian name and has become an official family name. Thus the conative function has been replaced by the referential function. Simultaneously the number of phoneme combinations on the referential plane has been enriched, since the sequence [z, o], although limited to the sphere of proper names, is now admitted in non-conative utterances.

Many linguists have drawn attention to phonic anomalies in the realm of proper names. These anomalies can be explained by the penetration of elements of the conative plane into the neutral plane. It seems clear that the adoption of conative (and expressive) elements by the notional language is one of the channels by which new phonemic elements may penetrate into the neutral sphere.

Up to this point we have treated only two categories of con- ative words, viz., calls to animals and “vocative” forms of proper names. It may be doubtful whether addressing animals can be considered as a linguistic manifestation in the full sense of the word, i. e. as a “glottic” phenomenon in O. Jespersen’s terminology. Utterances made to animals differ from those in a normal linguistic situation in that the “addressee” is not in command of the linguistic system. But since we have to do with utterances in which phonemic material is used, we may affirm that calls to animals still belong to “glottic” phenomena. These calls are not contingent sound sequences, but are based upon the phonemic system of the language and, accordingly, have the status of collective conventional signals.

In a discussion of the conative function, military commands deserve our special attention. Thus in many languages we find fronted vowels in military commands but not in other types of utterances. Thus Russian has fronted vowels only in the position after or before soft consonants. In the military command [s’âgam màrs] ‘forward march’ fronted [à] stands between two hard consonants.

We have so far treated only examples of conative phonology. Let us now turn to morphology and syntax.

The structural peculiarities of the Russian imperative, where agglutinative phenomena are in contrast with the over-all in- flexional character of the language, have been studied by Roman Jakobson¹⁰ and V. V. Vinogradov. ¹¹ Such accumulations of agglutinative elements as in projd-ëm-te -ê -ka ‘let us take a little walk’ are found only in the imperative, i. e. in a typically conative verbal form.

In English or in German the imperative is characterized by the absence of prefixes or suffixes, cf. come ! — I come, he comes, I came; German komm ! — ich komm-e, du komm-st, er komm-t. In this case the conative function is achieved by a zero morpheme.

One of the formed means by which conation is achieved is the modification of syntactic rules valid in the non-conative planes of the language. Examples are found in interrogative sentences in many languages. It has already been pointed out that the question, being a linguistic manifestation aiming at an immediate re-action (viz., an answer), has to be considered as one variety of conative utterance. In many languages inter- rogative sentences are marked by a special intonation and by the inversion of the word order of affirmative sentences; cf. French il vient ‘he comes’ — vient-il? ‘does he come?’, Slovak otec je doma ‘the father is at home’ — je otec doma? ^,is the father at home?’, German er kommt — kommt er?, English He is coming — Is he coming ? etc. Both the interrogative intonation and the inversion serve to evoke linguistic reactions.

Interjections are to be classified according to the specific function they fulfil. Along with clearly expressive interjections such as Russian [oj], [ox], [uf ], there exist conative interjections, such as hello, hush, Russian [ps], [a’u], used to keep contact between persons in a forest, Slovak kus, mars, curuk etc. Many of these conative interjections resemble imperatives and are partly treated as such. For instance the Russian inter- jections na ‘here you have’ and nu ^,come along’ take agglutina- tive suffixes like the imperative, cf. na-te-ka ‘here you have, hold this (polite form)’, nu-te-ka.

The study of the conative function and of its formal features is essential not only for synchronic linguistics but also for the elucidation of certain diachronic problems. In the process of linguistic evolution the delimitation of basic linguistic functions may become obscured, as shown in the case of proper names. A linguistic feature belonging, in a certain period, to the cona- tive function of language may lose its specific value and become a feature of the referential function, and vice versa.

The laws governing each of the basic linguistic functions are not identical for all languages. We would maintain that a detailed study of each of the basic linguistic functions is of paramount importance for an adequate understanding of the synchronic system as well as for the interpretation of numerous diachronic facts.

Notes

1. In a lecture delivered at the 10th Congress of Psycho- logists in Bonn, April 1927.

2. Kant-Studien 38.

3. Karl Buhler, Sprachtheorie, Jena 1934.

4. J. M. Korînek, Studie z oblasti onomatopoje. Prâce z védeckych ustavu 36, Prague, 1934. — V. Skalicka, Studie o mad’arskych vyrazech onomatopoickych, Sbornik filologicky 11, Prague, 1939•

5. N. S. Troubetzkoy, Grundzuge der Phonologie, Travaux du Cercle Linguistique de Prague 7, 18.

6. J. Laziczius, Problème der Phonologie, Ungarische Jahrbucher 15.

7. Grundzuge der Phonologie, 24.

8. J. M. Korînek, O jazykovem stylu [On Style in language], Slovo a slovesnost 7, 1941, 28-36.

9. Zeitschrift fur slavische Philologie, 1927.

10. Les enclitiques en slave. Atti del III Congres so Inter- nazionale dei linguisti, Rome, 1933, 384.

11. Sovremennyj russkij jazyk. Grammaticeskoje ucenije o slove, 2nd edition, Moscow, 1948, 59.

*Originally published in Russian under the title "0 prizyvnoj funkcii jazyka, " Recuenl Linguistique de Bratislava, I: 45 -57 ( 1948). rans slate by. V. IsaEenko,