Roman Jakobson (1959)
On linguistic Aspects of Translation
According to Bertrand Russell, “no one can understand the word ‘cheese’ unless he has a nonlinguistic acquaintance with cheese.”1 If, however, we follow Russell’s fundamental precept and place our “emphasis upon the linguistic aspects of traditional philosophical problems,” then we are obliged to state that no one can understand the word “cheese” unless he has an acquaintance with the meaning assigned to this word in the lexical code of English. Any representative of a cheese-less culinary culture will understand the English word “cheese” if he is aware that in this language it means “food made of pressed curds” and if he has at least a linguistic acquaintance with “curds.” We never consumed ambrosia or nectar and have only a linguistic acquaintance with the words “ambrosia,” “nectar,” and “gods” - the name of their mythical users; nonetheless, we understand these words and know in what contexts each of them may be used.
The meaning of the words “cheese,” “apple,” “nectar,” “acquaintance,” “but,” “mere,” and of any word or phrase whatsoever is definitely a linguistic - or to be more precise and less narrow - a semiotic fact. Against those who assign meaning (signatum) not to the sign, but to the thing itself, the simplest and truest argument would be that nobody has ever smelled or tasted the meaning of “cheese” or of “apple.” There is no signatum without signum. The meaning of the word “cheese” cannot be inferred from a nonlinguistic acquaintance with cheddar or with camembert without the assistance of the verbal code. An array of linguistic signs is needed to introduce an unfamiliar word. Mere pointing will not teach us whether “cheese” is the name of the given specimen, or of any box of camembert, or of camembert in general or of any cheese, any milk product, any food, any refreshment, or perhaps any box irrespective of contents. Finally, does a word simply name the thing in question, or does it imply a meaning such as offering, sale, prohibition, or malediction? (Pointing actually may mean malediction; in some cultures, particularly in Africa, it is an ominous gesture.)
For us, both as linguists and as ordinary word-users, the meaning of any linguistic sign is its translation into some further, alternative sign, especially a sign “in which it is more fully developed” as Peirce, the deepest inquirer into the essence of signs, insistently stated.2 The term “bachelor” may be converted into a more explicit designation, “unmarried man,” whenever higher explicitness is required. We distinguish three ways of interpreting a verbal sign: it may be translated into other signs of the same language, into another language, or into another, nonverbal system of symbols. These three kinds of translation are to be differently labeled:
1 Intralingual translation or rewording is an interpretation of verbal signs by means of other signs of the same language.
2 Interlingual translation or translation proper is an interpretation of verbal signs by means of some other language.
3 Intersemiotic translation or transmutation is an interpretation of verbal signs by means of signs of nonverbal sign systems.
The intralingual translation of a word uses either another, more or less synonymous, word or resorts to a circumlocution. Yet synonymy, as a rule, is not complete equivalence: for example, “every celibate is a bachelor, but not every bachelor is a celibate.” A word or an idiomatic phrase-word, briefly a code-unit of the highest level, may be fully interpreted only by means of an equivalent combination of code-units, i.e., a message referring to this code-unit: “every bachelor is an unmarried man, and every unmarried man is a bachelor,” or “every celibate is bound not to marry, and everyone who is bound not to marry is a celibate.”
Likewise, on the level of interlingual translation, there is ordinarily no full equivalence between code-units, while messages may serve as adequate interpretations of alien code-units or messages. The English word “cheese” cannot be completely identified with its standard Russian heteronym “сыр,” because cottage cheese is a cheese but not a сыр. Russians say: принеси сыру и творогу “bring cheese and [sic] cottage cheese.” In standard Russian, the food made of pressed curds is called сыр only if ferment is used.
Most frequently, however, translation from one language into another substitutes messages in one language not for separate code-units but for entire messages in same other language. Such a translation is a reported speech; the translator recodes and transmits a message received from another source. Thus translation involves two equivalent messages in two different codes.
Equivalence in difference is the cardinal problem of language and the pivotal concern of linguistics. Like any receiver of verbal messages, the linguist acts as their interpreter. No linguistic specimen may be interpreted by the science of language without a translation of its signs into other signs of the same system or into signs of another system. Any comparison of two languages implies an examination of their mutual translatability; widespread practice of interlingual communication, particularly translating activities, must be kept under constant scrutiny by linguistic science. It is difficult to overestimate the urgent need for and the theoretical and practical significance of differential bilingual dictionaries with careful comparative definition of all the corresponding units in their intention and extension. Likewise differential bilingual grammars should define what unifies and what differentiates the two languages in their selection and delimitation of grammatical concepts.
Both the practice and the theory of translation abound with intricacies, and from time to time attempts are made to sever the Gordian knot by proclaiming the dogma of untranslatability. “Mr. Everyman, the natural logician,” vividly imagined by B. L. Whorf, is supposed to have arrived at the following bit of reasoning: “Facts are unlike to speakers whose language background provides for unlike formulation of them.”3 In the first years of the Russian revolution there were fanatic visionaries who argued in Soviet periodicals for a radical revision of traditional language and particularly for the weeding out of such misleading expressions as “sunrise” or “sunset.” Yet we still use this Ptolemaic imagery without implying a rejection of Copernican doctrine, and we can easily transform our customary talk about the rising and setting sun into a picture of the earth’s rotation simply because any sign is translatable into a sign in which it appears to us more fully developed and precise.
A faculty of speaking a given language implies a faculty of talking about this language. Such a “metalinguistic” operation permits revision and redefinition of the vocabulary used. The complementarity of both levels - object-language and metalanguage - was brought out by Niels Bohr: all well-defined experimental evidence must be expressed in ordinary language, “in which the practical use of every word stands in complementary relation to attempts of its strict definition.”4
All cognitive experience and its classification is conveyable in any existing language. Whenever there is deficiency, terminology may be qualified and amplified by loan-words or loan-translations, neologisms or semantic shifts, and finally, by circumlocutions. Thus in the newborn literary language of the Northeast Siberian Chukchees, “screw” is rendered as “rotating nail,” “steel” as “hard iron,” “tin” as “thin iron,” “chalk” as “writing soap,” “watch” as “hammering heart.” Even seemingly contradictory circumlocutions, like “electrical horse-ear” (электрическая конка), the first Russian name of the horseless street ear, or “flying steamship” (jena paragot), the Koryak term for the airplane, simply designate the electrical analogue of the horse-ear and the flying analogue of the steamer and do not impede communication, just as there is no semantic “noise” and disturbance in the double oxymoron - “cold beef-and-pork hot dog.”
No lack of grammatical device in the language translated into makes impossible a literal translation of the entire conceptual information contained in the original. The traditional conjunctions “and,” “or” are now supplemented by a new connective - “and/or” - which was discussed a few years ago in the witty book Federal Prose - How to Write in and/or for Washington.5 Of these three conjunctions, only the latter occurs in one of the Samoyed languages.6 Despite these differences in the inventory of conjunctions, all three varieties of messages observed in “federal prose” may be distinctly translated both into tradition al English and into this Samoyed language. Federal prose: 1) John and Peter, 2) John or Peter, 3) John and/ or Peter will come. Traditional English: 3) John and Peter or one of them will come. Samoyed: John and/ or Peter both will come, 2) John and/ or Peter, one of them will come.
If some grammatical category is absent in a given language, its meaning may be translated into this language by lexical means. Dual forms like Old Russian брата are translated with the help of the numeral: “two brothers.” It is more difficult to remain faithful to the original when we translate into a language provided with a certain grammatical category from a language devoid of such a category. When translating the English sentence “She has brothers” into a language which discriminates dual and plural, we are compelled either to make our own choice between two statements “She has two brothers” – “She has more than two” or to leave the decision to the listener and say: “She has either two or more than two brothers.” Again in translating from a language without grammatical number into English one is obliged to select one of the two possibilities - “brother” or “brothers” or to confront the receiver of this message with a two-choice situation: “She has either one or more than one brother.”
As Boas neatly observed, the grammatical pattern of a language (as opposed to its lexical stock) determines those aspects of each experience that must be expressed in the given language: “We have to choose between these aspects, and one or the other must be chosen.”7 In order to translate accurately the English sentence “I hired a worker,” a Russian needs supplementary information, whether this action was completed or not and whether the worker was a man or a woman, because he must make his choice between a verb of completive or noncompletive aspect - нанял or нанимал - and between a masculine and feminine noun - работника or работницу. If I ask the utterer of the English sentence whether the worker was male or female, my question may be judged irrelevant or indiscreet, whereas in the Russian version of this sentence an answer to this question is obligatory. On the other hand, whatever the choice of Russian grammatical forms to translate the quoted English message, the translation will give no answer to the question of whether I “hired” or “have hired” the worker, or whether he/she was an indefinite or definite worker (“a” or “the”). Because the information required by the English and Russian grammatical pattern is unlike, we face quite different sets of two-choice situations; therefore a chain of translations of one and the same isolated sentence from English into Russian and vice versa could entirely deprive such a message of its initial content. The Geneva linguist S. Karcevski used to
compare such a gradual loss with a circular series of unfavorable currency transactions. But evidently the richer the context of a message, the smaller the loss of information.
Languages differ essentially in what they must convey and not in what they may convey. Each verb of a given language imperatively raises a set of specific yes-or-no questions, as for instance: is the narrated event conceived with or without reference to its completion? Is the narrated event presented as prior to the speed event or not? Naturally the attention of native speakers and listeners will be constantly focused on such items as are compulsory in their verbal code.
In its cognitive function, language is minimally dependent on the grammatical pattern because the definition of our experience stands in complementary relation to metalinguistic operations - the cognitive level of language not only admits but directly requires recoding interpretation, i.e., translation. Any assumption of ineffable or untranslatable cognitive data would be a contradiction in terms. But in jest, in dreams, in magic, briefly, in what one would call everyday verbal mythology and in poetry above all, the grammatical categories carry a high semantic import. In these conditions, the question of translation becomes much mare entangled and controversial.
Even such a category as grammatical gender, often cited as merely formal, plays a great role in the mythological attitudes of a speech community. In Russian the feminine cannot designate a male person, nor the masculine specify a female. Ways of personifying or metaphorically interpreting inanimate nouns are prompted by their gender. A test in the Moscow Psychological Institute (1915) showed that Russians, prone to personify the weekdays, consistently represented Monday, Tuesday, and Thursday as males and Wednesday, Friday, and Saturday as females, without realizing that this distribution was due to the masculine gender of the first three names (понедельник, вторник, четверг) as against the feminine gender of the others (среда, пятница, суббота). The fact that the word for Friday is masculine in some Slavic languages and feminine in others is reflected in the folk traditions of the corresponding peoples, which differ in their Friday ritual. The widespread Russian superstition that a fallen knife presages a male guest and a fallen fork a female one is determined by the masculine gender of нож “knife” and the feminine of вилка “fork” in Russian. In Slavic and other languages where “day” is masculine and “night” feminine, day is represented by poets as the lover of night. The Russian painter Repin was baffled as to why Sin had been depicted as a woman by German artists: he did not realize that “sin” is feminine in German (die Sünde), but masculine in Russian (грех). Likewise a Russian child, while reading a translation of German tales, was astounded to find that Death, obviously a woman (Russian смерть, fem.), was pictured as an old man (German der Tod, masc.). My Sister Life, the title of a book of poems by Boris Pasternak, is quite natural in Russian, where “life” is feminine жизнь, but was enough to reduce to despair the Czech poet Josef Hora in his attempt to translate these poems, since in Czech this noun is masculine život.
What was the initial question which arose in Slavic literature at its very beginning? Curiously enough, the translator’s difficulty in preserving the symbolism of genders, and the cognitive irrelevance of this difficulty, appears to be the main topic of the earliest Slavic original work, the preface to the first translation of the Evangeliarium, made in the early 860’s by the founder of Slavic letters and liturgy, Constantine the Philosopher, and recently restored and interpreted by A. Vaillant.8 “Greek, when translated into another language, cannot always be reproduced identically, and that happens to each language being translated,” the Slavic apostle states. “Masculine nouns ‘river’ and ‘star’ in Greek, are feminine in another language as река and звезда in Slavic.” According to Vaillant’s commentary, this divergence effaces the symbolic identification of the rivers with demons and of the stars with angels in the Slavic translation of two of Matthew’s verses (7:25 and 2:9). But to this poetic obstacle, Saint Constantine resolutely opposes the precept of Dionysius the Areopagite, who called for chief attention to the cognitive values (силе разума) and not to the words themselves.
In poetry, verbal equations become a constructive principle of the text. Syntactic and morphological categories, roots, and affixes, phonemes and their components (distinctive features) - in short, any constituents of the verbal code are confronted, juxtaposed, brought into contiguous relation according to the principle of similarity and contrast and carry their own autonomous signification. Phonemic similarity is sensed as semantic relationship. The pull, or to use a more erudite, and perhaps more precise term - paronomasia, reigns over poetic art, and whether its rule is absolute or limited, poetry by definition is untranslatable. Only creative transposition is possible: either intralingual transposition - from one poetic shape into another, or interlingual transposition - from one language into another, or finally intersemiotic transposition - from one system of signs into another, e.g., from verbal art into music, dance, cinema, or painting.
If we were to translate into English the traditional formula Traduttore, traditore as “the translator is a betrayer,” we would deprive the Italian rhyming epigram of all its paronomastic value. Hence a cognitive attitude would compel us to change this aphorism into a more explicit statement and to answer the questions: translator of what messages? betrayer of what values?
1 Bertrand Russell, «Logical Positivism,» Revue Internationale de Philosophie, IV
(1950), 18; cf. p. 3.
2 Cf. John Dewey, «Peirce’s Theory of Linguistic Signs, Thought, and Meaning,»
The Journal of Philosophy, XLIII (1946), 91.
3 Benjamin Lee Whorf, Language, Thought, and Reality (Cambridge, Mass.,
1956), p. 235.
4 Niels Bohr, «On the Notions of Causality and Complementarity,» Dialectica, I
5 James R. Masterson and Wendell Brooks Phillips, Federal Prose (Chapel Hill,
N. C., 1948), p. 40f.
6 Cf. Knut Bergsland, «Finsk-ugrisk og almen språkvitenskap,» Norsk Tidsskrift
for Sproavidenskap, xv (1949), 374f.
7 Franz Boas, «Language,» General Anthropology (Boston, 1938), FP, 132f.
8 Andre Vaillant, «Le Préface de l’Évangeliaire vieux-slave,» Revue des Études
Slaves, XXIV (1948), 5f.
Roman Jakobson - On Linguistic Aspects of Translation - full text