Navigation – Plan du site

In Principio Erat Verbum :Traditional Principles Concerning Words and Their Order in Translating

Frederick M. Rener
p. 15-35


On attend généralement du traducteur consciencieux qu'il se montre fidèle à la structure de la phrase du texte original. Divers facteurs rendent cette tâche ardue et parfois impossible à accomplir. Parmi eux il y a la nature même du texte à traduire. Mais un facteur beaucoup plus décisif est la nature de la langue cible avec ses propres contraintes concernant l'ordre des mots dans la phrase. A travers l'histoire de la traduction, tant les théoriciens que les praticiens ont été confrontés à cette situation.
Cet article veut donner un aperçu des principes théoriques et des solutions pratiques auxquels ont recouru les traducteurs des XVIe et XVIIe siècles. L'entreprise était rendue plus difficile encore par le fait que la langue source était le latin, avec sa structure de phrase des plus complexes, et aussi par le fait que la syntaxe de la langue vernaculaire n'était pas encore stabilisée.

Haut de page

Notes de la rédaction

† Au moment de mettre sous presse Palimpsestes n° 7, nous apprenons avec tristesse la disparition de Frederick M. Rener le 27 janvier 1993. Nous saluons ici la mémoire de l'auteur d'un grand livre : Interpretano : Language and Translation from Cicero to Tytler. Amsterdam : Rodopi, 1989, 367 p.

Texte intégral

1It is of particular interest to be discussing aspects of translation at a time when walls are being torn down, barriers and curtains removed which had barred communication for decades. Removing walls has been the activity of translation since its inception. In fact, it is to walls, commonly called language barriers, that translation owes its existence. But there is more. The image of the wall served in the past to signify not only a division but also the very nature of language. This is true of Latin in particular. The operation of giving words the correct order is called in Latin constructio because it involves putting together words (verba struere) much like the mason puts together stones. The product of this operation is then a structura, a wall built with words.

2Seen in this context, translating means transferring the structura from one language to another and acquires, therefore, a close affinity with the art of masonry but in a rather unconventional manner. Since the translator must handle two languages he must be a specialist not only in building but also in wrecking. He has the double task of de-constructing before he begins the con-structing.

3The present is devoted to a comparative study of the modalities, the feasibility and the limitations of this double specialization in a few modern cases. My task is to introduce the theme of this volume by giving an outline of the situation in the past, when translating was still in principio, i.e. in its early stages. In doing so, I hope to provide a background to the articles which follow and thus give the discussion a broader horizon. Keeping in tune with the times, this will call for tearing down still another wall, the one dividing les anciens et les modernes. I hope to be able to show that at least on this point there are no querelles to report.

4I intend to deal with this topic in two major sections. The first focuses on the description of the classical constructio, i.e. the procedures employed during these centuries when placing words in the sequence required for their conveying a message. The second part focuses on translation and intends to give an outline of the translator's handling of the sentence both in the analytical (de-constructing) and in the creative (con-structing) phase. This second part is subdivided into two sections of which one is theoretical and relates the rules found in treatises of the time ; the second is practical as it examines a few concrete cases of this transfer. The material in this paper represents an amplification of some portions of my study on translation bearing the title Interpretatio.

5The principle employed in classical antiquity for this constructio is a progressive adding of parts, the smallest of which is the word. The construction begins, then, with the single word and thus we can say that in principio est verbum. Even though words carry meaning singly, they are intended as parts of larger unit, which is why Aristotle defined the word, in his Ars poetica, as "a part with meaning" (p. 59). Consequently, communication can take place only when such single parts are joined into sentences provided that this joining conforms to specific rules. It is only when the proper building code is applied that the structura acquires stability and fulfills its purpose.

6In the final chapter of his Le Tretté de la grammere françoeze (1551), Louis Meigret, the XVIth-century French grammarian whom we shall meet later in the discussion, having discussed the single parts of speech introduces constructio as the action of building a wall ("dresser un bâtiment") when he writes : "Voilà donque quant à ce que concerne les matières necessaires (i.e. the stones) pour dresser un bon bâtiment de language" (p. 194)

7He then gives the official term for this operation in Greek (syntaxis) and Latin (constructio), before trying to supply the still non-existent French term : "... que nous pouvons appeller bâtiment ou construccion, ou ordonance bonne de paroles" (p. 194).

8He is quick to underscore the extreme importance of this operation. The structural material is useless unless the wall is built with the proper tools ("plomb et ligne") and is thus given stability : "Car si nous ne les savons ordonner de sorte qu'on en pouisse tirer quelque sense, nous ne seront point différents de ceux qui font provision de toutes matières necessaires a dresser quelque gros édifice, sans tenir-aucune raison de comodité, ne de plomb, ne de ligne..." (p. 194).

9As mentioned above, the operation by which this "bâtiment" will be erected is called constructio or also compositio. What does the finished product, the sentence, look like and how is it arrived at ?

10This is not an easy question for the simple reason that the sentence is a product of two specialized disciplines, each with its own proper aim. As a result, a sentence can have two forms : it can be a purely functional structure without frills ; it can also be one loaded with ornaments. The former is the end product of the grammatical operations, the second a product of the rhetorical operations. Each one calls for a specific order of words : the functional structure employs the so-called ordo naturalis, its ornate counterpart the so-called ordo artificialis. Humphrey in his treatise on translation brings into play the traditional naturalars antithesis when he contrasts the two. He ascribes the former to natura (ordine naturae) and the latter to ars (artificio compositionis) :"... cum verba non ordine naturae, sed artificio compositionis ... struuntur." (497). It would be wrong to regard the two orders as antithetical and mutually exclusive. Rather, they are complementary in the sense that the ordo artificialis presupposes the ordo naturalis upon which it builds by modifying many of its features.

11In his Lexikon Philosophicum (an important repository of traditional knowledge) Rodolphus Goclenius, a renowed teacher at the University of Marburg at the end of the XVIth century, condensed the traditional notions in his description of compositio. He points out that there are three kinds of compositio, namely "grammatica, oratoria, et poetica" He writes :

Est compositio quaedam Grammatica tum vocum singularium tum vocum coniunctarum. Illa in Etymologia, haec in Syntaxi spectatur. Est Oratoria seu Rhetorica orationis ; est Poetica metri et carminis (s.v.).

12We have to keep in mind that for the primary sources "language" means Latin language. Even though a dead language, Latin played an extremely important part in the mentality of the time and became the model for the vernacular languages. The rules of grammar, the rhetorical devices and, of course, the aesthetic ideals were taken from Latin. The closer a vernacular came to the model, the higher was its standing. It is to be expected that in the early stages several Latin linguistic features, including the word order, were taken over by the vernacular. Meigret, among others, complains about the tendency of his contemporaries "de suivre le style Latin et d'abandonner le notre" (p. 195).

13The desire to follow the model often ran against grave obstacles represented by the patterns used in French which had been established by custom. Meigret remarks that such latinized sentences though using native words nevertheless render the sense obscure and offend the ear of the reader on account of the "mauvaise ordonnance" :

... que combien que leurs paroles soient naïvement françaises, la mauvaise ordonnance rend toutefois le sens obscure avec un grand mecontentement de l'oreille du lecteur et de l'assistence (p. 195).

14The first person to notice such non-transferable features in language was not the grammarian but the translator. He can thus be regarded as the first comparative grammarian who was able to determine the individual characteristics of each of the languages. The notion that each language had a linguistic personality all its own is, therefore, not a realization which surfaced in the XVIth century with Meigret, as Scaglione seems to suggests (p. 153), but goes back to the early stages of translation.

15Order in general can be defined as a sequential arrangement of parts. In speech, order is a sequential arrangement of Aristotle's "parts with meaning" for the purpose of conveying meaning. This sequence presupposes a hierarchy of sorts by which the placement of each part can be determined. The criterion for such hierarchy can be time, or size, or dignity, or function. It can also be a combination of these.

16In speech, parts are sorted according to their function in the sentence. There were eight groups to which a word could belong, groups known as parts of speech. Just by looking at the names of some of these groups, one can deduce that they have a lower position on the scale. The prefixes in the names of these items point to an auxiliary function : ad-verb, ad-jective, pronoun, pre-position, con-junction. Only two parts of speech, the noun and the verb, have a leading role in the sentence.

17This duumvirate was puzzling to many thinkers and led philosophers and grammarians to engage in an age-long discussion about principalitas i.e. which of these two parts had the most important function in the sentence, which of the two was primus inter pares. At the beginning of the XVIth century, this abstract dispute received a new form when it was changed into a story with a dramatic plot, in Andreas Guarna's booklet entitled Grammaticae opus novum published in 1511. The book was later known by its subtitle Bellum grammaticale and became such a success as to make any author green with envy.

18Johannes Nolte, the editor of a 1908 study and edition of Guarna's work, informs us that during the following two hundred years, the book became known all over the continent, even in remote areas such as Finland. It was reprinted at least 106 times in different countries particularly in the centre of Europe. Nolte counted 60 editions in Germany, 28 in France, and 13 in Italy. Translations were made into Italian, German, French, English, and even into Swedish. He mentions two cases in which the story was changed into Latin plays (tragicomoedia), one in Oxford and one in Munich, which were publicly performed.

19The book deals with a civil war in an otherwise idyllic country, known as Provincia Grammaticae and the bone of contention is principalitas. As we shall see below, the region excelled for its location and fertility and also for its particular form of government as it was ruled by two kings, Amo (the verb) and Poeta (the noun), who lived in perfect harmony. One day, after a good meal and much wine, this amiable relationship came to a sudden end. Mutual understanding gave way to a long and progressively angry contention which led to the an actual Grammaticale Bellum Nominis et Verbi regum de principalitate orationis inter se contendentium, as the author writes in the subtitle.

20In the first stage, this war was one of words as the two contendents employed rational arguments in their discussion such as the argument ex facto, showing their mutual dependence. The noun can accomplish little without a verb and the verb even less without a noun. Then followed arguments ex auctoritate. The noun claimed to be superior on grounds of age, because he was first in the history of the universe and because Deus is not a verb but a name. The verb retorted by quoting the opening sentence of St. John's Gospel : In principio erat verbum.

21Since rational arguments failed to convince, the two kings resorted to war. Each king calls to arms his vassals : in Amo's army we find the variety of adverbs (time, manner, place) and of verbs whereas Poeta calls the pronouns, the adjectives, the prepositions. The two armies gather on the plain of the conjunctions and are separated by the river Sive. A fierce battle ensues which brings casualties representing defective verbs or nouns. It would have raged all day if it were not for the intervention of a deus ex machina in the form of a violent storm (tempestate ingens superveniens aquarum vis) which showers tons of water on the hot heads and succeeds in bringing about a cease fire.

22The conflict was solved by arbitration. Three outstanding Italian grammarians are called to pronounce the verdict which succeeds in pacifying both kings although none of them is a winner. The jury decreed that when the noun was in the casus rectus (nominative or vocative) it ruled over the verb ; when it was in a casus obliquus (accusative, dative, genitive), the verb ruled over the noun.

23The primary goal of this book was pedagogical, namely to employ a novel approach in teaching the pupils the regularities and irregularities of Latin grammar. At the same time, it drew attention to the two principal parts of speech and their respective place in the sentence. Furthermore, by giving each part the position of a king (rex), the story reinforced the notion of order and regimen, the law governing the concatenation of the parts within the sentence.

24Throughout these centuries, the rules for the correct word order in the sentence were taught by the grammatica normativa which is not to be confused with the speculativa. A well-known grammatical manual used in schools was the Doctrinale puerorum (around 1200 A.D.) by Alexander of Villedieu in which the rules were schematized and reduced to verses for easier memorization.

25This book, too, was a long-lasting success for it was still used almost three hundred years later. In the last decades of the XVth century, there appeared an anonymous manual of Latin grammar which is of some interest for the students of translation. The author underscores the novelty of his method in which translation from the pupil's mother tongue, in this case German, is employed as an excellent device for learning Latin. The book is called Exercitium puerile grammaticale (1485). In the process of explaining the new method, the author covers also the rules governing the ordinatio constructibilium, the order of the structural parts.

26We learn that there are seven general rules for correctly ordering these parts. They are taken from the Doctrinale but the author rephrases them and thus makes them more understandable not only to his pupils but also to the modern reader. The list contains the rules for the regular word order (regulariter). What emerges here is a standard sentence in which the parts of speech appear in the following order : the nominative (subject), a verbum personale (a finite verb), nouns in the oblique cases either as an accusative (direct object) or as a dative and accusative (indirect and direct object.) The last item is the genitive which is governed (regimen) by the noun which precedes it.

27This sequence is called ordo naturalis because it is supposed to follow the natural path of the human thought (Humphrey's ratione naturae) as expressed in the triad agens/actio/passio. In such a sentence, an agent is followed by the action which in turn is followed by the effect of this action. The remaining parts are governed by regimen, i.e. a required sequel of the governer (régissant) and the governed (régi). The Exercitium offers as an example the expression "filius patris".

28I must mention parenthetically that the term ordo naturalis is sometimes employed in a slightly different sense. At issue here is not a sentence but a sequence of nouns which have to be placed according to the hierarchy determined by nature (dignitate naturae). Gasparino Barzizza writes : "Quae dignitate naturae in eodem genere vel in diversis praestant si earum rerum verba in oratione ponantur, minus illa dignis saepius anteferantur" (qtd. in Sonkowsky, p. 269). Thomas Wilson in his Art of Rhetorique (1553) gives a concrete example : "Some will set the Cart before the Horse as thus. My mother and my father are both at home, as though the good man of the house did weare no breches, or that the graie Mare were the better horse". He continues by saying that even though such things may actually happen, "in speaking at least let us keepe a naturall order, and set the man before the woman for maners sake" (qtd. in Scaglione, p. 156-57). Barzizza mentions other doublets in Latin which have been taken over by most European languages. In addition to viri et feminae which is the model for Thomas Wilson, Barzizza mentions dies ac noctes, virtutes et vitia, secundae res atque adversae (qtd. in Sonkowsky, p. 269).

29By following the seven rules in the ordinatio constructibilium, one arrives at a sentence pattern for the simplest and at the same time the most reliable statement, a statement which is sure to reach its destination. It comes as a surprise to learn that its use is restricted to didactic texts (ad exponendum) and when communicating with unschooled persons (rudis). The reason behind this attitude is not connected with grammar but with human psychology.

30In his Ars rhetorica, Aristotle brings to our attention an interesting feature of human nature. He says that experience teaches that things which are constantly repeated generate a feeling of overabundance (satietas) : "Quodcumque autem uno modo semper se habet, satietatem generat" (qtd. in Tesauro, p. 122). This feeling is associated with disgust and revulsion. Aristotle, therefore, suggests change and variety : "Varietas iucunda est" (ibid.).

31The principle that "varietas iucunda est" found application also in speech. Paradoxically, regularity, the distinctive feature of the normal sentence, may under certain circumstances become its weakest point. Faced by the sameness, the listener's attention slackens to the point of non-reception of the message despite its formal perfection. Variety in the selection of words (synonyms) and in the structure of the sentence is the proven remedy.

32This fact is one of the main reasons, though not the only one, behind the tendency to modify the regular sentence. This explains the readiness shown by the author of the Exercitium to complete his presentation of the word order by providing a second list containing seven exceptions (impedimentaliter) immediately following the seven principal rules (regulariter). Accordingly, the initial position of the sentence is no longer taken up by the noun (subject) but by a variety of other elements such as a dependent clause, an accusative with infinitive construction, an ablative absolute, and even an oblique case (accusative).

33The sentence, however, in which varietas becomes the dominant feature is the rhetorical sentence or periodus. In this sentence, words are not only "parts with meaning" but also parts equipped with acoustical, indeed musical characteristics : sound and rhythm. The position of each word in this sentence depends on its ability to fit a certain rhythmic pattern and produce a pleasant succession of sounds.

34This sentence is a complex structure with several subdivisions (colon, comma) often arranged symmetrically or antithetically. They are given structural visibility by means of special devices (figurae elocutionis) too numerous to be mentioned here. Two of them deserve our attention, however, because they are connected with two points seen above, namely repetition and regimen. In the periodus, several rules proper of the regular sentence are intentionally broken for artistic purposes. Thus subdivisions stand out because of the repetition of the same word either at its beginning (anaphora) or at the end (epiphora). Similarly, a genitive is often anteposed to its governing nominative (anastrophe or inversio). These intentional transgressions generate suspense and surprise in the listener and thus prevent boredom from setting in. Such "misplacements" by design go under the term ordo artificialis.

35Up to this point, we have covered several stages in the process of constructio :from the single words (parts of speech), to the skeleton sentence in expository texts, down to the rich display of symmetrical structures combined with pleasant musical effects in the rhetorical periodus. However, this is not the end of the line. The process continues by additional combinations of parts of larger dimensions.

36At this point, the sentence itself is regarded as a part which combines with others to form a larger unit which we shall call the text. As in earlier stages, combining parts should not be equated with mere juxtaposition. In this text every sentence is skillfully cemented to the next by several means called iuncturae. The outcome of all these operations is a harmonious and pleasant flow of words without impediments, a text in which to borrow Batteux's words : "tout coule sans effort" (p. 100).

37This is, then, the "bâtiment" which will become the object of translating. In his activity, the translator is thus confronted with a sequel of sentences, some short some long, some simple some complex, some built ordine naturae and the majority built artificio compositionis. His task is to proceed from sentence to sentence ("per sententias," as the author of the Exercitium urges in his paragraph "summopere animadvertendum") and to analytically take it apart in order to erect, on foreign grounds, as exact a replica as possible.


38We come now to the second and more important part of our considerations, the one which deals with the active phase of translating. We intend to briefly examine the methods and the problems connected with this difficult enterprise of restructuring the original text in the target language. This is the phase which puts great demands on all the talents of the translator. Batteux points out that only a person who has never engaged in this activity regards it to be easy : "Il n'y a que ceux qui n'ont jamais essayé de traduire les Auteurs anciens qui puissent douter combien cette entreprise est difficile" (p. 100). Obviously, they see not the labor but only the finished product "où tout coule sans effort" (p. 100).

39It is well known that the past employed two different approaches known as ad sensum and ad verbum, a terminology which acquired semi-official status especially after St. Jerome. It should be noted that these terms are far from reliable. In the course of centuries, they received a variety of interpretations ranging from strict to lax. Thus ad verbum stood at times for extreme literalness in which words, their gender, number, and placement were transferred to the target language without regard to meaning. On the other hand, ad sensum was at times interpreted as a complete disregard of the original wording.

40The decision about the approach to be used was not left to the translator but was subject to other factors. One of them was the nature of the text. Texts dealing with a subject matter where precision was at stake (theology, metaphysics and abstract sciences) called for a literal rendering. Other material was translated more freely.

41If one consults the theoretical sources, especially those of the XVIth century in which many treatises were written, one gets the impression that the literal approach is to be preferred without regard to the nature of the text. This is true at least as far as programmatic statements are concerned. After all, fidelity is the translator's main virtue and this must extend to the word order. Humphrey, for instance, requires the translator to be identical to the original : "... alius ne sit, at idem ..." (p. 403). Vives calls for the retention of all ornamentation :"... figurae et reliqua orationis ornamenta conservari debent" (p. 234). Fausto da Longiano is opposed to any changes in the word order "... hassi da osservare l'ordine de le parole senza trasporle" (p. 22v). Huet, the last of the Latin theoreticians, shares the general view : "Nec vocum ordinem temere deserendum"(p. 30).

42Having once stated the general principle, all of these authors are quick to add an important proviso concerning certain limitations. Vives employs the formula "quoad fieri possit," (p. 234) echoed by Fausto da Longiano's "quando acconciamente far si possa" (p. 22v).

43This vague formulation is replaced in most sources by a specific reference to the proprietates linguae as the main impediment to the total transfer of the word order. Peletier for instance says : "Le Translateur doeve garder la propriete et le naïfe de la Langue au laquele il translate" (p. 110). The term proprietates linguae is sometimes replaced by variants such as facultas linguae, consuetudo linguae, and also genius linguae which must be regarded as synonymous. Thus Humphrey, to quote at least one instance, uses two expressions : "In summa alius ne sit, at idem, quantum quoque modo suae linguae facultate et proprietate consequi posse" (p. 403).

44Not all languages show the same rigidity when it comes to the order of words in a sentence. Some are rather flexible while others allow no deviations. Generally speaking, inflected languages offer more variety than their opposites. This explains why the author of the Exercitium, for instance, gives no specific instructions in matters of word order to the pupil translating Latin into German. Evidently, the two systems were regarded to be compatible precisely because the inflectional endings of words give clear signals as to the function of a particular word in the structure.

45Surprisingly enough, Italian in the past displayed a good measure of flexibility despite its being uninflected. This is true at least of the examples cited by Fausto da Longiano in his treatise on translation. Thus the opening sentence of the Genesis which he translates from Greek is rendered with "In principio fece Dio il cielo e la terra," which is the word order in the original, but in Italian it displays an inversion of the verb and the subject. Similarly, Aristotle's famous saying "Omnes homines natura scire desiderant" is taken over without any changes : "Tutti gli uomini per natura sapere desiderano" (an inversion of two verbs). On the other hand, Cicero's sentence, "Malo virum, qui pecunia egeat, quam pecuniam, quae viro" structured as it is artificio compositionis, calls for a different treatment. Instead of his own, Fausto quotes Boccaccio's translation : "Voglio avanti huomo, che abbia bisogno di ricchezze, che richezza, che abbia bisogno d'huomo." He also justifies the repetition of "abbia bisogno" on grounds of language characteristics : "Riplico due volte il Boccaccio per levare la durezza de la voce 'egeat,' che nel Latino ella non era necessaria" (p. 26r).

46The situation changes radically when one turns to the French language. In the XVIth century, both Louis Meigret and Barthélemy Aneau point to the different word order in Latin and French. Aneau extols the French word order because it follows the ordo naturalis : "... il suit le droit ordre naturel, sans entremescler les dictions" (qtd. in Scaglione, p. 154, n. 19). Meigret discusses at some length this difference and brings an example of such "misplacements" in the Latin sentence, "Gallos vicit Caesar," which begins with a direct object. One should not overlook Meigret's remark that in Latin such a word order is used "pour parler elegamment" (p. 195).

47He questions the validity of such a statement on grounds of the natural word order. What meaning, he asks, can we get from a statement where the verb occupies the place reserved for another part thus going against the norms which the custom of our language has set, so that we can understand one another ("commune intelligence") ? To tell the truth, if we compare the Latin style with ours we shall find that it is contrary to what we usally do, so that if one looks at the "ordre de nature" we shall find our language ranking higher than Latin. The Latins completely reverse the natural order so that instead of placing the "agent" first, they put the "passive" first and the agent after the action : "... par ce moyen le passif qui par l'ordre de nature doit être le dernier en clause, est le premier en prolacion" (p. 195). This is a genuine case of putting the cart before the horse or, to use a French version, "la charrue devant les beufs" (qtd. in Scaglione, p. 154, n. 19).

48It is common knowledge that the battle against Latin and for French was fought with this same argument not only in the XVIth century but even in the late XVIIIth century when Rivarol, in his essay De l'universalité de la langue française claimed that French alone "par un privilège unique" preserved the natural order and the clarity deriving from it (Œuvres, p. 48-49). Scaglione who covers this long diatribe found only one solitary dissenting voice, Fénelon, who half a century before Rivarol saw in this rigid word order an impoverishment of the language : "on a appauvri, desséché, et gêné notre langue... " Indeed the presence of the strict sequence subject-verb-object leads to boredom (Aristotle's satietas) caused by the absence of variety : "C'est ce qui exclut toute suspension de l'esprit, toute attention, toute surprise, toute variété, et souvent toute magnifique cadence" (qtd. in Scaglione, p. 235).

49As a transition from theoretical argumentations to more practical considerations, I would like to bring to your attention a portion of the fourth letter from Batteux's work De la construction oratoire which appeared in 1763. It is the portion in which he discusses "latinismes de construction", i.e. some important points to keep in mind when transferring word order from Latin into French. He prefaces his discussion with a statement of principle concerning the first and foremost requirement in a translator, namely his thorough familiarity with the proprietates linguae of both languages : "La premiere chose nécessaire au Traducteur est de savoir à fond quel est le génie des deux langues qu'il veut manier" (p. 104).

50Just like the theoreticians mentioned above, Batteux favors structural fidelity : "toutes les constructions... doivent être conservées" (p. 107). However, contrary to previous practices which customarily dismantled the rhetorical periodus, reducing it to coordinated clauses, he insists that a periodus must be rendered as it is found in the original. As a demonstration he brings a periodus "d'une longueur extrême" (26 lines) taken from Cicero and adds his own translation. Unfortunately, its length does not allow a closer examination on our part. He also requires that features such as symmetry or antithesis be retained in translation. In case of physical impossibility, there is an obligation ("s'acquitter") to provide an equivalent substitute. Thus Cicero's statement : "Animi imperio, corporis servitio magis utimur" is rendered as ; "Nous nous servons de l'esprit pour commander, du corps pour obéir" (p. 114).

51More interesting is the second set of cases in which owing to the particular nature of the original, the translator is forced to depart from the original : "... il faut entièrement abandonner la manière du texte qu'on traduit" (p. 117). In this section he deals with the Latin transgressions against regimen mentioned earlier. He lists three principal instances in which in French one must adjust the sequence between the "mots régissant" and their "régis." The first instance is the well-known case of the object before the verb discussed above by Meigret Batteux employs bellum intulit as an example which he translates with : "il a porté la guerre." The second case concerns the position of an adjective. Latin, for instance, has furens bellua but French must postpone the adjective and say : "la bête furieuse." The third instance shows the frequent inversion of the "régissant" and its "régi" (genitive before the nominative) which French does not tolerate. Thus "urbis magnitudo" is to be rendered with "la grandeur de la ville" (p. 119). He concludes by stating the principle behind these changes. They belong to the proprietates which are sanctioned by custom : "Il est d'usage de suivre cet ordre dans la prose française" (p. 119).

52I cannot fail to mention an interesting element which Batteux requires in translation, an element which has not appeared in previous sources, namely the idea of "vivacité de sentiment" (liveliness, sprightliness, sparkle). He describes it as fire or luminosity :" tantôt c'est un feu qui brûle et qui éblouit ; tantôt c'est une lumière douce qui égaye et ne fatigue point (p. 120). At an earlier point in the same letter, Batteux offers an interesting example in which this "feu" comes into play. At issue is a passage taken from Cicero : "Neque potest is exercitum continere imperator, qui seipsum non continet" (p. 110).

53Batteux offers three translations. The first is as follows : 'Un général qui ne règle point sa conduite, ne peut régler une armée." The second is more literal : "Un général ne peut régler une armée, qui ne peut se régler lui-même." The latter detaches the relative clause from its antecedent and Batteux, therefore, regards as a "latinisme". The former version is correct but lacks the sprightliness ("feu") of the original. Batteux prefers the rendering of M. Fléchier, although he is forced to admit that the translator has failed to keep the order of ideas of the original. Here is Fléchier's rendering ; "Quelle discipline peut établir dans son camp celui qui ne peut régler sa conduite ?" In Batteux's eyes, this version has the meaning and also "le même feu." He concludes : "Ainsi l'exemple de M. Fléchier nous donne une double leçon" (p. 110).

54A surprising "leçon" if one takes into account Batteux's general rule stated earlier that "toutes les constructions... doivent être conservées." In Fléchier's version, an affirmative statement is turned into a rhetorical question, words of the original have been changed ("exercitum" is replaced by "camp"), new notions are added ("discipline"), others are removed ("imperator" becomes "celui"), others are amplified ("régler sa conduite"). By using the interrogative construction Fléchier even succeeded in placing the direct object before the subject. There is no denying, however, that the "feu" is there.

55Batteux's discussion and exemplification display the familiar sight found in studies on translation. A few selected fragments are taken out of context and used as evidence for the point at issue. As a departure from this practice, I have chosen as an example to be examined a coherent text in its entirety, namely the opening sentence of Guarna's Bellum grammaticale containing the laudatory description of the "Provincia Grammaticae". Despite its small size for a periodus, it is an interesting specimen in itself. What makes it particularly interesting for our discussion is the fact that, thanks to the research of Johannes Bolte, we possess a series of translations which are different in terms of the target language (three in French, and one each in Italian, English and German) and in terms of the century in which they were made (XVIth, XVIIth or XVIIIth century).

56Andreas Guarna opens his work with the following sentence :

Grammaticae provinciam ceterarum omnium, quae ubique terrarum fama celebres sunt, pulcherrimam felicissimamque esse nemo ferme vel mediocriter eruditus est qui ambiget : quippe cum amoenitate situs, salubritate caeli, frugum omnium copia rebusque aliis, quibus mortalis haec vita carere non debet, nulli cedat, tum illustrium virorum nutrix semper exstitit et alumna (p. 4).

57Let us first subject the original to a closer scrutiny. Though didactic in its purpose, the book is written in a medium rhetorical style with a generous sprinkling of "misplacements" of the natural order. Prominent is the consistent inversion of the genitive ("régis") and nominative ("régissant") case as in "grammaticae provinciam," "ceterarum omnium," "copia frugum omnium," "nutrix virorum illustrium" (in this latter case a postposition of the adjective). There is an inversion involving a pronoun in "mortalis haec vita." Other points worth mentioning are the synonymical doublet "nutrix et alumna" (from "alo", to nourish) in the concluding portion, and especially a classical example of a cursus velox at the very end : "exstitit et alumna." The passage is a periodus with two cola, one consisting of the proposition, the second of the demonstration. Each colon is subdivided into several subordinating relative clauses. The first colon is a concatenation of a principal and three dependent clauses arranged backwards (impedimentaliter) so that the subject ("nemo") is placed after several dependent clauses while an accusative with infinite opens the sentence. The connecting link "quippe" introduces the second colon with an ellipsis (the subject is missing). This colon is subdivided into two parallel clauses built around the particles "cum... tum."

58The translations are arranged in the chronological order and the orthography found in Bolte's introduction to his edition.

59a) An anonymous French translation made roughly forty years after the first edition (1546, published in Basel by Jacques Estauge from Lyon), La guerre des noms et desverbes :

Il n'est nul tant soit moyennement sauant, qui face doubte que de tous pays qui ont renomee, celuy de Grammaire ne soit le plus beau, et le plus heureux : tant pour la plaisance du lieu, du bon aer, et de l'abondance de tous fruictz et autres choses necessaires a ceste uie mortelle : que pour ce qu'il est nourisseur et entreteneur de toutes gens moult estimez (p. 26).

60b) An English translation made twenty years later in part from the preceding French translation (William Hayward, London, 1565), The Discourse of the Grammar Warrebetweene two Kings, the Noune and the Verbe, contending for the chiefe Place inOration, wherin thou shalt see the Verbe to prevayle :

There is none so simple in judgement, that doubteth the soyle of the grammar to be the fairest and happiest of all the renouned parts and provinces of the world, as well for the pleasantnesse of the place wherein it is situate, being in good and holsesome ayre, and aboundaunt in all fruites and other good things, without which this mortali lyfe may not easilye be passed : as also for that shee hath alwayes and ever beene the nourse and bringer up of all people of renoune (p. 29).

61c) A second French translation made roughly a century after the first edition (Pierre Roger, Paris, 1616) :

Il n'y a personne qui ayt tant soit peu d'erudition, qui ne sçache que la province de la Grammaire est une des plus belles, et des plus riches et heureuses qui soient en toute la terre habitable, quelques celebres et renommées que soient les autres, d'autant qu'elle ne cède à aucune, soit pour la beauté de la situation, soit pour la bonté de l'air et pour l'abondance de toutes sortes de fruitz et de toutes autres choses qui sont necessaires pour la vie de l'homme. Aussi a elle tousiours esté comme la mere nourrice de plusieurs grands et illustres personnages (p. 27).

62d) A German translation of the middle of the XVIIth century (Johann Buno, Dantzig, 1650) :

Dass das Land Grammatica unter allen Provintzen, so irgend in der Welt bekant seyn, die Schönste und Herrlichste, daran wird beynahe Keiner, der auch nur halb gelehrt, zweifeln. Sintemahl selbige, an Liebligkeit des Orts, Gesunder Luft, an Uberfluss allerley Früchte, und allen andern Dingen, deren das Menschliche Leben nicht entbehren kan, Keiner etwas nachgibt, als auch an Berühmten und Gelärten Leuten, welche Sie jederzeit gehäget und erzogen.

63e) An Italian translation a whole century later (Angiol Maria Ricci, 1741), La guerra grammaticale :

Non vi ha alcuno anche mezzanamente ammaestrato, che metta in dubbio essere la provincia della Grammatica, di tutte l'altre, che per fama chiare sono in tutta quanta la terra, la più bella, e la più benavventurata. Imperciocché a nessuna cedendo per l'amenità del posto, per la salubrità dell'aria, e per la dovizia di tutte le cose, senza delle quali star non dee questa mortal vita, fu eziandio d'uomini illustri madre sempre, e nutrice (p. 31).

64f) A French translation in a bilingual edition at the beginning of the 19th century (H. B. Gibaut, Poitiers, 1811), La guerre grammaticale entre les deux rois le Nom et le Verbe, disputant de la prééminence dans le discours :

La province de Grammaire est située dans la plus belle et la plus heureuse de toutes les contrées ; c'est un fait connu des moins instruits. Elle ne le cède à aucun pays du monde, soit par l'agrément de ses sites, soit par la pureté de son ciel, la fertilité et l'heureuse abondance de toutes choses nécessaires dans cette vie mortelle. Au sein de cette terre favorisée furent nourris et élevés les hommes qu'a célébrés la renommée ... (p. 28).

65Comparing the ways the structure of the periodus has been transferred into the target language, one cannot overlook the fact that only German, an inflected language, is in the position of preserving the original architecture. The translator begins with the dependent clause having first resolved the accusative with infinitive construction. Just as in Latin, the subject and the principal clause occupy the last part of the first colon. The connective link introduces the second colon which shows a few modifications in the wording (the "nutrix et alumna" are changed into the verbs "gehäget und erzogen"). There is also a change in the sentence structure in the last part of the colon, the "tum" section, which is no longer an independent clause with a finite verb ("exstitit") but one dependent on the verb "nachgeben" ("cedit"). This modification compels the translator to use the dative plural ("berühmten und gelärten Leuten") to replace the genitive of the original ("illustrium virorum").

66By contrast, all the other languages had to completely restructure the first part by putting the subject at the beginning. Italian and English use an accusative and infinitive, whereas French must resort to a resolution. These two languages introduce participial or gerund construction ("being..." and "cedendo...") not present in the original. German and Italian link between the two cola with "sintemahl" and "imperciocchè." It is worthy of note that all translators have reversed the Latin anastrophe consistently employed by Guarna and placed the genitive after the noun on which it depends.

67The parallel formation "cum ... tum" is preserved at least in part by all translators. The entire structure is found in English with "as well... as" and in the 1546 French translation with "tant... que." German and Italian have only the second particle with "als auch" and "eziandio." The 1616 French translation, having used "soit" twice in the first part, preferred to place the "tum" section as a separate principal clause introduced by "aussi."

68Other minor changes were caused by lexical deficiencies of the target language such as a noun equivalent of "salubritas" in German and English. Both languages resort to adjectives "good and holsesome"(!) and "gesunder." Except for German, all the remaining languages lack a verb equivalent for "carere" which called for alternative solutions. French used a relative clause "qui sont necessaires", English a passive construction "may not easyly be passed," and Italian used a verb "stare."

69This short excursion into the practical field of translating was a salutary experience. It shows the variety of ways in which the same text can be translated even in one and the same language and by the same token reveals the great reservoir of expedients a language puts in the hands of the translator. The important condition is, of course, that these expedients be not only known but also properly employed. Each translator puts his own stamp on his rendering.

70Among the translations just seen, the quality of the achievements varies but there is no evidence that the quality improves as time passes and languages progress. A tangible example is provided by Gibaut's translation published in a bilingual edition in 1802 approximately two centuries and a half after the first French version. We have until now made no references to this translation because it is so different from the others.

71Unlike all previous translations using uninflected target languages, Gibaut's version begins with the words found in the original ("grammaticae provinciam") but he changed the accusative to a nominative case. It is the first sign of the unprofessional approach employed throughout the sentence, unprofessional in the light of the generally accepted rule that the sentence structure should not be tampered with except in cases of incompatible structural patterns between the two languages.

72Gibaut has completely disarranged the complex structure of the original periodus and replaced it with a series of simple sentences which have little in common with Guarna's rhetorical ornamentation. Furthermore, his version shows additions and deletions when compared with Guarna's sentence. What surprises most is the fact that this translation appeared in an edition in which the original was available to the reader on the opposite page. We do not know the reasons for Gibaut's approach but we are justly assuming that he knew the French language enough to do justice to the original.

73Even though poor in its quality, this translation is valuable because it brings to our attention an important factor in translating, a factor which has until now received little notice, namely the translator himself. In my book on translation, I pointed out that Humphrey seems to be the only Latin theoretician to touch upon this aspect when he employs the phrase "quantum in nobis est" (p. 479). The obvious interpretation of this expression would be to regard it as a reference to the translator's natural talent, his preparation and his skill. However, the phrase can be a reference to an additional quality listed by Humphrey among the virtutes of a translator.

74Even though last, this virtue is by no means the least. On the contrary, the author devotes twenty pages (pp. 180-200) to singing its praise and inculcating its importance in translating. The quality is called diligentia and, according to Humphrey's synopsis, affects three aspects of the translator's activity, namely the choice of words ("vis verborum"), the integrity of the content ("nihil excidat," "nihil addendum") and the retention of the sentence structure ("ordo non transponendum").

75Humphrey showers diligentia with page-long praises. Among other things, he declares it to be the indispensable key to success not only in translation but in any activity :"... nec sine hac quidque unquam praeclarum et magnum effeceris." (p. 180) At one point of his eulogy, Humphrey makes use of a rhetorical question which is worth remembering especially after reading Gibaut's translation. It is also worthy of serving as the conclusion to our considerations.

76What is the use, he asks, of possessing other virtues, no matter how exceptional and distinguished they may be, if the stimulus and the spur ("calcar"), as it were, of "diligentia" is absent : "Quorsum enim virtutes caeterae, quantumvis eximiae et insignes, si non stimulus hic et quasi calcar adsit diligentiae ?" (p. 180). How true ! Even the most thorough knowledge of the "genie des deux langues" and the presence of other eminent professional qualifications are of little use unless they are coupled with a conscientious effort and personal committment on the part of the translator.

77Exercitium puerile grammaticale Ed. Johannes Müller in : Quelleschriften und Geschichte des deutschsprachigen Unterrichts bis zur Mitte des 16. Jahrhunderts. Gotha : Thienemann, 1882.

Haut de page


ARISTOTLE. De arte poetica. Tr. I. Bywater, ed. T. E Dorsch, Harmondsworth : Penguin Classics, 1965.

BATTEUX, Charles. De la construction oratoire. Paris : Desaint et Saillant, 1763.

BOLTE, Johannes. Andrea Guarnas Bellum Grammaticale und seine Nachahmungen. Monumenta Germaniae Paedagogica, v. 43.

FAUSTO DA LONGIANO, Sebastiano. Del modo di tradurre d'una lingua in altra secondo le regole mostrate da Cicerone. Venice : Griffio, 1556.

GOCLENIUS, Rodolphus. Lexicon philosophicum. Frankfurt : Becker, 1613.

HUET, Petrus D. De interpretatione libri duo. The Hague : Arnold Lee, 1683.

HUMPHREY, Lawrence. Interpretatio linguarum seu de ratione convertendi et explicandi auctores tam sacros quam prophanos. Basel : Frobenius, 1559.

MEIGRET, Louis. Le Tretté de la grammere françoeze. Ed. Wendelin Foerster, Heilbronn : 1888.

RIVAROL, Antoine. Œuvres complètes. Paris : Collin 1808.

RENER, Frederick M. Interpretatio : Language and Translation from Cicero to Tytler. Approaches to Translation Studies, v. 8, Amsterdam-Atlanta, 1989.

SCAGLIONE, Aldo. The Classical Theory of Composition from its Origin to the Present.. University of North Carolina Studies in Comparative Literature, v. 53., Chapel Hill, North Carolina ; University of North Carolina Press, 1972.

SEBILLET, Thomas. Art poétique francoyse (1555). Ed. F. Gaiffe. Société des textes français modernes, v. 14., Paris : Cornely, 1910.

SONKOWSKY, Robert P. "A Fifteenth-Century Rhetorical Opusculum." Festschrift in Honor of Berthold L. Ullman. Ed. Charles Henderson Jr., Rome : Edizioni di storia e letteratura, 1964. v. 2, pp.259-281.

TESAURO, Emanuele. Il cannochiale aristotelico. Ed. A. Buck., Bad Homburg : Gehlen, 1968.

VIVES, Johannes Lodovicus. Opera omnia. Ed. G. Majarans, Valencia : Monfort, 1782.

Haut de page

Document annexe

Haut de page

Pour citer cet article

Référence papier

Frederick M. Rener, « In Principio Erat Verbum :Traditional Principles Concerning Words and Their Order in Translating », Palimpsestes, 7 | 1993, 15-35.

Référence électronique

Frederick M. Rener, « In Principio Erat Verbum :Traditional Principles Concerning Words and Their Order in Translating », Palimpsestes [En ligne], 7 | 1993, mis en ligne le 23 décembre 2010, consulté le 21 mai 2018. URL : ; DOI : 10.4000/palimpsestes.633

Haut de page

Droits d’auteur

Tous droits réservés

Haut de page
  • Logo Presses Sorbonne Nouvelle
  • OpenEdition Journals