Introduction

1Of the many articles that have been written over the years on the subject of legal translation, only a few address the history of legal translation.1 This lack of interest is surprising since legal translation predates even Bible translation. For example, it is generally accepted that "the oldest known recorded evidence of legal translation is the Egyptian-Hittite Peace Treaty of 1271 BC" (Sarcevic 1997, 23). It would be both impossible and futile to try to pinpoint the reasons for this lack of interest. One of its consequences, however, is that many false or misleading statements have been made about legal translation through the ages. This article will discuss one of the most persistent and pervasive of those myths, which claims that the legal text was translated very literally for many centuries because of its authoritative status. Sarcevic, one of the foremost scholars in the field of legal translation, states that until the seventeenth century translations were "strict literal translations" and that only in the eighteenth and nineteenth centuries did translations become "literal translations."2 She provides the following figure on the evolution of legal translation (Sarcevic 1997, 31–32):

SARCEVIC'S MODEL OF THE EVOLUTION OF LEGAL TRANSLATION

2However, Sarcevic does not take into account the idea that legal translation is only possible through the appropriation and reinvention of a legal text according to the period in which the text is translated. In this article, I will argue that the act of translation is closely bound to the translational goal of the translator, to the period in which the text is translated, and to the legal culture3 to which the text is translated and transferred, not to the authoritative status given to the original text. To support my argument, I will examine the possible origins of the myth of literalness in legal translation and its reformulation in the twentieth century. I will then provide examples from a translation of the Institutes done in thirteenth-century France.

Possible origin of the myth

3The Corpus Iuris Civilis was compiled in the Byzantine Empire in the second half of the sixth century under the auspices of Justinian I (482–565). This compilation of Roman law consists of five separate texts: the Codex constitutionum, the Digest or Pandectes, the Institutes, the Codex Repetitae Praelectionis and the Novellae Constitutiones. In the proeminium of the Digest (Constitutio Tanta 21), Justinian prohibited anyone from appending "any commentary to these laws [Digest], save only insofar as he may wish to translate them into the Greek language in the same order and sequence as those in which the Roman words are written (kata poda, as the Greek call it)" (Watson).4 This restriction on the translation of the Digest, which is considered to be the oldest known written rule restricting the type of translation to be used when translating a legal text, had various consequences through the ages. For example, Stein states that Constitutio Tanta 21 was the main reason why the Corpus attracted "relatively little attention" (Stein 1999, 35) in the Byzantine Empire, since few Byzantine jurists could read and understand Latin well enough to make use of the compilation.

4This regulation on translating the Digest seems to have contributed to the myth of literalness in legal translation in the twentieth and the twenty-first centuries. For instance, Sarcevic quotes Constitutio Tanta 21, explaining its inclusion in the Digest by the influence of the literalness topos used in Bible translation. According to Sarcevic, literalness in legal translation results from the authoritative status given to legal texts, a status similar to that given to the Bible. Because legal texts are authoritative documents, they were translated literally (Sarcevic 1997, 24–25). Furthermore, Sarcevic, following a short review of various articles and studies, states that legal texts were translated literally until at least the beginning of the seventeenth century; that the relationship between translator, "author," and public has been sterile for close to two millennia; and, finally, that the first "challenge to the literal translation of legal texts did not come until the twentieth century"5 (Sarcevic 23ff). Notwithstanding the importance or value of that research or the influence of Tanta 21, the studies so far on the practice and evolution of legal translation through the ages are too few and far between to permit general statements to be made on the methods used in translating legal documents or on the evolution of legal translation.

Corpus Iuris Civilis and its authority during the Middle Ages

5As stated above, Sarcevic explains the use of extremely literal translation techniques by the authoritative status given to legal texts. The inherently normative or authoritative aspect of this kind of text should, according to Sarcevic, have contributed to the trend of strict literal legal translation. If her reasoning is valid, all translations of legal texts done before the twentieth century should, regardless of the period in which the texts were translated or the translational goal of the translator, be either literal or strict literal translations. Since the thirteenth-century translation that will be examined in this article is a translation of a text contained in the Corpus Iuris Civilis, the status of this legal compilation during the Middle Ages must now be described.

6The Corpus Iuris Civilis was compiled in the Byzantine Empire at the end of the sixth century. It was "rediscovered" by the West at the end of the eleventh century. By the twelfth century, the Corpus was considered in the West to be an auctoritas6 in the abstract sense of the word. It was extensively studied and glossed by the Glossator and Post-Glossator Schools. Various glossators compared the Corpus to the Scriptures. It was regarded as sacred and was given almost Biblical authority.7 The glossators called the Corpus "sanctio sancta," "sacratissimae leges," and "donum Dei." One glossator writes that the compilation is "a divine precept coming from the mouth of the princes [des préceptes divins exprimés par la bouche des princes; our translation]," that "the Holy Spirit has spoken through them [que l'Esprit Saint a parlé en elles; our translation]," and that the compilers of the Corpus "had received power from God... so one could say that the laws were made by God [ont reçu pouvoir de Dieu... et qu'on peut donc dire que les lois ont été faites par Dieu; our translation]" (Thireau 1993, 18).

7The Corpus Iuris Civilis was considered to be an authoritative text during the Middle Ages. Therefore, if Sarcevic is correct in considering the use of literalness in legal translation to be due to the authoritative status of legal texts, then the various translations made of the Corpus Iuris Civilis during the Middle Ages should all have been strict literal translations or at least literal translations. In order to test Sarcevic's conclusion, I will now analyze the methods and techniques used in translating the Institutes.

Verse translation by Richard d'Annebaut

8In this section, I will analyze certain aspects of a translation of the Institutes that is a cas d'espèce8 from thirteenth-century France. It is the verse translation of the Institutes9 attributed to Richard d'Annebaut. The translation was finished in 1280 according to its epilogue; it is written in verse, and contains approximately 24,000 lines, including an eighty-four-line prologue and a forty-line epilogue presumably written by the translator. We have only one manuscript of this translation, Harley 4477.2, and one incunabulum printed in Paris in 1485 by Antoine Cayllaut. This verse translation has never been edited.

9Here I will examine some of the most frequent translation strategies and techniques used by d'Annebaut in order to determine if, regardless of his reasons for translating the Institutes, he translates it literally or extremely literally because of the authoritative status this legal textbook had during the Middle Ages. If this is the case, then Sarcevic's reasoning and conclusions are correct. Three elements of d'Annebaut's translation will be examined: 1. the prologue and epilogue; 2. the use of synonyms to translate certain legal terms; and 3. the translation strategies and techniques used by d'Annebaut in three sections of the Institutes.

Prologue and epilogue10

10D'Annebaut in his prologue does not broach the literal or free translation topoi. He simply states that he will translate the Institutes in verse form (Si les translateray en rime / Ou consonant ou leonine [17–18]). He further states that his translation is done to educate a schoolboy, Bertrand d'Escalphepié, and that he hopes that it will help the boy to learn and understand Latin (A commencier ceste besoigne / Ne met ung enfant de gascogne / Qui m'est ballie a introduyre / Et a ensaigner et a duyre / Et a tenir lay bien soubz pie. / Bertran a nom Deschalphepie / Frere est Raymont qui les se veut. / Se il y veult garder suvent / Il y pourra asses apprendre / Et plus legierement entendre / Le latin quant il le verra / Et trouver ce que il querra. [21–32]). Furthermore, d'Annebaut explains in the epilogue that students will be able to collate the French and Latin ([faire] collation / de françois contre le latin [34–35]), enabling them to understand what their teachers are telling them, which will be all in Latin (ce que les maitres lor diront / Que tout en latin lor lirons [39–40]).

11What can be inferred from d'Annebaut's comments? First, by referring to the Latin original, he positions his translation as a service text to the original and to its auctoritas. Secondly, d'Annebaut does not specify in either his prologue or his epilogue why he versifies the Institutes. He does state that he translates to help a young man understand the Latin text. From what d'Annebaut writes, it is logical to assume that he translated the text for pedagogical reasons and sees his translation as a pedagogical tool.11 In this sense, d'Annebaut can be included in the category that Peter Dembowsky calls service-translator (Dembowsky 1986) — one who brings attention to the dependence of the translation on the original and to its didactic intention. It is therefore probable that d'Annebaut adapted his translation techniques and strategies to the translational goal he claims to pursue, namely, to translate for "a child of Gascogne [un enfant de Gascogne; our translation]." Furthermore, the combination of the versification process imposed on the text and the translational goal pursued by the translator should have resulted in a translation that is far from being strictly literal or literal as defined by Sarcevic because of the metrical and pedagogical constraints the translator imposes on his translation work.

Use of synonyms12

12One of the pedagogical strategies favoured by d'Annebaut is the use of synonyms when translating Latin terminology. The first example is the translation of the Latin expression ab intestat, meaning "without a will" (Section 3.1.7). D'Annebaut, in his translation, alternates between using a calque and gallicizing the Latin expression. This ambivalence is partly due to the versification of the text. When a calque facilitates versification without hampering the transfer of content, he uses the Latin expression: Que pour intestat est tenu / Cil dont l'heritage est venu. The same rule applies to gallicizing a term: in the very same section (3.1.7), he translates the Latin by Qu'ilz fut celui certainement / Qui cil mourut sans testament. This translation strategy could confuse the reader; however, it does have the advantage of teaching the reader the Latin equivalent for the French translation and of referring him or her, indirectly, to the auctoritas of the Latin.

13Here is another example. Section 33d (4.6.33d) states that when a person takes a stipulation from someone else that gives him or her a choice between two things (for example, between a slave or an amount of money), and the claimant then asks specifically for one of the things promised, this constitutes an overclaim. This is because the person taking the stipulation no longer has a choice. Section 33d then goes on to give some examples of overclaiming.

14The first example is as follows: si quis generaliter hominem stipulatus sit et specialiter Stichum petat [having stipulated generally for a slave, he claims Stichus in particular]. D'Annebaut, translates this passage by: Aucun fait que l'en luy promete / Ung serf qui mencion n'est faicte / Ne de son nom ne de son prix / Ne quel mestier il a apris / Mais ung serf tout generalment / Cil demande especialment / Tel comme il le veult avoir / De cestluy devon nous savoir / Qu'il sourdemande apertement. In the first line, we see that d'Annebaut translates stipulatus by que l'en luy promete. However, at the beginning of Section 33d he translates stipulatus by convenant. Once again he destabilizes the text by multiplying the number of equivalents for the same Latin word. The third and fourth lines (Ne de son nom ne de son prix / Ne quel mestier il a apris) are an overtranslation, but they do have a certain academic flavour: all good teachers repeat or emphasize what they consider to be the important elements. Furthermore, he does not transfer the Latin name "Stichus"; he simply explains it with the words "his name." Finally, the last two lines (De cestluy devon nous savoir / Qu'il sourdemande apertement) exemplify the rule of law explained in Section 33d. One problem in all versified translation, medieval or modern, arises from metrical constraints. D'Annebaut uses a formula to fill out line 8, for example: De cestluy devon nous savoir. We find throughout his translation other similar formulae: Mes nous devons cecy entendre; Que nous entendons à traiter; L'en doit savoir certainement; De ce devon estre certain; Nous devon savoir cy endroit; Et nous devon ici entendre; L'en doit en verite savoir. Such formulae attract the reader's attention and are one of the characteristics of d'Annebaut's translation. Of course, this formulaic writing is normally a rhetorical habit devoid of meaning. However, in d'Annebaut's case, it authorizes his direct involvement in the text, permitting him to guide his reader. He comes between the source text and the translated text in calling attention to something he finds important. Thus, we can hear the voice of the translator changing the register of the text.

15The other two examples given in Section 33d are aut generaliter vinum stipulatus, specialiter Campanum petat, aut generaliter purpuram stipulatus sit, deinde specialiter Tyriam petat [he stipulates for wine and specifically claims Campanian, or he stipulates for purple and claims Tyrian]. D'Annebaut's translation of this passage is Et autressy est il vrayement / S'il a de sa paine mise / Tant que pourpre luy est promise / Sans nommer le lieu ni la terre / Ou il la convient aller querre / Ne la valeur de la tainture / Cil demande par aventure / Telle de selle ou de sest / Ce est bien sourdemande faite." The first thing that stands out is that he does not translate the reference to wine. He prefers to put all the emphasis on purple dye. The second line (S'il a de sa paine mise) has a connotation that is absent from the Latin, namely, that purple is given only to those who work hard. Also, instead of translating the word Tyrian, he prefers to spell out the meaning and even elaborate on it by explaining to his reader that, when purple is promised to someone without any details as to its origin or its quality, he cannot specify which kind of purple is to be given. The last line (Ce est bien sourdemande faite) repeats the rule of law. Here we can see that d'Annebaut transfers the general idea present in the Latin, but intervenes in the translation in order to explain it more fully to his reader, in keeping with his translational goal.

Analysis of sections 1.1 (introductory paragraph), 1.1.1 and 4.18.4

16I have reproduced the following three short sections of the Institutes in order to give a general idea of d'Annebaut's translation methods and strategies, and to determine if he in fact translates the Latin text literally. The first two sections define the terms Iusticia (justice) and Iuris prudentia (jurisprudence).

D'ANNEBAUT, FIRST EXCERPT

17In these two sections, Iusticia is defined as the constant and perpetual desire to give each individual the rights to which he or she is entitled, while iuris prudentia is the science and philosophy of law in acquiring the knowledge of things divine and human, and the science of justice and injustice. Both Latin definitions are general, abstract, and philosophical in nature. D'Annebaut succeeds in translating the general idea contained in the Latin. However, his translation of the introductory paragraph puts into concrete form the rule expressed in abstract fashion in the Latin text. Furthermore, the versification process imposed on the text results in a glissement de sens (a semantic shift) by shifting the emphasis to different aspects of the text. D'Annebaut begins by specifying that justice is the stable (estable), steadfast (ferme) and perpetual (pardurable) desire that each person have without delay (sans attendue), his own thing (chose) returned to him. The use by the translator of a synonymous binomial combined with the repetition of the term volente (desire) in line 26 (Volente ferme et pardurable) puts the emphasis on the steadfast quality of the desire. Furthermore, d'Annebaut commits a slight glissement de sens when he translates the Latin ius suum cuique tribunes by Que a chacun sans attendue / Soit sa propre chose rendue. His translation is a concrete rendition of the abstract idea expressed in the Latin. The reason for this shift from a theoretical viewpoint to a practical one seems to be the metrical constraints the translator has imposed on himself. A result of this shift is a change in register from a theoretical and philosophical definition in the Latin to a practical definition applied to the practice of law in the Old French. The idea in both languages stays the same ("justice must be rendered"), but the perspective changes.

18D'Annebaut's translation of Section 1.1.1 reveals some of the effects of versification on the text. One can see that the translation is more prolix than the original. The original has twelve words while the translation has thirty-four words. By removing the formulaic writing used by d'Annebaut in order to respect the rhyme scheme and to draw the reader's attention to aspects that the translator considers important, one can see that d'Annebaut transfers the two elements found in the definition of iuris prudentia:

D'ANNEBAUT, SECOND EXCERPT

19Of the thirty-four words in his original translation of this section, twenty-one remain once the formulaic writing is removed — only nine more than the Latin original, which is normal considering that Latin is a very economical language. This simple exercise demonstrates quite clearly that d'Annebaut's translation is far from being a strict or a literal translation of the Latin original. He makes quite an extensive use of formulaic writing in lines 29 (sans doubtance), 30 (Que homme ait en soy pourvance), 31 (comme certaines) and 33 (congnoistre adroit). The use by d'Annebaut of formulaic writing in this section seems to have two main goals. The first one is to respect metrical constraints; the second is to draw the reader's attention to certain aspects he considers important. The last words of line 29 (sans doubtance "without any doubt") draw the reader's attention to something the translator considers important, while lines 30 and 31 repeat the idea that a wise person must know the meaning of the law (Que homme ait en soy pourvance / De congnoistre comme certaines: "That a man must be sufficiently wise / To know as a fact"). Finally, the last two words in line 33 repeat, once again, the idea that a person must know the law. Here the translator uses the term adroit, which allows him to respect metrical constraints, but at the same time he brings the reader's attention to the idea of law/rights by using the préfixe d'intensité "a" + droit. D'Annebaut's use of formulaic writing authorizes his direct involvement in the text, permitting him to guide his reader. He comes between the source text and the translated text by calling attention to something he finds important in the text. Here, d'Annebaut becomes a teacher.

20The last example comes from Title 18 of Book 4, which deals with criminal trials. Section 4 describes the punishment reserved for anyone committing the crimes of adultery, of homosexuality or of the non-violent seduction of an unmarried girl or respectable widow.

D'ANNEBAUT, THIRD EXCERPT

21The first point to note is that, once again, the Old French translation is far more prolix then the Latin original. The former has one hundred and forty-two words, the latter sixty-two. This wordiness is caused by the use of formulaic writing and by the need to respect the rhyme scheme. Lines 19, 25, and 30 are good examples of the effects of versification on the text (Encontre le salut de l'ame / Se ilz estoient par verite / Qui n'estoient pas de grans merites). The second point is that d'Annebaut does not translate the title lex Iulia de adulteriis or the sanction imposed on anyone committing this crime (gladio punit). In both cases, he refers his reader to his translation of the previous section, which gives the title of the law (Julius) and the punishment (Que il en doit mort soustenir).13

22Lines 8 to 10 refer to the crime of adultery (Et non pas ceux tant seulement / Qui autry mariage brisent / Et estrange femme ravisent). The translation explains that the law will punish those who break another's marriage by ravishing the woman or wife. Here we can see that the translator takes for granted that only a man will commit the crime of adultery! The Latin uses the pronoun quae (qui) in the nominative neutral form, which does not indicate the gender of the adulterer. It is difficult to determine exactly what caused this translation error. D'Annebaut could have used the term person instead of femme. Maybe he was tired or let his belief in the purity of women cloud his judgment. Not all translation mistakes can be explained.

23The second crime addressed in Section 4.18.4 is homosexuality. The Latin defines the crime of homosexuality in this way: qui cum masculis infandam libidinem exercere audent [those who dare to indulge their unspeakable lust with males]. D'Annebaut's translation is highly explicit. His translation is Mais contre ceulx qui la droiture / Corrompent de droite nature / Mais qui prennent hommes pour femmes / En destruisement de leur ames. This translation is quite vivid and the imagery used by d'Annebaut leaves absolutely no doubt as to the exact nature of the crime. He specifies that those who take a man as one would a woman will lose their soul. He therefore adds the sense that, homosexuality is not only a crime in secular law but also a crime in the eyes of God. This connotation is not in the Latin text. We can see that d'Annebaut, in this excerpt, intervenes in the text. He explains it and makes sure that there can be no possible misinterpretation by the reader.

24The last crime covered in Section 4.18.4 is the seduction by a man of an unmarried girl or widow. D'Annebaut stays relatively close to the Latin. For example, he translates flagitium (rape, adultery, and so forth) by meffait de fornication (17) and stupraverit (to bring dishonour to, to tarnish) by corrompte (20). However, line 19 (Encontre le salut de l'ame) is an ajout du traducteur. This line rhymes with line 20, which is probably the reason why d'Annebaut added it. It gives a definite moralistic flavour to the translation that is absent from the original. Once again, the voice of the translator is heard in the translation.

Conclusion

25What can be inferred from the various examples? First and foremost, d'Annebaut's translation of the Institutes is not a strict literal or even literal translation of the Latin as defined by Sarcevic. He adds or removes various elements in order to adapt the text to its intended readership. In this sense, by translating the Institutes in Old French and in verse form, he reinvents and appropriates it. He invents a versified legal language in Old French, which forces him to change the register of the text. He usually succeeds in transferring the general meaning of the Latin text, but he changes the formulation and lays greater emphasis on certain aspects. It would appear that for him the transfer of the content of the Latin text is not conditional on the transfer into Old French of the linguistic form of the original. The consequences of this action are to adapt the Institutes to verse form, to destabilize the text and to add to its network of meanings.

26Secondly, the relation between d'Annebaut, his translation, his readership, and the "author" of the Institutes cannot be characterized as sterile. He intervenes frequently in the text by using various techniques. He guides his reader, reminding him of what he considers important and explaining it when he feels the need. By choosing to versify the Institutes, he destabilizes the text and reinvents it to correspond to his translational goal, which is to educate a child. He has an interventionist approach to translation, which frees him from the Latin text.

27On a more general note, it can be inferred from this short analysis of d'Annebaut's translation that the history of legal translation still needs to be written. This specific translation is not a literal or a strict literal translation. However, too little is known about the various translational strategies and techniques used through the ages in translating legal texts to be able to make general statements on the evolution of legal translation. A comprehensive study of the development of legal translation must still be done.