JOURNAL OF ENGLISH STUDIES- VOLUME 3, (2001-2), 217-236 CHAUCER’S “MAKYNG ” OF THE ROMAUNT OF THE ROSE

Despite its being the first testimony of Chaucer's genius, the interest of modern criticism in the Romaunt has mainly focused on the issue of authorship, whereas the efforts to assess this text as a translation have been limited both in their number and in their scope. This paper discusses Chaucer's translation of the Roman de la Rose, and provides an evaluation of Fragment A from a modern traductological perspective, while taking account of contemporary theoretical positions. First, this article compares the Romaunt with Chaucer's later translating practice. Second, taking into account that the immediate audience of the Romaunt would have been cognizant of French, this essay considers the pragmatic function of this translation. Finally, I reconstruct some of Chaucer's decisions in the translation process, and then I present the translation strategies he adopted in order to create an English metapoem which replicated the spirit of the Roman, thus proving the adequacy of English for poetic expression

"Chaucer's translation of the Roman de la Rose represents his first significant literary endeavor" 1 . Undertaking a poetical translation of the French original must have been a challenging project if we take into consideration the situation of the English language and literature in the second half of the fourteenth century: the contact of English with French since 1066 had promoted a unidirectional movement of literary imitation by English writers of 1. A shortened version of this article was read at the 36th International Congress on Medieval Studies, Western Michigan University, May 3-6, 2001. continental themes, which, however, fostered the equality neither of the English literary tradition with the French, nor of the status of their respective languages. Chaucer, therefore, found himself in a somewhat mediocre literary tradition, and with a language that, lacking the prestige of French and Latin, was not adequately equipped for the expression of "sentement".
The principal interest of modern criticism in the Romaunt has consisted in determining the extent of Chaucer's participation in its production, 2 while the efforts to assess this text as a translation have been rather limited both in their number and in their scope (see Dahlberg 1999: 27-32). This essay proposes to evaluate Chaucer's Romaunt from a modern traductological approach by looking at translation as process and not as product (Djordjević 2000: 12), while still taking into consideration contemporary views.
The only contribution that discusses the intrinsic value of the Romaunt is an article by Caroline D. Eckhardt (1984), who also complains about the scant attention paid to Chaucer's text. Eckhardt's conclusions have been generally accepted as the standard reading of the translation, although her article is not free of methodological errors. First, she makes a brief reference to the theoretical formulations on translation by Saint Jerome and Dryden, but later she fails to apply these descriptive theories to Chaucer's work (cf. Dahlberg 1999: 30). Next, Eckhardt proceeds to compare the Romaunt to the French text with the purpose of observing, "where the translation is accurate, how that accuracy is achieved and, where it is not, what the value of the departures might be" (Eckhardt 1984: 46); that is to say, Eckhardt is committed to examining the Romaunt as a literary product, without taking into account the causes that determined or the conditions that influenced such a product.
For an appropriate understanding and evaluation of the literary implications of this translation in the development of Chaucer as a poet, we have to consider not only its relation to the source text, but also the process that has generated this literary product. I subscribe to Basil Hatim and Ian Mason's view (1990: 4-5) that "the resulting translated text is to be seen as evidence of a transaction, a means of retracing the pathways of the translator's decision-making procedures". Obviously the reconstruction of the translation process and of all the factors intervening in the "makyng" of the Romaunt cannot be absolute, but its investigation will foster a better appreciation of the text than the "product-to-product" comparison. 3 JORDI SÁNCHEZ MARTÍ 2. The issue of the authorship of the translation is not completely settled, yet consensus has been reached in attributing to Chaucer Fragment A, and in considering not improbable that he was the author of Fragment C. Dahlberg (1999: 3-24) offers a detailed revision of the positions held by the critics in relation to the authorship of the Romaunt. This paper concerns itself only with Fragment A.
The adoption of modern postulates for the assessment of Chaucer's translation in place of the contemporary traductological tenets requires a justification. The translation of "olde bokes" was constant throughout Chaucer's literary career. Although Chaucer never produced a treatise on translation wherein he exposed his views on this issue, we find some allusions to translation scattered in his works that give us an idea of his views. All his comments agree in adopting the Hieronymic posture condensed in the quote "non uerbum e uerbo, sed sensum exprimere de sensu" (Jerome 1910: 508). Chaucer, therefore, expressed his commitment to preserve the original meaning, as for example in Troilus and Criseyde: And of his song naught only the sentence, As writ myn auctour called Lollius, But pleinly, save oure tonges difference, I dar wel seyn, in al, that Troilus As in his song, loo, every word right thus As I shal seyn... (I.393-98) 4 Nonetheless, this promise has only declarative value, since the adoption of Saint Jerome's view became axiomatic in the Middle Ages. In fact, "none of the works in which he (i.e. Chaucer) claims to have adopted the Hieronymic posture can be considered unequivocally a sense-for-sense translation" (Machan 1989: 57). In general, the violation of the Hieronymic principle comes from the imposition of the translator's intentio upon the sense of the author or intentio auctoris, and it results in the manipulation of the materia that Chaucer found in his sources, mainly by the interpolation of glosses. The effect of this hermeneutical intervention is the effacement of the original text and its substitution by a new version that has absorbed the original matter and has cast it in the mould of a new intentio auctoris. 5 The fusion of exegetical translation and the manipulation of the intentio auctoris characterizes Chaucer's creative principle and his attitude toward the "olde bokes"; as Shoaf (1979: 64) indicates, "Chaucer plowed under the old fields of poetry to prepare for the harvest of the new science".
Despite the deviation from the original that this practice implies, the resulting texts were perceived as legitimate translations. From a modern perspective, however, these Chaucerian creations would be considered adaptations or versions at most, due to the prescriptive element attached to the 4. Cf. The Canterbury Tales, VII.961-64, VIII.78-84. All citations from Chaucer are from Benson et al. (1987), unless otherwise noticed. 5. Copeland (1991, esp. pp. 186-202) proves how the exegetical practice of academic discourse was adopted in vernacular translations, and Chaucer was no exception. concept of translation. For instance, one of the seminal works in the field of translation studies provides the following definition: "Translating consists in reproducing in the receptor language the closest natural equivalent of the source-language message, first in terms of meaning and secondly in terms of style" (Nida and Taber 1969: 12). The insistence on the translation equivalence prevents us from approaching the Chaucerian versions from a present traductological stand.
Nonetheless, the case of the Romaunt is rather different from the rest of Chaucer's translations. His purpose here was to render the original faithfully both linguistically and stylistically, thus displaying a clearly modern approach. Unfortunately the Romaunt does not include a prologue where the translator describes his purpose and the method of his translation. Yet the Prologue to the Legend of Good Women provides us with a relevant testimony uttered by the God of Love: "For thow," quod he, "art therto nothing able. Yt is my relyke, digne and delytable, And thow my foo, and al my folk werreyest, And of myn olde servauntes thow mysseyest, And hynderest hem with thy translacioun, And lettest folk from hire devocioun To serve me, and holdest it folye To serve Love. Thou maist yt nat denye, For in pleyn text, withouten nede of glose, Thou hast translated the Romaunce of the Rose, That is an heresye ayeins my lawe" (lines F. 320-30) This allusion is one of the arguments used for ascribing the Romaunt to Chaucer, but it is also a testimony to Chaucer's acknowledgement of having altered his translation strategy after producing the Romaunt, since this is the only case in which "it nedeth nat to glosse" (G.254). The God of Love is reprimanding Chaucer not only for having undertaken the translation of a poem that contravenes the God's principles, but also for having produced a version "in pleyn text", i.e. without Chaucer's mediation. 6 In the God of Love's opinion this intervention would be desirable in order to purge the text from all the supposedly subversive contents of the original, and thus provide a reading more favorable to the God of Love. Queen Alceste takes Chaucer's side and adduces the conventional argument of the translator who denies his responsibility for the contents of the text: JORDI SÁNCHEZ MARTÍ 6. Cf. G.85-86: "For myn entent is, or I fro yow fare, / The naked text in English to declare".
He may translate a thyng in no malyce, But for he useth bokes for to make, And taketh non hed of what matere he take, Therfore he wrot the Rose and ek Crisseyde Of innocence, and nyste what he seyde. (G.341-45) Alceste, rather ironically, tries to portray Chaucer as a mere traducer who does not judge the subject of his translations, and who, adopting a selfeffacing role, simply transfers a text from one language into another; nevertheless, she admits that Chaucer "useth bokes for to make", claiming full responsibility in those instances. This translating dichotomy is exemplified by the Romaunt and the Troilus, respectively, and Alceste proclaims Chaucer's innocence in relation to the production of the Romaunt on the basis of his low-key role as a translator. The word "innocence" here incorporates its semantic duality of Chaucer being guiltless of the crime imputed to him, and also naive, probably alluding to his youthful inexperience.
Although these words insinuate that Chaucer favored some degree of intervention in the process of translation, their implications have to be considered cautiously, because they are pronounced by the narrator in his dream. We can, however, affirm that Chaucer was aware of having altered his translating practice after the Romaunt, which is the only instance of a relatively faithful translation by Chaucer according to modern standards, thus justifying our approach. This fundamental contrast between the two Chaucerian modes of translation has been overlooked by some critics when describing Chaucer's translating practice, since they have focused on translations after the Romaunt, yet giving general validity to their conclusion. For instance, Machan (1986: 62) depicts Chaucer's translations in general in the following terms: "they all involve, though in various ways, the incorporation of material from other texts or the inclusion of original and significant Chaucerian additions", a statement that cannot be applied to the Romaunt. 7 Jiří Levý (1967) has defined the activity of translation as a decision-making process. He discusses explicitly the types of choice that the translator has to make during the stage of reformulation of the source text into the target language. The translator, however, starts making decisions long before. In order to comprehend the rationale of the translator in adopting some solutions, first we will have to delineate the function of the translated text. 8 There is no mention by Chaucer of what might have been his purpose in undertaking the 7. Shoaf's views (1979) are also based on those later translations, without any specific reference to the Romaunt.
8. Roberts (1992: 7) defines the function of translation as "the application or use which the translation is intended to have in the context of the target situation". translation of the Roman into English, so we can guess its function only from contextual elements (cf. Hatim and Mason 1990: 12). Chaucer was educated in a courtly milieu and it seems likely that he composed his first short poems in Anglo-Norman (see Robins 1978, andWimsatt 1982). Although Anglo-Norman was in decline toward the end of the fourteenth century, it was still a widespread language and it was even used in England for the creation of significant literary works (e.g. Gower's Mirour de l'Omme). Conversely, English had been relegated to instances of more familiar and less sophisticated communication, thus showing a diglossic pattern (Burnley 1989: 41, 48). We can infer that the immediate audience for the Romaunt also belonged to courtly circles and was cognizant of French, thus having no pragmatic need for Chaucer's mediation, making his effort apparently redundant. 9 If the main purpose of this translation was not to encode the source text in the target language to promote its understanding and circulation among his audience, what was Chaucer's intention?
Chaucer engaged in this translation as a linguistic test run: he wanted to try out the capacity of English to attain higher spheres of expression. Later in his career, when Chaucer had developed greater sense of authorship, he commented on the difficulties he still encountered when trying to adapt English to the mode of poetic expression of French Marguerite poetry: Allas, that I ne had Englyssh, ryme or prose, Suffisant this flour to preyse aryght! But helpeth, ye that han konnyng and myght, Ye lovers that kan make of sentement (LGW, Chaucer left the translation unfinished, but the outcome of the test must have been successful, since thenceforth his only language for literary creation would be English. This personal reassurement about the poetic possibilities of English would not be Chaucer's only gain from this translation: the Romaunt would confer prestige both on Chaucer and on English. As Tim W. Machan (1989: 66) notes, Chaucer "obtained status and authority (...), for if the sources he translated -or claimed to translate-had prestige, this prestige was necessarily a part of his own texts". 10 Nonetheless, the objective of improving English and preparing it for more elaborate discourses would always be in JORDI SÁNCHEZ MARTÍ 9. On the education of the English aristocracy, see Orme (1983). 10. For this translation Chaucer received the praise of Eustache Deschamps in his famous envoi, where he refers to Chaucer as "grant translateur, noble Geffroy Chaucier". The date of Deschamps's poem is not completely established: although traditionally 1385-86 is thought to be the most likely (Benson et al. 1987: xxiv), the date 1377-80 has also been argued by Kooijman (1980). Chaucer's agenda, as can be surmised from the prologue to A Treatis on the Astrolabe. 11 After having discussed the Romaunt's function, let us now proceed to analyze the translation process, which commences with the selection of the source. Why did Chaucer pick the Roman de la Rose? With the reading of the Roman Chaucer discovered the world of courtly allegory and of "fin'amors", which was new to him. The fascination he felt for this text would not be transitory, since Chaucer's encounter with the French poem would make an indelible impression on his poetics, as critics have extensively discussed. 12 Chaucer therefore preferred a poetical work that was ahead of his literary tradition in the establishment of his "programme", in Anton Popovič's terminology. 13 The translation of poetry has always raised the issue of its translatability. Dante (1995: 30) himself expressed the limitations of poetic translation: "E però sappia ciascuno che nulla cosa per legame musaico armonizzata si può della sua loquela in altra transmutare sanza rompere tutta sua dolcezza ed armonia". 14 The same view was subscribed by Jacob Grimm, a great translator himself: "A faithful translation of a true poem is impossible; for in order not to be any poorer than the "original" it would have to be identical with it" (quoted in Frank 1991: 118). The list of those who negate the possibility of translation could be much longer, including names such as Ortega y Gasset, Benedetto Croce, and Roman Jakobson. The force of their argument can be offset by bearing in mind Nida's words (1969: 483): "Rather than being impressed by the impossibilities of translation, anyone who is involved in the realities of translation in a broad range of languages is impressed that effective interlingual communication is always possible, despite seemingly enormous differences in linguistic structures and cultural features". Chaucer had certainly not been trained as a translator, yet presumably he had great linguistic competence in French -which would be similar to that of a 11. Here Chaucer argued for the adequacy of English for scientific discourse and set it among languages such as Greek or Latin: "This tretis ... wol I shewe the under full light reules and naked wordes in Englissh, for Latyn canst thou yit but small, my litel sone. But natheles suffise to the these trewe conclusions in Englissh as wel as sufficith to these noble clerkes Grekes these same conclusions in Grek; and to Arabiens in Arabik, and to Jewes in Ebrew, and to Latyn folk in Latyn" (ll. 25-33).
12. For a survey of the literature written on the impact of the Roman on Chaucer's works see Dahlberg (1999: 32-46).
13. Popovič (1976: 23) indicates that the selection of the source text usually "corresponds to the prevailing literary, cultural and social standard".
14. "Therefore everyone should know that nothing harmonized according to the rules of poetry can be translated from its native tongue into another without destroying all its sweetness and harmony", trans. Richard H. Lansing (1990: 18). bilingual-, and he was familiar with the French literary tradition. As a poet he could exploit his creative powers to transfer the original into his mother tongue in the most faithful manner, despite the constraints imposed by versification. Did Chaucer entertain thoughts about the impossibility of his task? As I pointed out at the beginning of this paper, Chaucer engaged in the translation of this poem as a means to test and prove the literary capabilities of his mother tongue; to admit the impossibility of such translation would have implied to acknowledge the inadequacy of English.
Thus Chaucer decided to undertake the translation of the Old French text, and began to compose a metapoem, as James S. Holmes designates this type of translation on the basis of Barthes's meta-language. 15 The next choice that Chaucer had to make concerns the prosodic schema of the metapoem; that is to say, he had to decide which form of versification would be appropriate. Guillaume de Lorris's poem is written in octosyllabic couplets; thus it is not surprising that Chaucer inclined toward the couplet of eight syllables with four beats, which was the most common form in Middle English. If Chaucer had pondered more carefully over the place of his metapoem in the English literary tradition, he would probably have preferred a verse form more appropriate for the expression of this new poetic sensibility. Later in his career Chaucer adopted the longer line with five stresses, which had been rarely used thitherto and which might have been a better option for his translation.
Derek Brewer (1998: 78) criticizes Chaucer's choice for considering it "reminiscent of the English metrical poems". Chaucer opted for a mimetic form, following Holmes's typology (1988b: 25-6), trying to imitate the meter of the source text; by his retention of the original form, Chaucer confirmed his fidelity to the original.
Before the actual writing of his metapoem, Chaucer would have done a thorough reading of the original, which would have informed his understanding of the source text, thus having a substantial impact on the translated version. Robert de Beaugrande in his chapter "The Role of Reading in Poetic Translating" comments on the importance of the reading activity in the translation process, and he adds that in contrastive studies "when errors are noticed, they tend to be attributed to the translator's writing strategies rather than his or her reading strategies" (Beaugrande 1978: 25). Despite Chaucer's undeniable linguistic competence, being able to figure out the meaning of all lexical items is a rather complicated task if we bear in mind that fundamental aids, such as bilingual dictionaries, were not available at that time. Scholars have blamed his misunderstanding of the original for some JORDI SÁNCHEZ MARTÍ 15. Holmes defines metapoem as "the poem intended as a translation of a poem into another language" (1988a: 10). negative shifts in the Romaunt. 16 For instance, in relation to the rendering of "N'estoit fardee ne guigniee" (line 1004) as "No wyndred browis had she" (A 1018), 17 Alfred David indicates that "the translation may introduce browis through misunderstanding "guignee" as "plucked"" (Benson et al. 1987(Benson et al. : 1106cf. Dahlberg 1999: 108). Another example is the use of "ridled" (A 1235) to convey the French "cueillie e iointe" (line 1213), which Langlois (1920, vol. 2, p. 306) believes to be a mistranslation. These comprehension problems are not limited to lexical, but they are also caused by syntactical difficulties: Of fruyt hadde very tree his charge, Nul arbre n'iot qui fruit ne charge, But it were any hidous tree Se n'est aucuns arbres hideus, Of which ther were two or three.
Dont il i a ou trois ou deus (A 1352-54) (lines 1326-28) 18 In addition to these cases pointed out by other critics, I have identified other difficulties of Chaucer in the negotiation of meaning. Curiously enough Chaucer seems to have problems with the transfer of alien literary terminology: English does not offer Chaucer any equivalent for the Old Occitan terms "serventois" and "rotruenges", but it is unusual that he does not attempt to include their meaning in his translation by applying some strategy, such as borrowing or paraphrasis, as he does on other occasions. This is, therefore, an instance of loss, since the translation "jargonying" ("twittering of birds", MED) deprives that sentence from the prosopopeic component of the Roman.
16. Popovič defines negative shift as "an incorrect solution of information caused by misunderstanding of the translation" (1976: 16).
17. All quotations from the Roman de la Rose are from Sutherland (1967), and the text of the Romaunt is quoted from Dahlberg (1999).
18. As Alfred David argues, "the translator misunderstood the French, which says that there were at least one or two of every kind of fruit-bearing tree except a few that were too ugly" (Benson et al. 1987(Benson et al. : 1106. See also Dahlberg, (1920: 120 Throughout the whole translation Chaucer tried to provide as much detail as possible to facilitate the comprehension of the text; it is therefore strange that these explicative remarks are not included. This makes me believe that Chaucer had difficulties understanding the words "charme" ("hornbeam") and "bricons" ("harebrained"), and preferred just to introduce a redundant comment in the first case, and to omit it in the second.
The encoding of the original message into English should be perceived as a creative act, only limited by his degree of faithfulness to the original and by the versification. But this mediation implies the negotiation of the meaning of the source text, and thus the imposition of the translator's interpretation. How can we observe the intervention of the translator? What is the effect of such mediation? I will attempt to answer these two questions by conducting a comparative analysis of both poems.
Caroline Eckhardt argues in her article that "the most obvious quality of the Romaunt as a translation is certainly its very high degree of literal reproduction of its source" (p. 46). Despite having been a generalized opinion, the literality of the Romaunt is simply unfounded, since not more than one hundred lines could be considered a word-for-word reproduction of the French original. Furthermore, not all these lines should be classified as instances of literalism, 19 since many of them are just accurate and natural English translations, as for instance "That she hadde suffred day and nyght" (A 309) for "Qu'el soffroit de iorz e de nuiz" (line 300), "For nakid as a worme was she" (A 454) for "Qu'ele estoit nue come uers" (line 445), and "In sich a gise that he hir kyste / At all tymes that hym lyste" (A 1291-92) for "En tel guise qu'il la bessoit / Toutes les foiz qu'il li plessoit" (lines 1269-70). Although it is also true that on some other occasions Chaucer imitated both the syntactic structure and the lexical material, and sometimes even the rhyme of the original. 20 Thus, "Par quel art ne par quel engin / Ie porroie entre el iardin" (lines 499-500) is rendered "By which art or by what engyne / I myght come into that gardyne" (A 511-12); "Qui ne fust en son droit asise. / Mout fu bien uestue Franchise" (lines 1216-17) is translated "That it nas in his right assise. / Full wel clothed was Fraunchise" (A 1237-38). Other instances of literalism are A 466-67, 1049-50, 1191-92, 1200-02, 1325-26, 1601-02, to the point of a practical fusion with the source text: "Largesse hadde on a robe fresh" (A 1187) is given for "Largesce ot vne robe fresche" (line 1160).
Despite the conspicuousness of these examples for the comparative critic, they are in fact an exception, since Chaucer was far from being a servile translator. Chaucer's adoption of what Vinay and Darbelnet (1995: 31) designate as oblique translation methods reveals the translator's compromise to generate a text that works in the target language while moving away from the phraseology of the source text. One of the methods of oblique translation he adopted is transposition, which "involves replacing one word class with another without changing the meaning of the message" (Vinay and Darbelnet 1995: 36). This method is used profusely by Chaucer throughout his translation and shows his fidelity to the sense of the original while avoiding a word-for-word reproduction. The examples are diverse and include verbs rephrased as nouns: "But to my joye and my pleyng" (A 598) for "Qu'a moi iouer e solacier" (A 587); and also a noun instead of a verb: "To be aqueynted with Richesse" (A 1139) for "...l'acointance / De Richece..." (lines 1119-20); adjectives instead of nouns: "So feirs and daungerous was he" (A 1482) for "Plains de desdaing e de fierté" (line 1450). The enumeration of all the instances would be tedious, so I will simply indicate some other occurrences: A 552, 691, 1107-08, 1287, 1350, 1407, and1482. It is worth mentioning, however, that on some occasions Chaucer used this translation method to intercalate his own interpretation of the original. For instance, "Qu'el sembloit estre enlangouree" (line 202) is rendered as "Hir semed to have lyved in langour" (A 214). Here, in addition to the syntactic transposition, there is also semantic variation, since the English version replaces the French present infinitive with a perfective form, thus endowing it with a more palpable and vivid temporal dimension than the neutrality of the French. "E gent mignotement baler" (line 744) is translated as "And folk daunce and mery bene" (A 760), which implies semantic deviation despite the correspondence among the lexical items, since the adverb in the French text modifies the ver "bales", whereas its translation is describing "folk".
The other method of oblique translation relevant to this comparative study is modulation, defined by Vinay and Darbelnet (1995: 346) as "a translation method consisting of changing a point of view, an evocation, and often a category of thought". Chaucer used this method extensively in his translation, so I will limit myself to present some illustrative examples: "Que toute rien d'amer s'esfroie" (line 85) is rendered "Whan love affraieth al thing" (A 91), which is an instance of modulation by reversal of forms. This transformation is caused by the use of the reflexive pronominal form in the French text, which is always difficult to convey in English: in the original "toute rien" is the subject responsible for the action (the agent), but at the same time it is the object (the patient); Chaucer's translation, however, chooses "love" as the grammatical subject and "al thing" as the object of the verb, thus representing a more direct description. There is also modulation when changing an active sentence into the passive voice and vice versa: "E la chalor aual descent" (line 1545), is construed as "And that the heete descendid is" (A 1575); "Cele por qui ie l'ai empris" (line 41) is translated as "For whom that it begonnen is!" (A 43). There is also a relevant occurrence of free modulation, in which a literal translation would seem vague in the target language: "Et si fet dou seignor sergent" (line 870) is rendered "And he can wel these lordis thrallen" (A 882). Here Chaucer reveals his fidelity to the original meaning, but it is expressed in fluent English. There are other occasions on which he combined modulation with transposition: "Con cele qui mout fu iree" (line 318) is translated "As she that was fulfilled of ire" (A 326). Here the past participle "iree" becomes a noun in English (transposition), and the substitution of "fulfilled" for "mout" denotes a different conceptualization (modulation). Another instance of the integration of both methods is the rendition of "Ci ne set conseiller nus" (line 1585) as "Heere lith no rede ne witte therto" (A 1615), in which the English version adopts an impersonal construction and replaces the verb "conseiller" with the nouns "rede ne witte".
All these examples consistently show the Romaunt's closeness to the original. Now, however, I would like to pay attention to two other cases of modulation in which there is no deviation from the sense of the source text, but which have significant narratological implications. The first instance seems to be just a change from the passive into the active voice: "Ce est li Romanz de la Rose, / Ou l'art d'Amors est tote enclose" (lines 37-38) is rendered "It is the Romance of the Rose, / In whiche al the arte of love I close (A 39-40). This transformation helps Chaucer with the rhyme, but the explicitation of the first person at this point also serves to assert a distinctive authorial voice for the translation. The neutrality of the French passive sentence is displaced by the greater definiteness of that "I" that can only be identified with the translator's voice, i.e. with Chaucer. 21 This differentiation from the narrator of the source text persists throughtout the translation, and is achieved mostly through the method of explicitation, as will be shown below.
The other instance of modulation that I want to comment upon also reflects the same type of conflict between both narrators: "Nou metrai pas en obliance" (line 982) is rendered "As fer as I have remembraunce" (A 996). On JORDI SÁNCHEZ MARTÍ 21. Eckhardt has used this example to substantiate her conclusion that in the Romaunt there is "a somewhat stronger awareness of the narrator as character" (1984: 52). the surface it is just a case of substitution of an affirmative sentence for a negative one without semantic consequences (see Vinay and Darbelnet 1995: 37). But, in fact, there is a change of attitude, since the certainty of the French narrator is downplayed with a weak and conditional statement by the translator, who expresses his lack of command of the matter of the poem. Previously, there was another case of modulation hinting in this same direction: "Mès de ce ne fet a parler" (line 764) is rendered as "But herof lieth no remembraunce" (A 782).
Another translation phenomenon deserving our consideration is dilution, a strategy that aspires to maintain the semantic correspondence between the source text and the translation, but implies a stylistic variation at the level of expression. Chaucer's use of dilution -"the translation technique of spreading one meaning over several lexical items" (Vinay and Darbelnet 1995: 341)displays his recreation of the materia of the poem: even though the translation does not depart in substance from the original, Chaucer opted to employ more direct and precise language, thus providing a more exact image of the fictional world. The translation of the French pronominal verb "se taire" caused Chaucer some trouble, but he exploited the occasion and introduced his own understanding in the translation: "Li oisel, qui se sont teü" (line 67) is rendered as "The briddes that haven lefte her song" (A 71), and "Orendroit m'en couendre teire" (line 1413) is translated as "I mote my tonge stynten nede" (A 1441). In both examples the English version diverges from the original and pronounces a tonal distance in preferring greater concreteness and vividness. The mediation of the translator can also be felt in the expression of mortality: "S'ele morust, ne granz pechiez" (line 349) is rendered "Ne synne, although her lyfe were gone" (A 358), and "Ne fu d'ome mortel oïe" (line 668) is translated as "Was herd of man that myght dye" (A 676). In both cases Chaucer decided to make explicit the meaning of "morust" and "mortel" respectively, which results in an emphasis on the transcendence of death. Here there is no linguistic difficulty that may justify Chaucer's choice of dilution, thus showing the translator's intent.
This tendency toward greater clarity is enhanced by the extensive use of specification and explicitation: "A shift towards greater specification will produce a transeme the meaning of which is made more precise, by either the addition of extra words or the use of words with a less general meaning" (Shuttleworth and Cowie 1997: 159). This translation strategy permits Chaucer to describe the fictional world as closer to the narrator, and thus to make it more visible to the reader. Hence "These theves and these smale harlotes" (A 191), by means of the demonstrative adjectives and the affective "smale", gives a more immediate picture than the French "Les larrons e les ribaudiaus" (line 177). Likewise the line "The watir that so wel lyked me" (A 121) introduces an element of subjectivity which is absent in the original "Cele eue qui si bien seoit" (line 115). The same desire for concreteness explains the rendering of the French indefinite pronoun "neant" (line 446) as "peny" (A 246); the translation of "Se li tens fust .i. poi diuers" (line 446) as "And if the wedir stormy were" (A 455); "Cheueus ot blons com bacins" (line 527) as "As ony basyn scoured newe" (A 540); "Ainz fu clere come la lune" (line 996) as "And clere as the mone lyght" (A 1010).
The list of instances of specification could be further enlarged, but there is still another translation mechanism that requires more attention, since it was used the most in the Romaunt: explicitation. The process of explicitation "is brought about by the translator filling out ST [the source text], for example including additional explanatory phrases, spelling out implicatures or adding connectives to "help" the logical flow of the text and to increase readability" (Shuttleworth and Cowie 1997: 55). This strategy is the main cause for the expansion of the translation -thirty-five lines longer-, and also reveals Chaucer's intervention more noticeably. This was directed toward the construction of the fictional world of the translation according to Chaucer's interpretation of the original, and once more this method produces a text endowed with greater preciseness than the Roman.
This phenomenon can be observed mainly in the addition of linguistic material which, although it does not contradict the sense of the original, sets a different tone to the translation and limits its referential capability. In the following example Chaucer preferred to introduce a physical detail ("by her throtes") and omit the redundancy that imply "pechiez" and "maus": And that is routh, for by her throtes Si est granz pechiez e granz maus Ful many oon hangith at the laste Qu'en la fin maint en couient pendre (A 192-93) (lines 178-79) Chaucer's variation confers greater graphic quality to his version, which is one of the main characteristics of the Romaunt. The translation of "Qui enuoisiement chantoient" (line 494) as "That songen thorugh her mery throtes" (A 507) displays the same tendency. Chaucer's attention to detail is best exemplified by the meticulous description of Narcissus's movements at the fountain, which contrasts with the succintness of the original: And doun on knees he gan to fall, Sus la fontaine toz adenz And forth his heed and necke he straught Se mist lors por boiure dedenz To drynken of that welle a draught (lines 1481-2) (A 1514-6) On other occasions Chaucer's intervention consists in expressing overtly the covert implications of the original text; consequently, the translated text is endowed with greater denotation, but its power of suggestion is drastically diminished. Thus Chaucer erased any possible shades of meaning when dealing with age. First, in the description of Old Age, the Romaunt points out that she was shorter "Than she was wonte in her yonghede" (A 351), whereas the original simply states "De tele come el soloit estre" (line 341). Similarly, when the narrator comments upon the fact that elderly people tend to be cold, the translation blatantly affirms "Her kynde is sich whan they ben olde" (A 412); the remark in the Roman is gentler and it asks for some sympathy from the audience: "Bien sauez que c'est lor nature" (line 404). When later the narrator is depicting Youth, Chaucer's narrator immediately imposes his view, limiting the readers in their reception of the text: while the Roman reports that Jonece "... n'auoit pas encor passez, / Si con ie cuit, .xii. anz d'assez" (lines 1261-62), the Romaunt reads "That nas not yit twelve yeer of age, / With herte wylde and thought volage" (A 1283-84). Also, in the description of the garden in which Narcissus's pool is found, Chaucer's narrator depicts it as a space "On whiche men myght his lemman ley / As on a fetherbed to pley" (A 1421-22), where the original only states that "Ausi i pooit l'en sa drue, / Couchier come sus une coute" (1394-95). The subtly suggestive effect created by the narrator of the French text vanishes in the translation with the unconcealed imposition of Chaucer's reading by the addition of the verb "to pley" ("to play amorously; make love, engage in sexual intercourse", MED). 22 Not all the cases of explicitation, however, are informed by a manipulative force. There are other occasions in which Chaucer simply highlighted the subject of his sentence with a clarifying purpose. In the 22. Weiss (1985: 210) considers that this interpolation is appropriate and states that "Chaucer's addition of the verb "to pleye" is a most delightful touch in the same direction [i.e. toward greater concreteness]". portrait of Poverty, Guillaume de Lorris started by saying immediately after a period that "Des autres fu un poi loignet" (line 453), whereas Chaucer properly preferred to furnish his text with greater cohesion by writing "And she was putt, that I of talke / Fer for these other..." (A 463-64; emphasis added). The same reason moved him to explicitate "...this lettre of which I telle" (A 1543), when the French only has "...li escrit..." (line 1511). In other cases Chaucer revealed the implicature of the original, despite its obviousness; for instance, when he is talking about the feats that a Knight had accomplished for Largesse, his beloved, the French texts reads "por s'amie" (1187), whereas Chaucer's insists in clarifying even the evident, and says "...for the love of his lemman" (A 1209).
In general, the effect of this translation strategy hinders the expression of "sentement" envisioned by Chaucer, and at the same time it provides the reader with a less demanding text, since the translator has already resolved the blurry areas for his audience. The greater directness of the Romaunt had already been indicated by Caroline Eckhardt and by Alexander Weiss, but both fail to identify and discuss the mechanisms generating this effect. Nonetheless, we must be careful when addressing the issue of explicitation, because it is a natural tendency of most translations, as George Steiner (1998: 291) has indicated: "the mechanics of translation are primarily explicative, they explicate (or, strictly speaking "explicitate") and make graphic as much as they can of the semantic inherence of the original".
Finally, there is another factor that offers information on the translator's view with relation to his own production: the treatment of cultural references from the source text. Chaucer consistently retained the cultural allusions of the original without any effort for naturalization. All the French geographical references are kept without any variation: "...in all the rewme of Fraunce" (A 495), "...in all Arras" (A 1234), "...for Parys..." (A 1654). The preservation of these allusions can be explained not only by the translator's fidelity to the original, but also as the result of a decision made by Chaucer in the process of translating: he wanted to create a text that may be perceived as culturally foreign because of its contents and framework. Hence Chaucer even promoted this effect of exotization through the addition of allusions not included in the original, but which are coherent within the frame of reference of the source text: 23 "...iusqu'en Ierusalem" (line 542) is rendered as "Fro Jerusalem unto Burgoyne" (A 554).
Chaucer's attitude was not going to hamper the reception of his translation because of the familiarity in England with French issues. There is one occasion, however, on which Chaucer omitted one geographical allusion, precisely because it would be devoid of significance for an English audience: "Si n'ot mie nés olenois [i.e. of Orléans]" (line 1194) is translated as "Her nose was wrought at poynt devys" (A 1215), which represents an interpretation of the French. 24 Later Chaucer interpolates an informative note that sets the Romaunt as the translation of a French original, a fact which had not been indicated at all in the text: Though we mermaydens clepe hem here Que, pour leur uoiz qu'eles ont saines In English, as is oure usaunce, E series, ont non seraines Men clepe hem sereyns in Fraunce (lines 673-74) (A 682-84) 25 This remark together with A 1228, in addition to presenting the text as a translation, also serve to define the identity of the translation as a different text, composed in another language for another audience and with a different tone. The verification that the Romaunt is construed as an entity independent from the original comes from the following translation of another cultural comment: Somme songe songes of Loreyne Li autres notes lohorenges; For in Loreyn her notes bee Por ce cou fet en Loheraigne Full swetter than in this contre.
Plus beles notes qu'en nul raigne (A 766-68) (lines 750-52) Chaucer maintains the cultural note, yet he introduces a significant modification in establishing as the term for the comparison "this contre", which his audience would read as England (see Benson et al. 1987Benson et al. : 1105. The examination of the translation process through this comparative analysis of the Romaunt with the original can help us determine Chaucer's poetics of translation in this case. The most prominent characteristic of the translation is the absolute fidelity to the meaning of the original. As it has been observed above, when there are semantic deviations from the source text, these are minimal and they never contravene the intended meaning of the Roman. Despite this faithfulness, Chaucer is far from being a fidus interpretes, 26 since he in most cases avoided the literal word-for-word translation by applying a variety of translation strategies, which have been studied above. 24. See Langlois (1920: 306). Dahlberg says that "to judge from Ln's [Langlois] note, a flat nose " (1995: 363) 25. See Dahlberg (1999: 97). Cf. A 1228: "Whiche tree in Fraunce men cal a pyne". 26. For a review of the ideas attached to the fidus interpretes, see Copeland (1991: 168-178).
While following the sense of the original, Chaucer was also concerned to bring his text closer to his potential readership. Thus his mediation was geared toward creating a text endowed with greater tangibility, which is achieved by means of vivid descriptions that guide the reader through the allegorical world of the original. Chaucer's narrator observes this same world with a penetrating eye, and represents the fictional world by focusing on details which are sometimes absent in the original. It is at this point that Chaucer's dialogic interaction with the original is revealed, representing a Bakhtinian zone of dialogical contact in which the translator establishes a dialog with the source text, negotiating its meaning and defining its own perspective for the translation's particular conversation with its readership.
In sum, Chaucer was aware that fidelity does not necessarily imply obsequiousness, since he felt responsible for his own product, as can be inferred from the revealing line "al the arte of love I close" (A 40). He was also conscious that the translated text has its own identity, different from that of the original, and he hoped that it be perceived by his audience as belonging to the English literary tradition -thus his clarification "in this contre", A 768without disguising its being a translation -thus his allusions to the language of the original and the numerous borrowings-, a tendency which would evolve into the total appropriation of the original text, so characteristic of later Chaucerian translations.