Aspects of Translation Technique in Antiquity Brock, Sebastian Greek, Roman and Byzantine Studies; Spring 1979; 20, 1; Periodicals Archive Online pg. 69 Aspects of Translation Technique in Antiquity Sebastian Brock I N HIS De optimo genere oratorum (§ 14) Cicero discusses his own translations of Aeschines and Demosthenes, commenting as follows: nec converti ut interpres, sed ut orator, sententiis isdem et earum formis tanquam figuris, verbis ad nos tram consuetudinem aptis. in quibus non verbum pro verbo necesse habui reddere, sed genus omne verborum vimque servavi. non enim ea me adnumerare lectori putavi oportere, sed tanquam appendere. The same contrast between the work of the literary translator and that of the interpres is further brought out in his De finibus (3.15) where, in the face of criticism, Cicero defends his occasional use of neologisms, but goes on to assure the reader: nec tamen exprimi verbum e verbo necesse erit, ut interpretes indiserti solent, cum sit verbum quod idem declaret magis usitatum. Horace expresses very much the same view of translation in his Ars poetica (133): nee verbo verbum eurabis reddere fidus interpres. In other words, the translator of a literary text went about his work in a manner totally different from that of the fidus interpres, the hack translator, who produced slavish renderings of legal and business documents. As Franz Blatt once observed, talking of Latin transla­ tions from Greek, "on se trouve rarement en face d'une veritable traduction prechretienne." 1 Jerome endorses the same view as Cicero and Horace in his influential letter 57, addressed to Pammachius: ego enim non solum fa tear, sed libera voce prcifiteor, me in interpretatione graecorum, absque scripturis sanctis, ubi et verborum ordo mysterium est, non verbum e verba, sed sensum exprimere de sensu. 1 F. Blatt, ClMed 1 (1938) 220. 69 70 ASPECTS OF TRANSLATION TECHNIQUE Jerome's passing mention of "the holy scriptures, where even the word order is a mystery" indicates how the neat dichotomy between literary and non-literary translation, hitherto operative, breaks down with the advent of Christianity, or rather, as we shall see, with the advent of biblical translation. Jerome is offering no new ideal for translating the Bible here, he is simply reflecting the accumulated experience of over half a millennium of biblical translation. Owing to the prestige of this ideal of literal biblical translation, it eventually became the norm for virtually all translation from Greek into Latin until the Renaissance. In the sixth century, when Boethius translated Porphyry's Eisagoge, and in the ninth century, when John Scotus translated the pseudo-Dionysian corpus, each in turn remarks in the preface to his work that the literal style of translation adopted will make him liable to incur Horace's deprecatory label of fidus inter pres. 2 John Scotus excuses himself to the reader by explaining that he deliberately sets out to be an inter pres, and not an expositor, of the work. Cicero's ideal of the translator as orator, and not an interpres, is completely reversed. I It is the background to this literal style of translation, verbum e verbo, as opposed to sensus de sensu, that will be examined in the present paper. Since this atomistic approach is so blatantly in contradiction to most modern ideals of translation, it is worth stressing at the outset that its practitioners did not choose this method because of any inadequacy on their part for the task: the choice is a deliberate one, and my aim here is to attempt to bring out the motivation that led to its adoption. Furthermore, we should remember that the method was capable of very considerable sophistication, and some of the men who produced these astonishing versions obviously had a very wide command of the two languages with which they were working. The mediaeval western ideal of the word-for-word translation thus has its origin in biblical translation practice. How biblical translation came to adopt this approach in the first place is not at once obvious, and we need to go back to its inception, in the Septuagint (a title 2 See W. Schwarz, "The Meaning ofjidus interpres in Medieval Translation," YThS 45 (1944) 73-78. SEBASTIAN BROCK 71 originally applicable only to the Greek translation of the Penta­ teuch).3 With such a long tradition of biblical translation behind us, it is perhaps a little too easy to overlook the remarkable character of this earliest undertaking: no other oriental religious text of such an extent achieved the honour of translation into Greek. This very fact that such an undertaking was entirely without precedent, however, posed a problem for the Jewish translators: on what were they to model their translation? If one turns to the available evidence for translation in the Hellen­ istic world, it would appear that Cicero's radical distinction between literary and non-literary translation was already operative. The Greek version of some of Asoka's edicts, found on a bilingual Greek­ Aramaic inscription of the third century B.C. at Kandahar (Mghan­ istan), treats the original with considerable liberty, and the text has been deliberately presented a la grecque.4 The same wish to present the reader with a version in reasonably good Greek style seems also to have guided the Greek translator of the Demotic story of Tefnut, preserved in P.Lond. inv. 274.5 At the opposite end of the scale come the word-for-word transla­ tions of legal documents in Egypt and of senatus consulta and other government missives originating from Republican Rome. 6 Dating from rather later, the bilingual texts of Vergil and others, for pedagogical use, with the Latin and Greek in parallel columns, belong to the same world of very literal translation. 7 Since the Pentateuch was both a legal and at the same time a 3 See my "The Phenomenon of the Septuagint," Oudtestamentische Studien 17 (1972) 11-36, where some of the issues raised in this paper are touched on from the point of view of the early history of the Septuagint. 4 L. Robert, Journal Asiatique 246 (1958) 13. 5 See S. West, "The Greek Version of the Legend of Tefnut, " JEA 55 (1969) 161-83. 6 On these see P. Viereck, Sermo graecus quo senatus populusque romanus magistratusque populi romani ... usi sunt examinatur (Gottingen 1888), and R. J. Sherk, Roman Documentsfrom the Greek East (Baltimore 1969). From the early Empire the Greek version of the Monumentum Ancyranum is of particular interest from this point of view; on this see A. P. M. Meuwese, De rerum gestarum divi Augusti versione graeca (Amsterdam 1920). The interaction between the two styles is discussed by H. J. Mason, "The Roman Government in Greek Sources; the Effect of Literary Theory on the Translation of Official Titles," Phoenix 24 (1970) 150-59. 7 The texts can most conveniently be found in R. Cavenaile, Corpus Papyrorum Latinarum (Wiesbaden 1958); see also R. E. Gaebel, "The Greek Word-lists to Vergil and Cicero," Bulletin of the John Rylands Library 52 (1969/70) 284-325. 72 ASPECTS OF TRANSLATION TECHNIQUE literary text, the initial translators were faced with a dilemma, and their hesitation is reflected in the inconsistent nature of their transla­ tion. Thus, for example, the Hebrew idiom wayyitab b-cene, 'it pleased' (lit. 'it was good in the eyes of'), is rendered in Genesis LXX once idiomatically ~xap'Y'J S~ @apaw (45: 16) but in the other two occur­ rences more literally Ka~ 7JP€CfXV ot '\6yo£ ~vav'rtov (34:18, 41 :37). More pedantically literal translations, such as apECK€£V ~v ocp()a,\fLoLc or aya()vv€£v ~VW1T£OV (= Heb. lipne, 'to the face of') were later to be devised. The Pentateuch in Greek probably goes back to the first halfofthe third century B.C. A hundred years later there is evidence to show that general opinion had clearly come down on the side of considering the Bible as a legal rather than a literary document (the standard method of quoting scripture, "as it is written," significantly derives from a legal context). This placed the Jews of the diaspora in a difficult position: either their Greek Bible required 'correcting' on the basis of the Hebrew to produce a more literal translation, or it had to be accorded the same position of prestige as the Hebrew original; in other words, the translation too had to be accepted as inspired. The latter position was adopted by most Jews from Egypt, and it is given expression first in the treatise Ad Aristeam, a work of the second half of the second century B.C., containing the earliest account of the legend of the seventy translators. Later on Philo develops the idea, and in his Life of Moses (2.40) he asserts that the Seventy should be called t€pocpav-rac Ka~ 7Tpocp7}Tac rather than EpfL'Y'JvEac (i.e. Cicero's interpretes indiserti). The alternative, to bring the original translation into closer line with the Hebrew, was the answer of unhellenized Jews in Palestine, whose work of correction, covering two or three centuries, is now beginning to be better known thanks to the chance find in the J udean desert of some fragments of the Twelve Prophets in Greek, dating from the turn of the Christian era, and the brilliant inter­ pretation of their significance by Pere Dominique Barthelemy.8 Later on we shall be examining some of the techniques of these revisers in the light of some other literalist translations in antiquity, notably the Roman official documents in Greek, the Greek transla­ tions of Vergil and the sixth/seventh-century school of translators 8 US devanciers d'Aquila (Leiden 1963). SEBASTIAN BROCK 73 from Greek into Syriac. First, however, it is important to look at some of the reasoning behind the initial choice of a literal rather than a free translation for any particular text and the conditions under which it is a practical option. In very general terms the sensus de sensu approach can be seen as bringing the original to the reader, whereas in the verbum e verbo translation the original acts, as it were, as Aristotle's unmoved mover, and the psychological effect is to bring the reader to the original. Seen in this light, it is obvious that the translator in antiquity had to make a momentous choice at the very outset. There are several different factors which condition the choice he makes, among which the most important are: the nature of the text he is translating, the relative prestige of the two languages concerned and the extent to which the source language is still widely known. The choice of the literal approach for school texts requires no special comment. The bilingual Vergil papyri served the same purpose as schoolboy cribs and the more sophisticated, highly literal Latin translations employed today by some orientalists. 9 Administrative and legal documents obviously need to convey to the reader the exact meaning of the original. This does not necessarily mean that the word-for-word translation is the most suitable, since the over-literal can be meaningless. In the bilingual semi-official honorary inscriptions from Palmyra, for example, it is often not possible to say whether the Greek or the Palmyrene is the translation: the essential content of the message is indeed identical but the phraseology quite different, each language following its own for­ mulae. Thus in an inscription from the sanctuary of Bel at Palmyra, dated A.D. 120/1, wehave: 10 rR (3oVA~ Kat 0 ofjfLoC MEAE'iv dans l'histoire du Canon," VigChr 3 (1949) 1-36. 18 See W. Schwarz, Principles and Problems of Biblical Translation (Cambridge 1955) 169-72. SEBASTIAN BROCK 77 had begun a translation from Egyptian of a book about Imhotep, but, overwhelmed by the nature of the task, he had given up; it was only after a divine visitation that he eventually took up the work again, T~V c~v (the god's) E1T£KaAECaJLEvoc 7TpovoLav Ka~ 7Tk'lpWOE~C Tijc cfjc OELOT7]TOC. Once the principle of translating a sacred text was admitted, the only compromise open in Palestinian Jewish circles-torn between the need (at least in some quarters) for a Greek translation and a consciousness of the overriding importance of the source text in Hebrew-lay in the adoption of a slavishly literal approach to biblical translation. The result was this long drawn out process of revision of the original Septuagint translations, culminating in the astonishing refinements of Aquila in the early second century. Later on, official Judaism turned against written translations of the Bible altogether. R.Judah b. Ilai, a pupil of R.Akiba at the end of the second century, held that "he who translates literally is a falsifier, while he who adds anything (by way of paraphrase) is a blasphemer." 19 The insistence that the Aramaic translations (targumim) should only be oral and not written down points to one of the factors that led to this new stance: at a time when Greek and Aramaic were gaining as vernaculars in Palestine at the expense of Hebrew, to allow a written translation into either of these languages was tantamount to encouraging the demise of Hebrew as a cultural language. The Hebrew Bible thus provided an ideal focus for cultural identity at a time when political aspirations had been effectively quashed. Many modern parallels to sensitivity on this issue will spring to mind, and it may well be that Egyptian reluctance to translate sacred books was inspired as much by these considerations as by doubts over the validity of translating a sacred text at all. In the Christian church, literal translation acquired an importance for different reasons. In the biblical sphere, because Christians for the most part adopted the attitude of Alexandrine Jewry and re­ garded the Septuagint as itself inspired, Aquila'S method ()OVAEVWV Tfj rE{1paLKfj M~EL (Orig. Ep. 1.2) was generally derided; nevertheless, controversy with Jews made it important for Christians to have a Greek biblical text which agreed with that current in Jewish circles. 19 Bah. Talm. Qjdd. 49a; Tosefta, Meg. 4.41. 78 ASPECTS OF TRANSLATION TECHNIQUE The result of this particular need was Origen~s monumental Hexapla,20 the inspiration for whose format may possibly have lain in the bilingual Vergil texts. Outside the Bible, literal translation rapidly became a requirement with the spread of heresy. Translators could-and did-interpolate their translations of orthodox writers and so pass off heterodox opinions under the auspices of some revered authority. Likewise, exact translations of works suspected of heterodoxy were required for the consideration of synods and councils. Marius Mercator produced the Latin translation of many Greek documents of the Council of Ephesus (431), and he prefaced his translation of a work of N estorius with the following words: 21 As far as I was able, I have tried to translate the work verbum de verba, lest I subsequently be found a falsifier rather than a true translator (verus inter pres ) • Precision in the translation of doctrinally suspect works was obviously essential. John Scotus, realizing that by translating Ps. Dionysius~ works into Latin he was introducing his readers to heady stuff~ is very careful to point out that he has translated verbum de verbo, and consequently allegations of dubious orthodoxy should be laid at the door of the original author and not the translator.22 Literal translation has thus become a double safeguard: for the reader, against the introduction of false or heretical views by the translator, and for the translator, against accusations by the reader of falsification of the thought of the original. The Graeco-Roman ideal of translation was primarily interested 20 So I have argued in "Origen's Aims as a Textual Critic of the Old Testament," Studia Patristica 10 (1970) 215-18. 21 Schwartz, Acta Cone.Oec. 1.5, pp.28-29. The passage is worth quoting as a whole: "Nestorii quondam episcopi Constantinopolitanae urbis nonnulla ad pLebem blasphemiarum dicta vel scripta ex Graeco in Latinum sermonem,fervore catholicaefidei ineitatus, curavi traniferre afidelibus linguae meae fratribus cognoscenda atque vitanda, in quibus verbum de verbo~ in quantum fieri potuit, conatus sum translator exprimere, ne prius falsarius magis quam verus postea probarer interpres. da igitur veniam, pie lector, si aut minus oratio lucuLenta est aut verborum ubicumque praesumptorum novitas aurem forte percukrit. elegi obtrectatorum Unguis magis exponi quam veritate sensuum exprimenda, ubi est omne defalsitate periculum, longius aberre. occupent igitur se ex hoc idem isti nostri disertuli ad singulas syllabas nostras scrutandas et verba rimanda: non id curo nec magni pendo, securus quod mihi de hoc opere nullus falsarii nomen inponat. scio etiam ab istis exprobanda nobis esse aliqua dicta vitiosa, quae nobis vis servandae Graecae proprietas extorsit." (On pp.395-415 there will be found useful Greek/Latin indices to the translation.) 22 Migne, P L 122.1032: ... sin vero obscuram minusque apertam praedictae interpretationis seriem iwiicaverit, videat me interpretem huius operis esse, non expositorem. SEBASTIAN BROCK 79 in form and the impact this had on the reader. In total contrast, the Jewish and Christian ideal was solely concerned with the content of the original, and biblical translation, once established, provides the model for translation technique in other areas as well: only a few fields, such as hagiography and other popular forms of literature, escape the deliberately fossilizing embrace of the literalist approach. Why concern for the content of the original did not lead to a desire to communicate that content in a more intelligible form, I have already tried to indicate: to communicate the message to the reader or hearer was the task of the expositor and not the inter pres. But there is also a further consideration which pointed to the atomistic verbum e verbo approach as the only acceptable mode of procedure. Given the view, widespread in antiquity, that what can be fully described must in some way be less than the mind that describes it,23 to translate an inspired text sensus de sensu would be to imply that the sensus of the impenetrable mysteries of scripture had been fully grasped by the translator; he would thus be guilty of the same presumption that Eunomius displayed when he claimed to be able to give a rational description of the nature of the Logos. Of course, in contexts where expositores with a knowledge of the source language were no longer readily available, tacit compromise had to be made, as Jerome's own practice in fact shows. To borrow an analogy from another field of academic activity, the aim of the literalist translators was to produce as it were a diplomatic edition of his text and not a critical one. And as such, the modern editor of their various source texts has ample cause to be grateful to these translators for producing their mirror versions. II In the second half of this paper we shall look at some of the tech­ niques of the verbum e verbo translators. Obviously some translations are literal because of the translator's lack of skill in either the source or the receptor language. Our concern is not with these, but with the highly sophisticated work of men who were probably bilingual in the 23 The point is expressed in a nutshell by the fifth-century Syriac writer John of Apamea (B.L., Add. 17167, foI.l38a): "he who is contained by knowledge cannot be Lord of that knowledge which contains him"; see also Greg.Nyss. v.Mos., PG 44.404-05. What seems to be a similar shift in emphasis from form to content can also be observed in the art of the period: see E. Kitzinger, Byzantine Art in the M alcing (London 1977) ch.1. 80 ASPECTS OF TRANSLATION TECHNIQUE two languages with which they were operating. I shall take my examples primarily from translations of Hebrew into Greek (biblical) and of Greek into Syriac (both biblical and non-biblical), since these are languages whose basically different structures pose obvious problems to the literal translator. My main sources will be Aquila and his predecessors for translation from Hebrew into Greek, and, for Greek into Syriac, the work of three early seventh-century translators who probably trained at the same monastery: Paul, bishop of Tella, who translated Origen's revised Septuagint text into Syriac (the 'Syrohexapla'), Thomas of Harkel, the author of the so­ called Harklean New Testament (a revision of an earlier revision of the Syriac New Testament made under the auspices of Philoxenus of Mabbug at the beginning of the sixth century),24 and Paul, bishop of Edessa, who revised a sixth-century translation of the works of Gregory of N azianzus. For comparison'S sake, however, I shall also draw upon some of the other manifestations of literal translation in the Graeco-Roman world, in particular the Republican documents from the east in Greek and the bilingual Vergil texts. In passing, it might be said that, since the late sixth and early seventh centuries witnessed the develop­ ment of literalizing schools of translation from Greek into Latin and Armenian as well as into Syriac, an interesting task for the future would be to examine the respective methods of these schools.25 As a preliminary remark, it should be noted that, except in the case of the Roman official documents, the receptor and not the source language was the translator's first language: this at once indicates that the abuse of syntax consequent upon this method of translation is deliberate and not due to incompetence. 24 Whereas Philoxenus' motivation for sponsoring a revision of the Syriac Peshitta New Testament was primarily theological, the interests of Thomas (and the two Pauls) were essentially philological; on this, see my "The Resolution of the PhiloxenianfHarklean Problem," forthcoming in the Festschrift for B. M. Metzger. 25 For Latin see, for example, L. Szymanski, The Translation Procedure of Epiphanius­ Cassiodorusin the Historia Tripartita Books I and II (Washington 1963), and F. Weissengruber, Epiphanius Scholasticus als Vbersetzer zu Cassiodorus-Epiphanius Historia Ecclesiastica Tripartita (SBWien 1972, Bd.283); for Armenian see H. Manandian, The Hellenizing School and its Development [in Armenian] (Vienna 1928); for Syriac see S. R0rdam, "Dissertatio de regulis grammaticis quas secutus est Paulus Tellensis in veteri testamento ex graeco syriace vertendo" (=pp.I-59 of his Libri Judicum et Ruth secundum versionem Syriaco­ Hexaplarem [Copenhagen 1859-61]), and my The Syriac Version of the Pseudo-Nonnos Mythological Scholia (Cambridge 1971) 34-44. SEBASTIAN BROCK 81 Translations can be literal in varying degrees, and a proper characterization lies partly in the successful identification of the translator's particular interests and concerns: most literalist trans­ lators concentrate their attention on certain features only. Man­ andian, in his classic study of the Armenian hellenizing school, was able to distinguish four different styles of translation there. Similar distinctions could certainly be made for the contemporary Syriac school of translation, and in the case of biblical translation from Hebrew into Greek it is now possible to isolate a number of stages of development in the work of the literalist revisers, culminating in Aquila. But details of these diachronic features are not our concern here; rather, we shall be looking at some of the limits to which literalism could be pressed. WORD ORDER AND FORMAL CORRESPONDENCE In the case of the biblical translator, for whom "the very word order is a mystery," reproduction of the word order of the source language is essential. The preservation of the word order of the original indeed serves as the most obvious mark of any literal trans­ lation, and in the case of the Vergil translations it is worth noting that this is more or less a requirement of the format. In most ex­ ponents of the method it is also of importance to render as far as possible each vocable with one that corresponds to it in grammatical function. This principle of formal correspondence produces many interesting refinements both in the work of Aquila and his pre­ decessors and in that of the seventh-century Syriac translators. Particles hitherto left untranslated by earlier translators are now carefully represented: thus the earlier Greek biblical translations render the Hebrew particles W-, gam, w-gam ('and', 'also', 'and also') indifferently by Kat, but Aquila and his predecessors wished to dis­ tinguish between these three words 26 and so rendered w- by Kat, gam by Kat 'Y€ and w-gam by Kat Kat 'Y€. Following the same principles Aquila neatly represents the Hebrew local suffix he by the Greek suffix -O€: thus OlKOVO€ for baytah, 'house-wards'. Aquila's choice of the Homeric OlKOVS€ rather than the Attic OlKaS€ may shed some light on the most notorious feature of Aquila's translation technique, his use of cvv+ accusative to represent Hebrew Jet, which can serve either to 26 There were hermeneutical reasons for this; see Reider in JQR N.S. 4 (1913/4) 336 n.33. 82 ASPECTS OF TRANSLATION TECHNIQUE introduce a defined object or as a preposition, 'with'. Was Aquila's starting point the adverbial use of cvv in Homer ?27 In that case it is likely that Aquila too was treating cvv as an adverb and not as a preposition when he translated the opening verse of Genesis EV ,J.. '\ I., 8 '\ \ \ , '\ '" , .... KE't'afl(XU.p EKTLCEV EOC CVV TOV ovpavov Kat CVV T'T}V y'T}V. The later Syriac translators are also careful to render all Greek particles with some regular formal equivalent: thus JL'V and apa are both transliterated (this in fact goes back to the fifth century), TE and yE are represented by kit, etc. Possibly there are traces of the same concern in some of the Vergil papyri: for the most part -que is represented by Kat, but in a few of the texts TE is employed as a better formal equivalent. Since every detail of the original text is potentially significant, it is not surprising that these translators are careful to distinguish between features of the original that would seem to be purely a matter of stylistic choice. Thus verbs of saying in Hebrew can take either of two prepositions, t- and 'et. This offers no difficulties for the Greek translator, since he can represent l- by the dative and 'et by 1Tp6c; Syriac in turn is able to render the Greek dative by t- and 1Tp6c by twat, 'towards'. 28 More problematic are cases where the receptor language has only a single vocable where the source language has two. Hebrew has two forms of the first person pronoun, 'ani and , anoki; in order to bring out this distinction in Greek Aquila has to resort to reserving EyW for 'ani and representing 'anoki by EyW ElJLt, regardless of what follows syntactically. Whereas Hebrew parataxis can readily, if inelegantly, be repro­ duced in Greek, the reverse is not the case, and one can only stand back and admire the skill and ingenuity exhibited by the seventh­ century Syriac translators in their work: wherever feasible the prin­ ciple of formal correspondence is preserved. A few examples must suffice here. Hebrew into Greek had already made much use of the construction EV TW+ infinitive to render the Hebrew 'infinitive . construct' with preposition. Syriac has nothing corresponding to the Hebrew infinitive construct and so resorts to employing a preposition + hay ('" TO) + d- introducing a verb in perfect/imperfect/participle; 27 See Barthelemy, op.cit. (supra n.8) 16. F. Field had already drawn attention to the Homeric element in Aquila, Drigenis Hexaplorum quae supersunt (Oxford 1875) xxiii. 28 lwat with verbs of saying is a regular feature of the later translation literature only; in early translations and in native Syriac writing it is extremely rare. SEBASTIAN BROCK 83 thus E7T~ TO {3aetA.€VCUL becomes lwat hay d-namlek, 'towards/that/that he reign'. 29 The Greek verbs elvaL and EXELV have no formal equivalents in Syriac, and in subordinate clauses these posed particular problems. By treating Jit, 'there is, exists', as a formal correspondent to elvaL and Jit l-, 'there is to', to EX€tV, the seventh-century translators managed to indicate the Greek subjunctive: John 14:3 (Harklean) (tva 07TOV €ill-i. eyw Kai. VIl-€LC) ~T€: ... tehwun "itaykun, 'you shall be/you are' (normal Syriac would have tehwun alone). John 3: 15 (Harklean) (iva 7Ta.C 0 7TLCTEVWV EV aimp) €Xn 'w~v alwvLOv: ... nehwe Jit leh bayye dal-calam, 'there shall be/there is/to him/life/which is for ever'. The second example gives the appearance of being a soloecism in Syriac if one goes by the normal Syriac idiom for 'he will have life' (nehwun leh ~ayye, 'there will be/to him/life'), for in the Harklean passage quoted ~ayye (plural) appears to be the grammatical subject to nehwe; what has happened, however, is that nehwe is simply used as a tense indicator, with no grammatical subject, and 'it leh has been taken as a syntactical unit equivalent to EX€LV, with ~ayye treated as the object, totally against normal Syriac usage. That this is the correct explanation is shown beyond a doubt by the occasional further extension: John 12: 8 (Harklean) TOVC 7TTWXOVC yap 7T/XVTOT€ €X€T€ Il-€(}' eaVTWV, EIl-€ S€ OV 7TCXVTOT€ EX€T€: l-meskine geT bkul;;:ban 'it lkun camkun. Ii den law bkulzban 'it lkun, 'the poor (prefixed by object marker) /for/at all times/there is/to you/with you; /me/but/not/at all times/there is/to you'. (Here meskine, which according to Syriac idiom should be the subject of 'it lkun, has the object marker 1- in order to correspond to the Greek accusative after €X€T€.) 30 One further syntactic calque involving EX€LV with an adverb is worth noting: 31 V7T0ll-0VTJTLKWC €XHV 'to behave enduringly': da-msaybTana'it hwa 'it [hun, 'enduringly/was/there is/to them'. (Here again the Greek construction is reproduced mechanically, totally against Syriac 29 For the benefit ofnon-Syriacists I give deliberately atomistic renderings of the Syriac. 30 Other examples in the Harklean will be found at Acts 9: 3 I, 2 Cor. 7: I, Phil. 3: I 7 and James 2:1. Paul of Edessa uses exactly the same construction; see Brock, op.cit. (supra n.25) 39. 31 See Brock, op.cit. (supra n.25) 39. 84 ASPECTS OF TRANSLATION TECHNIQUE idiom; hwa is simply introduced as a tense indicator, required by the context.) LEXICAL FEATURES; TECHNICAL TERMS Technical terms naturally represent a very important category, and the three possibilities open-transcription, etymological trans­ lation (usually a neologism) or cultural equivalent32-are well exemplified in the Greek documents of the Republican administra­ tion in the east, which use all three methods, the choice depending on the particular character of each term: thus dictator is transcribed, designatus becomes a7TooEoELYJLEVOC, while quaestor is equated with nX-JLlac.33 Aquila usually opts for the second method, etymological translation: pesa~, 'Passover', which is transliterated in the Septuagint becomes {l7TEp{JacLc in Aquila; Urim and Tummim, rendered by O~AWCLC Kat aA~OELa in the Septuagint Exodus and Leviticus, are given the more etymological translations c/JWTLcJLot Kat TEAwncELc by Aquila; the divine name Shaddai, too, is understood by Aquila as being composed of s-+ day, 'which/sufficient', and accordingly he renders it by