Some Pitfalls of Translation Greek

The completion of a Greek-Hebrew Index of the Antiochene Text in the Historical Books is an excellent opportunity to go through the whole translation process and detect the most common mistakes, the main difficulties met by the translators and the mechanisms employed to overcome them. It is like looking at the reverse side of the weave, giving an insight into the high degree of literal and formal correspondence between the Greek translation and the Hebrew parent text in most of the historical books. At the same time it offers

the opportunity to appreciate the limits of the formal equivalence and, in some cases, to get a glimpse into the presumed Vorlage of the translators.
For the Greek, our edition of the Antiochene text has been used as the basis of the analysis, and for the Hebrew the text of the Biblia Hebraica Stuttgartensia (BHS) ^.We are aware that the Masoretic Text (MT) is not identical to the Hebrew Vorlage of the translators.We use it in the comparison for practical reasons, since it is the only complete Hebrew text available for those books.Moreover, it should be emphasized that the MT agrees to a large degree with the Vorlage of the translators, as can be inferred from a look at the Index in which the formal, extant equivalent, is of first concern.One may wonder why priority is not given to Qumran texts of Samuel.The fact is, that apart from the fragmentary character (ca.8% of the text of Samuel) and occasional agreements with the Antiochene text, there is little evidence to define the textual affiliation of the Qumran fragments •^.
Much has been written recently on the use of formal or presumed equivalents in an index or concordance.E. Tov and T. Muraoka have diversely criticized the Hatch & Redpath's procedure, because these scholars adhered very closely to the formal, almost mechanical equivalence in their Concordance ^.For the Index of the -N.FERNANDEZ MARCOS and J. R. BUSTO SAIZ, with the collaboration of M.^ V.^ SPOTTORNO DÍAZ-CARO and S. P. Co WE, El texto antioqueno de la Biblia Griega I-III, TECC 50, 53, 60 (Madrid 1989-1996), and K. ELLIGER and W. RUDOLPH (eds.),Biblia Hebraica Stuttgartensia (5th corrected ed., Stuttgart 1997).
^ «However, insufficient evidence was found to affirm any link between L and 4QSam", except for L's dependence upon LXX, which was in turn dependent upon 4QSam"»: cf.E. D. HERBERT, «4QSam" and its Relationship to the LXX: An Exploration in Stemmatological Analysis», in IX Congress of the International Organization for Septuagint and Cognate Studies.Cambridge, 1995, SCS 45, ed. 5^/64:2 (2004) SOME PITFALLS OF TRANSLATION GREEK 343 Antiochene text we have adopted a middle road.We adhere to the formal equivalence when it looks plausible after a scrutiny of the sentence in both languages.We believe that by using this procedure we are ensuring that the reader can make the best use of the Index without going down the path of the subjectivity of the presumed equivalences.It is common knowledge that the abuse of presumed equivalents, while they may be tempting, can lead to suggestions that can be analysed in different ways by scholars, thus producing different reconstructions.Nevertheless, in some cases and as an aid to the reader, the presumed equivalent preceded by the abbreviation leg (= legit) is suggested between brackets ^.
In our search for the correct correspondence, the meticulous study of both the Hebrew and the Greek texts leads us to some further considerations.In some cases it is extremely difficult to decide whether the extant Masoretic text reflects a new equivalent for the extant Greek or whether, in fact, the Greek is being translated from a different Vorlage.Indeed, our knowledge of the Hebrew and Aramaic as well as of the Alexandrian Greek is limited, and I subscribe to the sound statement of R. Smend that «Eine Konkordanz muss in der Gleichsetzung, soweit eine solche überhaupt durchführbar ist, so mechanisch wie moglich verfahren und das Urteil der Zukunf iiberlassen» ^.This appreciation is also valid for an Index.Both extre-" ^ Just as Abraham Tromm did in his Concordance published in Amsterdam 1718 (Abrahaini Trommii Concordantiae Graecae Versionis Vulgo Dictae LXX Interpreturn..., I-II [Amstelodami et Trajecti ad Rhenum MDCCXVIII]).
We hope that a mine of useful information has been added through the new Greek words and new Hebrew equivalents preceded by an asterisk in the Index.Likewise we suggest the presumed reading in a number of obvious equivalences signaled with an obelus by Hatch and Redpath: auxiiobÔrjç, 'dry' in 1 Sam 23:14.15does not translate niiD, but is a doublet of the unknown geographic name f^n; Poppâç is a stereotype equivalent for \\ù^\ hence, it can be presumed that in 2 Chr 14: 9 the translator read m^ù)i instead of nriDii; in 1 Sam 13:7 it can be presumed that the translator read oniVT as participle of nnv (oi ôiapaívovieç) instead of the substantive 'the Hebrews' (on:;iv)).^ R. SMEND, Griechisch-Syrisch-Hebrãischer Index zur Weisheit des Jesus Sirach (Berlin 1907) p. X. mes should be avoided: the inclusion of Hebrew words among the new equivalents whose meaning is well outside the semantic field of the Greek word ^, and the systematic exclusion of a new Hebrew equivalent because it is not attested in other parts of the Septuagint.The good number of new equivalents marked with an asterisk in our Index attests to the richness and variety of the translation manifested through several new plausible correspondences.These equivalents are lacking in the Hatch & Redpath Concordance, be it because the Antiochene Greek terms are only attested in the deuterocanonical or apocryphal books (some of them without Hebrew Vorlage), or because they appear in the three Jewish translators whose Hebrew equivalents are not recorded in this Concordance.Moreover, Hatch & Redpath follow the Greek text of the codices Vaticanas, Alexandrinus and Sinaiticus plus the Sixtine edition ( 1587), but they ignore the Antiochene text which is very different from these manuscripts in the historical books.
A careful use of the Index allows the user to draw certain conclusions in relation to the different problems of textual criticism.The stereotype correspondence between two terms in Hebrew and Greek may lead to the restoring of a different reading from that of the MT for the passage in question.Thus, in 2 Chr 33:7 aicov has been introduced in the Greek translation for the MT Oìb^v.Taking into account that 99% of the ocurrences of oblv have been regularly translated by aicov, it can be deduced, in all confidence, that the Greek translator of this passage also read obiv, as was the case in the other ancient versions, and, consequently, it can be restored as a genuine reading instead of the dubious and uncertain Dlb^v of MT.On the contrary, in 1 Chr 17:16 we come across a different text critical panorama.MT reads «and what is my house, that you have brought me thus far (obn-iv)?».The entire Greek tradition inter-prêts the last part of the sentence as scoç aicòvoç.But this reading results from a phonetic confusion between the guturais v and n and, consequently, cannot be invoked as a sound witness to change the reading of the MT that makes sense.The frequent occurrence of the expression ëcoç aicòvoç in the Greek Bible has contributed to consolidate this reading in the Greek transmission.This is just a sample of the kind of textual criticism that can be made with the aid of the Index, and which is valid for a high proportion of common, abstract and concrete names where an almost stereotype equivalence is recorded.Notwithstanding, the critical judgement is more difficult to exercise in other names (for which the translator liked the variatio or the metaphorical or stylistic equivalence), and more especially in the verbs, where the array of equivalents is highly diversified: for instance, sïôcoXov corresponds to no less than ten Hebrew words, and Xajiipávsiv translates eighteen different Hebrew forms ^.
I shall now move on and try to identify the incorrect readings of the Greek tradition, especially of the Antiochene text, in order to highlight the genuine reading.I will deal with the translation processs and the text transmission in a reverse order, going back from a) the paléographie errors of transmission (inner-Greek corruptions), through b) the different vocalization performed by the translator and c) the interchange or confusion of similar letters in the Hebrew Vorlage, to d) some variant readings supported by a different Vorlage.For the last case, the Qumran fragments of Samuel open a window towards actual readings of the Greek confirmed by an extant Hebrew Vorlage different from MT.
^ As Muraoka observes, «It is obviously ill-advised to attempt to establish mechanical patterns of correspondence between Greek tenses and those of Hebrew»: T. MURAOKA, «Translation Techniques and Beyond», in Helsinki Perspectives on the Translation Technique of the Septuagint, eds.R. SOLLAMO andS. SlPlLA (Helsinki -Gottingen 2001) pp. 13-22: p. 20. 5^/64:2 (2004) A. INNER-GREEK CORRUPTIONS Using the parent text as a control, some Greek corruptions can be detected that have contaminated a part of or the entire manuscript tradition.A few examples taken from the Antiochene text, shared occasionally by the whole Greek tradition, will suffice to illustrate this phenomenon: -Aï^,-y6ç is the regular equivalent for the Hebrew iv 'goat'.However, in 2 Chr 31:6 we come across a formal equivalent of aiyoev for the Hebrew DWlp in the sequence «the tithe of cattle and sheep, and the tithe of the dedicated things that had been consecrated to the Lord their God» ^.In all probability the whole Greek tradition has been corrupted from àyícov to aiyoev.However, following the manuscript tradition we have restored aiyoev as did A. Rahlfs in his manual edition.The reason why this new reading, so alien to the original meaning of the Hebrew, succeeded in the text reception, is that it makes sense also in the Greek chain of words joined to the cattle and sheeps: Kai aòxoì iívsyKav STTiôéKaxa jiióaxcov Kaì TipoPáxcov Kaì éíciSsKaxa aiyoov, Kaì fiyíaaav xoò Kupíco Bsoe auxoev.
-The Antiochene reading of 2 Chr 16:14 must be characterised as an inner-Greek corruption: Kai SKXauaav aòxcò KÀ^aCaiv |Lisyá/lT|v for the Hebrew nbn> nonvy ib-iDn\yn ('and they made a very great fire in his honor').KaÍ£iv and Kaûaiç correspond better to the meaning of the Hebrew root ^n'\y, while K?taisiv translates regularly the root riDi.However, the paléographie confusion 8K?taüaav / SKauaav and KXaûaiv / KaCaiv, easy to detect in the cursive Greek writing, provoked the new reading in the Antiochene branch of manuscripts '°.The fact that the new reading makes sense in the context of the verse lead to its consolidation within the Greek tradition.In fact it is a doublet or alternative translation that Antio-^ The English translation of the Hebrew, when no otherwise said, is taken from B. M. METZGER and R. E. MURPHY (eds.),The New Oxford Annotated Bible with the Apocrypha (New York 1989).
'° Interestingly, the correct alternative reading Kai CKauaav aùiòv Kaûaiv jicyáÀiiv has been preserved exclusively in the Alcalá polyglot Bible.chene added to the reading of the current Septuagint resulting in the following sentence: Kai 87ioir|aav auioò ¿Kcpopàv \xzyòXr{V Kaì 8KXai)aav auxoD K^aCaiv jueyá^riv.In a text conceived for public reading it is essential that it has meaning.On several occasions the doublets of Antiochene fulfil this function by completing the sense or clarifying the context by means of an alternative reading attached to the reading of the majority.The same phenomenon of a doublet or alternative reading based on a paléographie, inner-Greek, corruption can be detected in the Antiochene text of 2 Chr 21:19: Kai OÙK ¿Tioíriasv aÙTOÒ ò À-aòç aûxoC èKcpopàv Kai Kka\)üiv Kaià Tf|v K^^auaiv TODv Tiaxspcov auioû.Again, this alternative reading has succeeded in the text transmission because it fits the context of mourning for Asa's death.
^" This term is the right Greek translation in the parallel passage of 2 Kings 22:14.Did the author of the Antiochene text take loO íiJ,aTio(pú?iaKoçfrom this parallel passage?There is no trace of Hexaplaric reading to 2 Chr 34:22; Field's reference points to alia exemplaria, in fact the reading of the Complutensian Polyglot which follows the Lucianic manuscript 108: cf.F. FIELD, Origenis Hexaplorum quae supersunt I (Oxford 1875 = Hieldesheim 1964).Tf|v (pvXáaaovaav xàç èvxoXáç.The reading of IDW as feminine participle by the translator (joining to the participle the article of the following word) generated an embarrassing interpretation in the target language and probably contributed to the succcess of the corrupt reading èvxoXáq instead of axoXáç.
-When the queen of Sheba visits king Solomon, she contemplated and admired all his wisdom and among other things in his palace she was amazed by the clothing of his servants, and by his cupbearers (vpv)Di, 1 Kings 10:5).The current Septuagint translates this part of the sentence xòv í|iaxia)nòv aùxoû Kai xoî)ç oivoxóouç aùxoC, in exact correspondence with the meaning of the root npv) in hiphil, 'give to drink'.However, in Antiochene we come across xòv í|Liaxia|Liòv aùxoC Kai xoi)ç eóvoúxouç auxoû ('his clothing and his eunuchs').It is the reading of Antiochene without variants, with a meaning far different from the original.It is not plausible to imagine a paléographie confusion at the level of the Hebrew between n¡7^D, the regular equivalent for oivoxóoç and ono, the stereotype equivalent for euvoûxoç.However, at the level of the Greek transmission the phonetic corruption by similar pronounciation of both terms due to the phenomenon of itacism provides a reasonable explanation.The reading also makes sense, and is consolidated in the text transmission of the Antiochene family of manuscripts ^'\

B. A DIFFERENT READING OF THE CONSONANTAL TEXT
The numerous passages characterised as aliter in the Index provide some information on misleading translations caused by a different reading of the Hebrew text and, occasionally, by homophonic translation.I emphasize that it is a typical phenomenon that occurs  -In 1 Sam 2:31 the Hebrew word :^' ny, 'arm', is read twice as V3J, 'seed'.The Hebrew sentence «See...I will cut off your arm and the arm of your ancestor's family» becomes in Greek Kai i8oì)... è^oXoGpeúaco xò arcépina aou Kai xò aTiépiua xou OIKOU XOC Tiaxpóç aou.
-In 1 Sam 15:9, the different vocalization plus the confusion of similar consonants leads to a new diverse sense in the target language quite different from the parent text.Saul and the people spared Agag, and «the best of the sheep... and the lambs (o^lDn), and all that was valuable».The plural of ID, 'lamb' or 'ram' is read and interpreted by the whole Greek tradition as xoov ajuTusXcovcov = D'>)D'i|, the plural of 0*15.
-In 1 Sam 16:20 the Hebrew onb mnn ^v)-) npn («And Jesse took a donkey with bread») is translated in Antiochene: Kai zXa^zv 'Isaaai ovov, Kai cTiéOiiKev aúxoò y^M-op âpxcov.In all probability this sentence arose from the double translation of a single word niDn with different vocalization as donkey ('liDq) and as a measure (nD n).Antiochene utilizes this recourse to double interpretation in order to solve the brachilogy of the Hebrew; the majority text of the Septuagint understood it as yó|uop.But only Antiochene makes a difficult sentence in Hebrew explicit.Interestingly, the Old Latin retains the Antiochene reading: Et accepit lesse asiniini et imposuit super gonior panis ' ' ^.
^'^ The Old Latin can be consulted in the apparatus of our edition of the Antiochene text quoted in note 2.An alternative reading of the consonantal text may produce a double interpretation that Antiochene incorporates willingly into the textual chain, as in 1 Kings 18:44, where the MT o>n, 'out of the sea', has been translated twice by Antiochene: uSoep aTcò GaXáaarjç ^^.
-The different vocalization of the MT may result in the interpretation of some proper names as common names or verbs, with a sense far from that of the original.In 1 Kings 6:3 (= MT 5:32a) it is stated that in the building of the temple «Solomon's builders and Hiram's builders and the Giblites did the stonecutting» (ibt7D>i o>!7i:\ni Dn>n >m riDbv) >n), translated by Antiochene: Kai fjvey-Kav oí üíoi Zo/lo|LioovToc Kai oí uíoi X8ipá|Li, Kai èvépallov aÓToúç («and Solomon's sons and Hiram's sons brought the stones and fashioned their borders»).The majority text of the LXX reads èneXsKriuav ('did the stonecutting') instead of fjveyKav, and puts the simple verb ëpallav instead of èvépaXov.But, what is more important, the translator read with different vocalization 'sons' (>Í5), not 'builders' (^A'l), and interpreted the proper name in the plural 'Giblites' '^ as a hiphil of bn>, 'circumscribe', probably read-^^ MT: «A little cloud no bigger than a person's hand is rising out of the sea» (nbv D>)3).The Old Greek àvótyouaa i35cüp probably read o>a nb^JD.Antiochene conflates both readings and obtains a meaningful sentence.
^^ People from Gebal called Byblos by the Greeks.ing Oib^^n, 'they fashioned their borders', a different reading accepted as emendation to the MT by some modern dictionaries ^^.
The homophonic translation may also explain some unusual equivalences in the Index.In these examples there is no reason for postulating a paléographie confusion at the level of the Hebrew language.However the similar phonetics of the Hebrew and Greek word may have influenced the selection of terms in the translation process in passages such as 2 Chr 33:6 sv yi] B6V8VVÓ|LI for the Hebrew 0)n"il *>>! («in the valley of the son of Hinnom»), or 2 Chr 30:10 èv TOD opsi 'E9pái|Li Kaì Mavaaax] for the Hebrew DnDN""^(nNi nv)3ni («in the country of Ephraim and Manasseh»).A phonetic connection exists between n and yf], ^IN and õpoç in Hebrew and Greek that might reasonably explain these uncommon translations.There may also be an underlying, diffuse conscience among Hellenistic Jews that Hebrew and Greek had something in common ^^ In 2 Chr 12:11 the guard of Rehoboam, whenever he went into the house of the Lord, would come along bearing the shields of bronze, «and would then bring them back to the guardroom» (NÎT^N Oìiv)m o>i¿in).The verse has been diversely interpreted by the Greek tradition.It is clear that the word Nn, 'guardroom' was not transparent for the translators ^^.They resolve the difficulty with a puzzling translation; the majority text of the Septuagint reads Kaì oí 87i:iaTpé(povT8ç sic àTiávTiiaiv xoev TiapaxpcxóvTODv.But the homophonic translation appears clear enough in the double sentence of Antiochene that includes 8Íç Tf|v xá^iv xcov Trapaxpexovxoev, a guess translation induced by the phonetic similarity of xáÇiç with HT\ ^°.Hatch and Redpath insert an obelus instead of the Hebrew equivalent of the septuagintal àTiávirjaiç, although the formal equivalence is beyond doubt.

C-INTERCHANGE OF SIMILAR LETTERS
Another source of misleading translations lies in the confusion of similar letters or groups of letters.The Index provides a mine of information on unusual equivalences going back eventually to a misreading of some consonants in the early square script.It is an accident of reading or copying; in the first case it arises in the course of the translator's deciphering of the Vorlage; in the second, it reflects a Vorlage already at variance with the MT.It is not to be excluded that a genuine textual difference underlies some of these variants, but in general it can be stated that the paléographie confusion at the level of the Hebrew script is the most plausible explanation.The most frequent interchange of similar letters occurs between n/i.
-In 1 Sam 23: 15-16.18-19, the city where David remains hidden in the wilderness of Ziph, Horesh (nv)nn), is translated systematically in Antiochene by Kaivi^, obviously read as nv>in.
-In 1 Sam 19:13.16 the uncertain Hebrew expression onvn n^lD, translated commonly as 'net of goat's hair' (Vulgate pellem pilosam caprarum), is interpreted in the whole Greek tradition as i^Tiap aiyoDv, by reading the first term as 115, iiver'.
-In 1 Sam 24:3 Saul went to look for David and his men •>DD b^ 0'>bv^n m:^, ususally translated as «in the direction of the Rocks of the Wild Goats».The majority text of the Septuagint reads STIÌ TCpóacoTCOV 'Eôôaié|Li, that is, a transliteration (cum variantibus).
"° In the parallel passage of 1 Kings 14:42 (= MT 14:28), the majority text of the Septuagint transliterates 0££, and Antiochene OsKoüe.-In 1 Sam 14:40 Saul says to all Israel: «You shall be on one side, and I and my son Jonathan will be on the other side» (*iivb mH nivb ...inN).The whole Greek tradition transmits in both cases £Íç Soü?t£Íav, reading 12V instead of niv.Interestingly, the Antiochene text adds, as a doublet, a new sentence with the correct sense according to the MT: Kai SÍTIS I^aohX Tipòç xòv Xaòv 'Y)LI£ÎÇ eGSuQe eiq êv |Liépoç, Kai éyò Kai 'IcovaGàv SGoixsQa 8Îç êv JLispoç.The alternative reading, in agreement with MT, is not supported by any Hexaplaric witness, and we are probably dealing with an early correction, already known to Josephus ^^ The double reading SouX-eíav/juiépoç, based on the interchange of i/n generated a new sentence.As is well known, a trend of the Antiochene text consists of joining double readings with small redactional retouches to clarify the meaning so that all the information of the preserved variants can be explicit for public reading.
The misreading of other graphically similar letters like >/i, i/\, n/D; D/n, Ù/2, T/i, ^/y, letters with similar phonetics like the sibilants Ì, V, \J, V); or the gutturals H, n, n, v, is also reflected in the Josephus, Ant.VI, 125: laxaiai ôè Kai aòiòç aùv TÔ) Tiaiôî Kai' àXXo juépoç.In 1 Kings 21:38 (MT 20:38), the prophet waited for the king of Israel along the road, «disguising himself with a bandage over his eyes» (v)^^"!:?^ noNi \yDnnn).The entire Greek tradition reads this part of the sentence Kai Kaxeôéaaxo év T8?ta|Liouvi TOÎ)Ç ócpOall-|Lioí)ç auToC.Hatch and Redpath give as the Hebrew equivalent of KaxaSsiv the hithpael of V^DD with a question mark.Muraoka ^^ put this root between double brackets signifying that the equivalent given by Hatch and Redpath is implausible.He pointed with an arrow to the qal of DDV as the true equivalent that should replace that of Hatch and Redpath.However, I think it is more plausible that the translator read in this passage the qal of v)nn, regularly translated in the Septuagint by Ô8Îv, Kaxaôsîv.The confusion of D/n in the Hebrew script is frequent and also between the sibilants v/\), while ÜDV in qal is regularly translated by aipeiv, éTiiysiLií---T.MURAOKA, «A New Index to Hatch and Redpath», ETL 74 (1998) pp.257-276; E. Tov, «Interchanges of consonants between the Masoretic Text and the Vorlage of the Septuagint», in Sha'arei Talmon, eds.M. FiSHBANE and E. Tov (Winona Lake 1992) pp.255-267, and A. GELSTON, «Some Hebrew Misreadings in the Septuagint of Amos», VT 52 (2002) pp.493-500.
-^ T. MURAOKA, Hebrew/Aramaic Index, p. 54.Csiv and, in my opinion, its confusion with v^Dn is less probable.In any case, it is just an example of how the presumed equivalents can be seen differently by diverse scholars.
The different reading based on paléographie confusion may affect not only isolated consonants but also a group of letters, the phenomenon of metathesis included: -In 1 Sam 8:16 the Hebrew reads «He will take your male and female slaves and the best of your young men (oDmni'TiNi, iuvenes óptimos in the Vulgate) and donkeys», while the Greek tradition interprets: Kai TOÎ)Ç SOUXODÇ \)\x,(hv Kai xàç ôouX^aç \)\x(bv Kai xd PoüKÓXia \)\x(bv xà àyaBà Kai xoùç õvouç ÙJLIOÒV.In view of the regular equivalence between npi and POUKOA^IOV, it can reasonably be presumed that the translator read ODnpi"TiNi. -In 1 Chr 22:9 the king Jehu searched for Ahaziah, «who was captured while hiding in Samaria» (^nDV)!NinxiD Nim iniDb>i).But the Greek tradition interprets unanimously: Kai KaxéÀ.aPovauxòv íaxpeuójLisvov èv Sajuapeía.Hatch and Redpath insert an obelus of uncertainty by iaxp8üó|Li8Vov as equivalent of Nin.However, given the regular equivalences of Kpúpeiv, KpÚTixsiv for the hithpael of Nin, and, likewise, the regular equivalence of iaxpeusiv for the hithpael of Non, it can be presumed that the translator read NDnriD.
Moreover, the confusion of final ì and final ^ leads to a quite different interpretation in the Greek of 1 Sam 28:14, when the woman medium evokes Samuel's spirit to Saul.To Saul's question concerning Samuel's appearance, the woman answers according to the MT: «An old man is coming up; he is wrapped in a robe» (v)>N b>v)3 nuv Nini nbv ^pT).However, Antiochene translates with the rest of the Greek witnesses "' ^: âvôpa õpBiov àvaPaívovxa ÒLTÍÒ xf|ç ~^ Only the manuscripts AN followed by a few cursives transmit õpGpioç, 'of the morning' or matutimis: cf. A. E. BROOKE, N. MCLEAN and H.St. J. THACKERAY, The Old Testament in Greek.Part I, I and II Samuel (Cambridge 1927).
(c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.esyfiç, àvaPsPÀ,r||iévov ôiTiXoíôa."OpBioç is a hápax of the Septuagint in this passage.The Hexaplaric witnesses represented by oí À.oi7roí read, according to the MT, npeG^mr\v.But õpBioç is used by Symmachus for the translation of Gen 1:27, the man's creation ^^.In this passage of Genesis Symmachus inserts an explanatory note relying probably on an exegetical tradition that emphasizes the most peculiar feature of the human being in contrast with animals, his upright stance, a tradition that can be traced back to Justin Martyr and other rabbinic sources ^^.In contrast, in 1 Samuel 28:14 it seems that the origin of the Greek reading is not exegetical but paléographie.The verb ^p\ is translated by ávop9o6v in the two occurrences of the Bible (Psalm 144:14 and 145:8).Moreover, it is well attested with the meaning of 'stand upright, erect' in postbiblical Hebrew as well as in Aramaic and Akkadian ^^.In any case, an exegetical tradition may have influenced this version since, according to the Midrash, when the spirits of dead people are evoked from the netherworld, only the kings appear upright, face first; the other persons rise feet first.This is, no doubt, why the woman recognised Samuel ^^.
Metathesis can be detected in some unusual translations, but it is especially visible in the transliteration of proper names.
In 2 Chr 28:3: Kai Sifiyaye xa xéKva auxou sv Trupi, for the MT \!)Nn i^n-nN *ivi'>i reflects a different reading from the verb v)Ki ...niv">i, in hiphil, a stereotype expression for «make pass through fire».In 2 Sam 22:13 it is said that «coals of fire flamed forth» (\i)N-' >bn> nvi).The cuiTcnt LXX translates literally: è^8Kaú9r|0"av âv6paK8ç Tiupóç.However, Antiochene interprets the whole sentence as ôif|X-9ov xáX-aCa  I believe that most of the commented phenomena can be explained as misreadings during the process of translation due to the incorrect desciphering of the Hebrew Vorlage.Consequently, they are of secondary character arising from an accident of the transmission, be it in the copying of the Hebrew text itself or produced by a misreading of the translator.It cannot be excluded, however, that some of these variants conceal a genuine reading.

D. TRACES OF A DIFFERENT VORLAGE
It is common knowledge that the Antiochene text is rooted in the Hebrew not only as part of the Septuagint tradition, but also due to the fact that it incorporates a set of Hexaplaric corrections according to the MT.Sometimes it is even closer to the MT that the rest of the Septuagint tradition.Moreover, S. Brock realized that not all the approximations to the Hebrew in Antiochene were of Hexaplaric provenance ^^.Thanks to the discovery of the Qumran documents for Samuel this statement has been confirmed.There are a few Antiochene deviations from the MT that are supported by 4QSam^ The relationship between the textual witnesses of the book of Samuel is very complex and, therefore, it is dangerous to make any kind of generalization.On the other hand, only with the full publication of the fragments and a thorough comparative study of all the witnesses can the net of relationship be ascertained.Provisionally, it can be stated that 4QSam'' was not the Vorlage of the Antiochene text; the lack of secondary agreements or conjunctive errors between both texts do not allow such a close relationship to be established ^\ For our purpose it will suffice to point out some agreements of Antiochene with 4QSam'^ leaving a full comparison of both witnesses for a further study.
-1 Sam 5:9: «And it ocurred that after they had brought it [the ark of God]» (iriN iipn nnN ^n>i), in the majority text of the Septuagint the translation is Kai éysvfiSii ixsxà xò fiexe^^Geîv aÚTf|v.However, in Antiochene we come across the following interpretation: Kai ¿yevexo ¿v xoe iiexeXQslv xf|v KIPCOXÒV Tipòç xoi)ç Y899aíouç.This version makes explicit the noun of the ark, translated literally by the pronoun in the Septuagint, but, what is more important, it mentions Gath (xoi)ç yeGôaíouç) as read in 4QSam'' (nn:^ 110 nnN ^n-^i), but absent in MT ^l -2 Sam 12:16: When the Lord struck the child that Uriah's wife bore to David, the king fasted «and went in and lay all night on the ground» (ni¿nN 1DV>1 p^ HII).The Vaticanus and his group of In two other cases, the reading underlying the whole Greek tradition is witnessed in Qumran, not in the MT: 1 Sam 2:8-9 the use of soXoyeïv in the Septuagint is transparent of the Qumran reading *jnn' >i ^^, not of a different or corrupted MT.And in 1 Sam 2:20 the current text of the Septuagint with àTioxíveiv as well as the Antiochene variant with àvxaTioSiôóvai are supported by the Qumran reading obv^^ ^^ instead of the ov^^ of the MT.' ATCOTÍVSIV and àvxaTioSiSóvai are regular equivalents for the piel of obv) in the Septuagint, while these two verbs are never used for ow.
These agreements between the Greek text, especially the Antiochene, and an extant, non-Masoretic, Hebrew, lead us to the conclusion that, in all probability, several other deviations of Antiochene are also rooted in the Hebrew.In this context I would like to point out a series of doublets in the Antiochene text whose origin can only be explained at the level of the Hebrew, a Hebrew text different from the MT.Such cases also confirm, from another perspective, that the Antiochene text is rooted in the Hebrew.A typical example will serve as an illustration: -In 2 Kings 2:23 while Elisha was going up on the way to Bethel, «some small boys came out of the city and jeered at him, saying (i!? nDNn in-ipt^pnn), «Go away, bald-head!Go away, baldhead!».The current Septuagint renders literally: Kai Tiaiôápia |iiKpà é^f|/^9ov èK xfiç TióÀ-scoç Kai KaxéTiaiCov aùxoû Kai SÍTCOV aùxoe.Notwithstanding, Antiochene emphasizes that the boys not only mocked him but also threw stones at him: è^fjXBs Tiaiôápia jLiiKpà SK xf|ç TióXscoç Kaì éXíOaCov aùxòv Kai KaxéTcaiCov aùxou Kai eXsyov auxoò 'Avápaivs, (paXaKpé, àvápaive, (paXaKpé.The use of é|U7iaíC£iv, KaxaTiaíCsiv for the hihtpael of obp is consolidated in the Septuagint.The use of XiBáCsiv, XiQo^oXelv for all the forms of bpo is also well attested among the Greek-Hebrew equivalences.Consequently, it can be deduced that ^^ E. ULRICH,The Qumran Text,p. 119 and A. FiNCKE,The Samuel Scroll,p. 9. ^^ E. ULRICH,The Qumran Text,p. 72 and A. FiNCKE,The Samuel Scroll,p. 10. this curious doublet ultimately relies on a different Vorlage with the reading bpp, or on the extant MT read with metathesis of consonants by the translator.Interestingly, the Old Latin retains only this second interpretation of the Antiochene text: pueri pusilli exierunt de civitate et lapidabant ilium dicentes: Ascende calve, Ascende calve.

E. CONCLUSIONS
Through the lens of translation, particularly of the Antiochene text in the historical books, I have tried to point out some of the pitfalls that may have occurred in the process of translation and transmission.An awareness of these mistranslations is the only way of correctly evaluating the Greek variants for the restoration of the genuine text.Some mistakes have been produced, such as inn^r-Greek corruptions, through the frequent copying of the manuscripts.Several mistranslations arose as a result of a different vocalization on the side of the translators.Other variant readings were produced by the confusion of similar consonants or groups of letters; these variants or alternative readings can be explained only at the level of the Hebrew.And finally, in a few cases, an extant, non-Masoretic, Vorlage has been detected in the Hebrew fragments of 4QSam^ These agreements open a window toward a textual stage when different Hebrew texts were in circulation.The Vorlage of the Septuagint (Old Greek) was one of them.MT is the only complete Hebrew text available, but we must be aware that the Greek tradition, when it deviates from the MT, may conceal another text, with a striking resemblance, but not identical to the MT.Some scholars maintain that the Vorlage of the Septuagint in the books of Kings is older and probably more genuine than the MT.
There are numerous passages in the Index preceded by the mention of the mark aliter.These draw our attention to the specific texts which should allow a continuous exercise of textual criticism
(c) Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.esSef6A:l (2004) SOME PITFALLS OF TRANSLATION GREEK 349 in the translation process at the first level of encounter of the two languages.Translation is a kind of reading and concretely the Septuagint is the first interpretation of an unvocalised Hebrew text.It is a kind of performance of the consonantal text, like a score, to use a musical metaphor.No doubt, in several cases it is clear that the translators were following a different reading tradition or an exegetical device, but in many other cases the end product can be analysed simply as a misreading.

( c )
Consejo Superior de Investigaciones Científicas Licencia Creative Commons 3.0 España (by-nc) http://sefarad.revistas.csic.esIndex.These phenomena have been recently dealt with by T. Muraoka, E. Tov and A. Gelston ^^.The examples abound, especially in the transmission of the proper names and other transliterated words.Herewith a handful of illustrations: