Before commencing with a large-scale comparison of Aquila, Targum Onqelos, et al. in my in-progress thesis,¹ I sought the most complete contiguous texts of the Three available. In the case of Aquila and Symmachus, this naturally led me to the Hexapla Psalter palimpsests, Ra 1098 and Ra 2005.² Taylor published the editio princeps of Ra 2005 in 1900 and Mercati that of Ra 1098 in 1958.³ Beyond these, an appendix in Busto Saiz’ published dissertation establishes a critical edition of Symmachus in the Psalter;⁴ he emends the texts of Ra 1098 and Ra 2005 using the likes of Field and Mercati’s «Osservazioni».⁵ Much more recently, Carrera completed a dissertation providing a new, diplomatic edition of Ra 1098 based on enhanced technological imaging.⁶ Whereas Mercati’s edition presents a slightly edited version of Ra 1098, somewhat in the vein of Rahlfs’ Septuaginta, Carrera transcribes the text as is without editorial corrections.⁷ He informs us that he will at least briefly discuss all differences between himself and Mercati.⁸ Thus, for Aquila, one only has Taylor’s edition in the case of Ra 2005, but one must choose between Mercati and Carrera for Ra 1098. In the case of Symmachus, one can begin with Mercati or Carrera for Ra 1098 and Taylor for Ra 2005 or, more simply, use Busto Saiz for both Ra 1098 and 2005. The scholar who does not start from scratch still has many decisions to sort through!

In the process of comparing every line of Aquila and Symmachus in the relevant passages provided by Field, Taylor, Mercati, Busto Saiz, and Carrera, I eventually opted to take the excellent work of Busto Saiz as my point of departure for Symmachus⁹ plus Mercati and Taylor for Aquila. Accordingly, my goal is to present improvements in the readings of Aquila and Symmachus that I have made to these starting points. Unlike the ongoing labours of the “Editio critica maior des griechischen Psalters” and Hexapla Project, I collect no new readings; I am largely at the mercy of Mercati’s «Osservazioni», which surpass Field in scope and specificity. Further, I am not comprehensively representing all readings of Aquila and Symmachus available: I only discuss what I believe to be the preferable readings of Aquila and Symmachus for the passages attested by Ra 1098 and Ra 2005 if I think the original readings differ from Mercati and Taylor’s editions of Aquila or Busto Saiz’ edition of Symmachus. By implication, my silence is meaningful – it indicates my judgment that my base texts are secure against whatever alternatives are presently available to scholars as per the above resources.

Conventions and Qualifications

The presentation of my critical work below follows a modified format of the Hexapla Project that Meade’s edition of Job 22–42 pioneers.¹⁰ I do not label my discussions “Notes”. My observations on translation technique are kept to a minimum, except where absolutely necessary for reconstruction. The rightward arrow (→) represents “translated as.” The Hebrew text (“HT”) comes from Accordance’s “Hebrew Masoretic Text with Westminster Hebrew Morphology (HMT-W4),” version 2.2, which I believe is preferable to BHS. In the case of the “LXX,” the text comes from Rahlfs’ Psalmi cum Odis,¹¹ with all its accompanying contingencies.¹² The versification follows the Heb. first, followed by Grk. in parentheses.

To minimise lengthy references, the last names of Field, Taylor, Mercati, Busto Saiz, and Carrera entail their editions.¹³ I refer to Mercati’s second volume of Psalterii hexapli reliquiae as the «Osservazioni»¹⁴ and, given its eminent value, I occasionally make reference to the CTAT volume on the Psalter as “Barthélemy.”¹⁵ I do not cite page numbers when citing the texts of the editions, as the chapter and verse adequately entail the location. I do, however, always cite page numbers in parentheses for the «Osservazioni» and Barthélemy as well as when Mercati has footnotes in his edition (i.e., not «Osservazioni»)¹⁶ or when I am citing from Carrera’s commentary. For convenience, I state that Ra 1098 gives attributions to αʹ or σʹ, while this is only implicit by columnar sequence. The sigla for the manuscripts follow Rahlfs numbers (where assigned) rather than, e.g., text-group or institutional shelf marks.¹⁷

The presentation of my work always encompasses Field, Mercati, Busto Saiz, Carrera, the «Osservazioni», and Barthélemy. As I am providing no new manuscript readings, the «Osservazioni» are particularly crucial because they give more detailed manuscript information than Field was able to provide.¹⁸ That said, only Field habitually quotes the Syh readings, from whom they invariably come. While in the notes I discuss salient differences in the readings revolving around dubious letters, implicit information, iota adscript versus subscript, etc., in the critical representation I omit such details. More significant emendations will become apparent in my comments.

There are certain verses of Ra 1098 where the text represents obvious scribal errors that do not warrant a full collation of all manuscript evidence. On the one hand, the scribe of Ra 1098 in several instances reveals dittography. Verse 28(27):7 is repeated in its entirety; I omit the second instance.¹⁹ Additionally, three minor instances of dittography occur in the column of Aquila. Mercati corrects only the second but I correct them all here rather than in fuller detail below: thus, read καὶ οὐκ εὔφρανας instead of καὶ οὐ καὶ εὔφρανας in 30(29):2b;²⁰ οὐκ ὄψεται instead of οὐκ ὄψεται οὐκ ὄψετ(αι)²¹ in 49(48):10b;²² and καὶ ἄν(θρωπ)ος instead of (καὶ) ἄν(θρωπ)ος καὶ ἄν(θρωπ)ος in 49(48):13a. On the other hand, the scribe omits an entire verse across all columns in 18(17):34²³ and 30(29):3; two lines across all columns in 46(45):11b; and one line across all columns in 18(17):29a. I have not undertaken to restore any of these. Beyond these scribal errors of Ra 1098, Mercati also has typos for Aquila which I correct here: in 35(34):19, read ψεύδ̣ους rather than φεύδ̣ους; in 49(48):11, read ἀσύνετος and εὐπορίαν instead of ασύνετος and ἐυπορίαν, respectively.²⁴

Provisional Critical Improvements to Aquila and Symmachus

18(17):34b Aq

The lacuna in Ra 1098 for ὑψώματά μου comes at the top line of the relevant folio, which is broken off across all columns in Ra 1098 (see Carrera, 189–90). Mercati represents traces of one acute accent (thus, καὶ ἐπι τὰ ⁎ ⁎ ⁎ ⁎ ⁎ ⁎́ ⁎ ⁎ στήσει με·). Both manuscripts attest the added article, which should be kept despite the doubts expressed in the «Osservazioni» (11).

18(17):36a Sym

Faithfully following Field, Busto Saiz gives the text of Symmachus as σωτερίας [sic] σου and states in his apparatus, “σου 1ᵒ] F(οἱ πάντες); μου O.39; ךָ- TH.” Field presents the corrupted attributions as he himself received them: “Οʹ. ὑπερασπισμὸν σωτηρίας μου. Οἱ πάντες· ὑπερασπισμὸν σωτηρίας σου. Sʹ [i.e., ϛʹ]. ὑπερασπιστῆρα τοῦ σωτῆρός σου.” The «Osservazioni» (12) suggest that the pertinent “Vat.” manuscript, Ra 1175, gives only a partial scholion to highlight the markedness of the lexeme ὑπερασπισμός. While other attributions require disentangling,²⁵ the reading of Symmachus is secure.

30(29):6d Aq

Mercati (23 and n.) reconstructs ⟨αἶνος.⟩ and comments that the circumflex is clear, while Carrera (223) transcribes α̣ἶ̣ν̣ ̣ ̣ and supplements with photos. We follow the text as both 264 and 1098 record it with πρωῒ, though the «Osservazioni» (77) suggest Aquila more likely translated πρωΐα.

30(29):10d Aq

As all the other columns of Ra 1098 translate -הֲ … -הֲ as μὴ … ἤ, one can too easily conceive of the copyist overlooking the nu and writing ἤ like for the others. Thus, follow Ra 264 (so «Osservazioni», 82).

31(30):20a Aq

The beginning of this folio of Ra 1098 does not preserve the first word of the verse. The several witnesses besides Ra 1098 that lump the attributions of Aquila and Symmachus together continue beyond the portions excerpted above, revealing similarities plus overlooked differences. Nevertheless, Aquila and Symmachus in all likelihood translated this first line identically, except for the latter adding the article which the former did not (see «Osservazioni», 126).

31(30):21b Sym

Ra 1098 leaves the ending of ἀντιλογι- implicit. Without further comment, Mercati and Carrera both transcribe ἀντιλογι(ῶν) as a plural, matching the Secunda’s μριβη, and Busto Saiz follows suit with ἀντιλογιῶν. This is entirely unexpected. The previous line of Symmachus has the singular παραδειγματισμοῦ for HT’s plural מֵרֻכְסֵי, suggesting Symmachus (like οἱ οʹ εʹ) and the Vorlage or reading tradition of the Secunda both harmonised in opposite directions: the Secunda with the plural μριβη (hence מֵרִיבֵי) to match מֵרֻכְסֵי, and Symmachus (sim. οἱ οʹ εʹ) with the singular παραδειγματισμοῦ (hence מֵרֹכֶס) to match מֵרִיב. Thus, read ἀντιλογίας.

32(31):9b–c Aq

Rahlfs 264 does not conflate Aquila and Symmachus in the way one would expect from Field (see «Osservazioni», 168–70).²⁶ The Syh does indeed combine them, as if translating (in my judgment) διὰ κημοῦ καὶ χαλινοῦ περιθέσεως τοῦ ἐπιστρέψαι οὐ μὴ ἐγγίσῃ πρὸς σέ. The «Osservazioni» (170) fittingly suggest an initial τοῦ fell out due to haplography with the preceding αὐτοῦ. We can represent this as a retroversion from Syh’s prepositional clitic rather than reconstructing it in brackets. Finally, the reading of Ra 1098 is represented above as ἐπιστρέψαι instead of Mercati’s ε⁎⁎επιστρέψαι because the “prefix ‘εἰσ-’ is not visible” according to Carrera (265).²⁷

The «Osservazioni» (170–71) prefer περιστρέψαι of Ra 264 over the idea of Aquila translating the Heb. hapax legomenon לבלום → ἐπιστρέφειν, as Aquila regularly translates לשׁוב → ἐπιστρέφειν but only uses περιστρέφεσθαι²⁸ once in translating the hitpael להתהפך (Job 37:12). But if Aquila indeed never uses the same Greek word to translate two different Hebrew words, then even using περιστρέφεσθαι once elsewhere is too much; we would have to conjecture that Aquila here used a third, unattested verb.²⁹ The best solution is the simplest: an original ἐπιστρέψαι was likely changed to περι- due to influence from Symmachus’ περιθέσεως.

32(31):11b Aq

In what could be a telling indicator of dependence, Ra 264 gives two quotations of Aquila: the first follows Ra 1098 (αἰνοποιεῖτε) and the second Ra 1175 (αἰνοποιήσατε).³¹ The previous two verbs of Aquila in 32(31):11a are εὐφράνθητε and ἀγαλλιᾶσθε; while the first is clearly subjunctive, the second is imperative but could be parsed as subjunctive. Plausibly, αἰνοποιήσατε originated through a scribe interpreting ἀγαλλιᾶσθε as subjunctive and then modifying ἀγαλλιᾶσθε to match.

The primary reason for including this stichos is that Mercati transcribes the final word as καρδίαν, which Carrera modifies to καρδίαι before noting, “Contra Mercati, this is a iota adscript, not a nu.” (266)³² There is, therefore, no discrepancy at all for the case ending of καρδία. Accordingly, we also do not need to dismiss – as the «Osservazioni» (175 n. 25) recommend – the attribution to Aquila (= Theodotion), καρδίᾳ, in Ps 94(93):15.

35(34):13a Aq

Mercati and Carrera alike transcribe Aquila’s translation of בַּחֲלוֹתָם as ἐν ἀρρωστίαις αὐτ(ῶν). The scribe of Ra 1098 seems to have been distracted both with the lines of Symmachus – on which see Mercati (55 n.; also «Osservazioni», 205–06), whom Busto Saiz follows in his edition – and of Aquila. Though the page from which the scribe copied likely had ἀρρωστίαι, the scribe unintentionally slipped in a sigma from the original line of Symmachus. In this instance, Syh ignores the difference between the number of Aquila and Symmachus. As per the «Osservazioni» (206–07), we follow the singular reading.

35(34):14b Aq

Mercati transcribes κα ⁎ ⁎ ⁎ ⁎ ⁎ and proposes ‘possibly κατέκυψα.’ (55 n.) Carrera transcribes κατέ̣κ̣υψα and then comments: “κατέ̣[κ]υψα—The first three letters are visible as well as the last three. The middle kappa is completely blocked.” (283)³³ Thus, we read κατέ⟨κ⟩υψα.

35(34):15 Aq

The verbal repetition of וְנֶאֱסָפוּ נֶאֶסְפוּ and the similarities between וְלֹא יָדַעְתִּי and וְלֹא־דָמּוּ invite fluctuation. 4QPsᵃ (4Q83) indeed reveals variation with the former, lacking וְנֶאֱסָפוּ and עָלַי by reading שמחו נ̊אס̊פ̇ו̇ ◦[ולא ידעתי.³⁵ With Ra 1098, the scribe or his source seemingly passes over the former lines we would expect from וְלֹא יָדַעְתִּי קָֽרְעוּ: the Secunda skips from νηχιμ (= נֵכִים) to ουλω δαμμου (= וְלֹא־דָמּוּ). The scribe thus registers surprise via the notation γρ(άφεται) (Mercati = ❘^ϼ) that Symmachus et al. has a “plus.”³⁶ The entire verse as Mercati and Carrera transcribe it stands as follows:

Table 1: Mercati’s transcription of Ps 35(34):15 according to Ra 1098³⁷

Table 2: Carrera’s transcription of Ps 35(34):15 according to Ra 1098³⁸

We partially follow the reconstruction that the «Osservazioni» (211–14, esp. 212) suggest of reading ⟨συνελέγησαν ἐπ’ ἐμέ⟩ for Aquila. The corrupted line of Ra 1098 has sufficient space to reconstruct the line for נֶאֶסְפוּ or עָלַי (cf. Carrera, 284), and we accept restoring ⟨ἐπ’ ἐμέ⟩ (= עָלַי) for Aquila. However, it seems entirely plausible that Aquila’s Vorlage lacked נֶאֶסְפוּ (thus Kennicott MS 147,³⁹ similar to 4QPsᵃ), so we do not accept the suggested restoration of ⟨συνελέγησαν⟩.

Our emending ᾔδειν to ⟨ἔγνων⟩ follows the «Osservazioni» (214 and n. 27), based on Ps 34(33):8, which has ιαδαθ for the Secunda and εγνως for Aquila.

35(34):16a Aq

The line in Ra 1098 reads ἐν ὑποκρίσει σεσι. Mercati (55 n.) insists we should ‘certainly read’ ἐν ὑποκρίσεσι, which Barthélemy (207) follows. As the parallel lines of σʹ and εʹ have ἐν ὑποκρίσει, the «Osservazioni» (217) are quick to conclude that Aquila actually had σεσι and that the scribe mistakenly copied the singular from the others: the original reading must have been plural because Aquila was pedantic and a Christian scribe could not have known that the Heb. word being translated was plural.

Mercati’s explanation is the obvious answer yet certain facts weigh against it. First, it does not mesh well with the scribal practices of Ra 1098, attested in dozens of cases where the scribe made corrections.⁴⁰ The pervasive way in which he corrected are (1) by writing over incorrect letters; (2) by erasing incorrect letters before rewriting them; or (3) by supralinear corrections (see Carrera, 58–59). In only one instance is there a sublinear correction (see Carrera, 318).⁴¹ Were the scribe “correcting” the text by his typical modus operandi, we would expect him to expand the final iota of ὑποκρίσει to make sigma before adding another iota (thus, ὑποκρίσεσι) or to write the correction supralinearly. Writing to the right is the typical way of presenting variant versions, not corrections. Second, no a priori reason exists why a Christian scribe could not know enough Heb. to “correct” Aquila. All that would be required is someone observant enough (1) to discern inductively the correlation between transcriptions in the Secunda and number, as with the plural form βαάνφη here, and (2) to expect that Aquila “should” have translated plural Heb. nouns in the plural.⁴²

In short, something unique seems to be happening here.⁴³ The scribe of Ra 1098 undoubtedly wrote σεσι in the margin. Why is the question. That the antigraph originally had ὑποκρίσεσι in the column of Aquila is not necessarily the best answer. Possibly, the antigraph had σεσι in the margin from some prior “corrector,” although in this case we might expect the scribe of Ra 1098 to write γρ(άφεται) as elsewhere – 32(31):7 (αʹ); 35(34):15 (σʹ), 17 (σʹ) – for an unexpected phenomenon that “is written” so in the manuscript before him. However, were the scribe of Ra 1098 registering an emendation for which he, as a copyist, had no textual precedent, and he therefore felt unfree to modify the text as it stood, perhaps this is what a marginal suggestion would look like as opposed to a correction. We adopt this explanation and leave the noun in question in the singular.

35(34):20 Aq

Mercati transcribes λαλήλουσι as is without writing “sic” in a footnote – his typical procedure when indicating a scribal error. The «Osservazioni» (236–37) discuss λαλήλουσι as if the form under consideration were an unexpected λαλοῦσι. Carrera transcribes λαλήλ̣ουσι and classifies it as a scribal error that apparently combines “the present indicative and the future indicative of ‘λαλέω.’” (287) But the simplest solution is to emend the third lamed to sigma and read it as an error for a future indicative.⁴⁵

Mercati and Carrera both depict a supralinear alpha above the final letter of ἀθρόα, yet they differ respectively in transcribing the ending as ἀθρό⁎ vs. ἀθρρ̣α̣. While Mercati’s transcription reflects a more expected form, Carrera (287–88) provides a photograph substantiating his choice. Whatever the correct reading of Ra 1098 is, the original reading of Aquila is likely a substantive form of the adjective ἄθροος ‘noiseless’ (the text above represents the corrected accent from ἀθρόα to ἄθροα). The Syh understood it so, unexpectedly taking ἄθροα as modifying ῥήματα while leaving γῆς and ἐπιθέσεων as genitives of a single head phrase; thus, ܡ̈ܠܐ ܕܡܢ ܫܠܝ ܕܐܪܥܐ ܕܨܕܘܐ ‘words (originating) from silence in the land of mockery’ (cf. «Osservazioni», 237–38).

35(34):23a Aq

Mercati transcribes ε̣ἰ̣ς κρ̣ί⁎⁎ and notes that he initially “wrote κρισιν, but it is unclear whether the word ends in σιν or μα.” (57 n.) He attempts to downplay his original transcription by saying the parallel columns influenced him («Osservazioni», 246–47, esp. n. 17). The real reason for his doubt, however, is that Aquila unequivocally translated משׁפט → κρίμα elsewhere, as in 89(88):31, and Mercati does not think it fitting that Aquila would translate a word of such frequency and importance with alternate Grk. terms. Carrera reads ε̣ἰ̣ς κρί̣σιν, without commenting on the variance from Mercati. In this instance, his lack of engagement with translation technique serves us by providing a bias-free transcription. Accordingly, we side with Carrera and Mercati’s initial transcription.

35(34):25b Sym

Busto Saiz omits Εὖγε, following the suggestion of Mercati (59 n.). But the «Osservazioni» (256) can offer no real basis for the decision, besides Εὖγε being a plus – hardly an argument worth buying. We therefore leave the text as it is.

35(34):27a Aq

Mercati gives the (corrected) text as αἰνέσαιεν but elaborates in the «Osservazioni» (261–62) that originally it read αινεσανσαν. In contrast, Carrera gives the corrected reading as αἰνεσαίσαν, the original being αινεσανσεν, and backs up his dissenting assessment with a photograph (292–93).⁴⁶

46(45):2a Aq

The scribe of Ra 1098 wrote the first line of 46(45):2a as follows:

Table 3: Mercati’s transcription of the first line of Ps 46(45):2a according to Ra 1098

We can only assume accidentally, Carrera without comment transcribes ἡμῶν instead of ἡμῖν for Symmachus. As for the scribe of Ra 1098, he wrote λανου in the second column instead of the first, an unprecedented error in the manuscript (Carrera, 318). The «Osservazioni» (320–21) recommend reconstructing Aquila anarthrously as ⟨θς̄ ἡμῖν⟩.⁴⁷ However, Aquila did occasionally add the article when not present in אלהים (see 18[17]:29; 46[45]:6bis, 8, 11), so one cannot assume he would not have done so. One can better account for the parablepsis of the scribe if Aquila and Symmachus identically had ὁ θς̄ ἡμῖν. Accordingly, we may reconstruct the first line as follows:

Table 4: A reconstruction of the original first line of the Hexapla of Ps 46(45):2a

46(45):8b Aq

Mercati transcribes θς̄ Ἰακωβ for Aquila and notes, “The apex clearly seen before θς̄ belongs to Ιακωβ b.” (77 n.; i.e., the preceding column of the Secunda, which reads ἰα̣κὼβ). Carrera, in contrast, transcribes ὁ θς̄ ἰακωβ for Aquila, although he does not explain the change in his commentary. I suspect Mercati tried to justify what he wanted to see: Aquila never adding so much as an article that lacked a correlating morpheme in the Hebrew (cf. 46[45]:2a above).

46(45):9a Aq

The original κατεργασματα was likely shortened to κατεργασμα in transmission («Osservazioni», 341),⁴⁸ although Aquila elsewhere translated a plural Heb. noun in the singular (e.g., 46[45]:5; 49[48]:4). Perhaps we should even explicate it as κατεργάσμα(τα), though Mercati and Carrera identically transcribe Ra 1098 as singular.

46(45):12b Aq

Mercati, whom Carrera follows, supplies the implicit ending of Aquila here as ὑπερεπαρτὴς ἡμ(ῖν), even though Aquila in 46(45):8 explicitly has ὑπερέπαρσ(ις) ἡμῶν. We should allow the repetitive nature of the refrain to guide us in explicating the ending of ἡμ- in v. 12b like the unequivocal genitive of v. 8b. Thus, read ἡμῶν in both cases. And yes, Aquila translated משׂגב differently in both verses!

49(48):4b Aq

Rahlfs 1098 plus Coislin. betray influence from the columns of οἱ οʹ and εʹ (= σύνεσεις) by degrees: 1098 only reveals the influence of their plural nominative ending, while Coislin. entirely replaces Aquila with their lemma (cf. «Osservazioni», 379–80). Aquila’s translation presupposes an elliptical construction in the second stichos with יְדַבֵּר → λαλήσει governing both clauses; see similarly 35(34):17; 89(88):26, etc.

49(48):8a Aq

Mercati and Carrera’s transcriptions differ respectively in one letter of one word: ἀδελφος̣ vs. ἀδελφοι̣. The latter comments, “ἀδελφοι—Scribal error. The scribe wrote either ‘ἀδελφοι’ or ‘ἀδελφον.’ Additionally[,] there is no visible accent.” (340–41)⁵⁰ It would seem Mercati allowed himself to be led by the other columns (all = ἀδελφος) in his transcription. If Aquila indeed translated ἀδελφον, he would be interpreting the preverbal אָ֗ח – note the revia – as a topicalised object, like Barthélemy (288–89) is inclined to interpret the Heb.⁵¹ If, however, Aquila translated ἀδελφοί, then he would have modified the number of the noun to match the plural addressees of 49(48):2–3,⁵² entailing the vocative case, though the plurality of the noun masks it. Repeatedly and unambiguously, Aquila used the vocative for singular addressees elsewhere (30[29]:13bis; 31[30]:6; 35(34):17, etc.). We follow this latter possibility and transcribe ἀδελφοί.

49(48):12a Sym

Busto Saiz’ apparatus records “οἰκῶν] οἰκιῶν F,” from whence comes the reading of Chrysostom. It seems Busto Saiz has ignored Mercati’s footnote (89 n.) indicating οἰκείων (= Carrera) as the actual reading. Rahlfs 1098’s οἰκείων reflects a change from Chrysostom’s οἰκιῶν due to iotacism,⁵³ so the two are in fact agreed. Rather than emending οἰκείων (i.e., οἰκία) to the synonymous οἰκῶν (i.e., οἶκος), we would do better to stick with οἰκία.

49(48):15c Aq

The «Osservazioni» (406) unnecessarily recommend following the original αὐτοῦ of 1178, which was rightly corrected to the plural (see Barthélemy, 301–03).

49(48):15c Sym

Busto Saiz excludes the full reading of Symmachus from manuscripts beyond Ra 1098 (on which see «Osservazioni», 404–06; Barthélemy, 302), but we incorporate it. One could assume Syh’s ܕܝܠܗܘܢ 2ᵒ entails a variant αὐτῶν, but more likely the Syriac translator harmonised the number to the preceding pronoun.

89(88):33a Aq

Carrera transcribes ἀθεσίας and comments (346) that the reading is clear. In contrast, Mercati transcribes ἀθεσίαν̣ and notes (91 n.) that the reading of the nu is uncertain. The «Osservazioni» (411–12) clarify the convoluted attribution of Ra 264 (omitted by Field), which supports the plural reading of Aquila. Hence, Mercati’s expectation of how Aquila would have translated has once again mislead his transcription.

89(88):45a Aq

The text of Aquila simply reads κεκαθαρισμ- without the ending specified. Carrera explicates it as κεκαθαρισμ(όν) and Mercati as κεκαθαρισμόν⁵⁵ but the latter remarks in a footnote, “Sic. l. κεκαθαρισμ⟨έν⟩ον.” (95 n.) It seems Mercati explicated the ending and forgot he had done so. Regardless, giving the text as κεκαθαρισμός implies an unlikely neologism with the partial reduplication, leaving us with a text that needs “correction” if we follow Mercati.⁵⁶ We should simply represent Aquila as the perfect passive participle κεκαθαρισμένον.

89(88):46a Aq

Mercati transcribes Ra 1098 as νεανι̣ο̣τήτ(ων) αὐτ̣οῦ, Carrera differing only in which letters are dubious: νεαν̣ι̣ο̣τήτ(ων) αὐτοῦ. The ending of the lemma in Ra 1175 is also implicit, which the «Osservazioni» (431) plausibly suggest was changed from νεανιοτήτ(ων) to νεοτήτ(ων) due to influence from Symmachus – in Ra 1098, νεό̣τητ̣ο(ς). The question then becomes whether to explicate the lexeme νεανιότης as plural (so Mercati, Carrera, and Field) or singular in both instances.

While the «Osservazioni» (431) correctly observe that Aquila translated עלמות in Pss 9:1 and 46(45):1 with νεανιότης, they fail to acknowledge that in the former instance the Grk. noun is singular but plural only in the latter: עַלְמוּת → νεανιότητος vs. עַֽל־עֲלָמוֹת → ἐπὶ νεανιοτήτων, respectively.⁵⁷ By comparing Aquila’s translation of the nearly identical Heb. expression in Job 33:25b, יָשׁוּב לִימֵי עֲלוּמָיו → ἐπιστρέψει εἰς ἡμέραν νεανιότητος αὐτοῦ,⁵⁸ we see a truth evident elsewhere also: Aquila translated plurale tantum nouns as singular or plural contextually. In 89(88):46a, we should explicate Aquila as singular both because of the similarity of the expression in Job 33:25b and because the reading Ra 1175 represents would be easier to explain if the endings of Aquila and Symmachus were both singular.

Conclusion

From the nearly 30 verses presented above, mostly of Aquila but also of Symmachus, several characteristics deserve mention. In many instances (18[17]:34b; 30[29]:6d; 35[34]:16a; 46(45):2a; 49[48]:8a, 15c; 89[88]:33a, 46a), Mercati’s conviction that Aquila should or would have always translated pedantically has hampered his text-critical judgments. The same is true of both Mercati and Busto Saiz in the case of Symmachus for 35(34):25b. On a couple occasions (18[17]:36b, σʹ; 35(34):15, αʹ), the preferable reading reflects a Vorlage different from the Leningrad Codex, which prior editors have been reticent to acknowledge. Finally, Carrera’s recent work repeatedly improves upon Mercati (30[29]:6d; 32[31]:9b–c, 11b; 35[34]:14b, 23a, 27a; 49[48]:8a; 89[88]:33a). The reason I have not emended the text of Ra 2005, preserving portions of Aquila and Symmachus for Ps 22(21):15–18, 20–28, is simple: in the few verses that remain, I would not emend Taylor’s work on Aquila nor Busto Saiz’ on Symmachus.

Whether my efforts constitute a modest contribution toward the ongoing efforts of establishing a critical text of Aquila and Symmachus, I leave to others to decide. At the very least, I have attempted to justify the reasons for my editorial judgments in a way that invites critique. Living in the twilight of the efforts of the Göttinger Septuaginta and Hexapla Project as we do, I hope the present work will anticipate the dawn.