The Samareitikon , the “κατὰ Σαμαρειτῶν” Marginalia in Codex M, and P. Carl 49

This paper explores the relationship between the Samaritan Greek translation of the Pentateuch, i.e., the Samareitikon , and an obscure 5th cent. fragmentary papyrus of Exodus, Carl 49. The latter has been recognized previously as transmitting a text of the Septuagint which was obviously revised towards some kind of Semitic source. It is argued here that the Semitic base upon which Carl 49 was revised was not Jewish but Samaritan. This is based on a textual analysis of the fragment which reveals important connections with the Samaritan textual tradition, specifically the Samaritan Targum. Further, this analysis may possibly be confirmed by external evidence, namely an obscure marginal reading designated κατὰ Σαμαρειτῶν found in codex M, the heavily annotated 7th cent. Octateuch ms.


The Samareitikon: A Brief Historical Introduction
Western scholars first encountered readings purported to have come from a Samaritan Greek version of the Pentateuch with the publication of the Sixtine lxx in 1587. Therein, Pierre Morin, the edition's collator, supplied at the end of each chapter selected variant readings culled from catena mss from the Vatican Library. Some of these readings were designated by the curious moniker τὸ σαμαρειτικόν, or "the Samaritan". This same collection of readings was Marsh Vetus Testamentum 70 (2020) 285-297 reprinted one year later in the Latin-Sixtine and was also incorporated into Johannes Drusius' commentary on the hexaplaric fragments (1622). Editions and studies of lxx including these readings made no real significant comments on the origins or nature of this τὸ σαμαρειτικόν, that is until the same were reprinted yet again in the London Polyglot (1653-57) edited by Brian Walton. In his lengthy introduction, Walton noted (Proleg. XI, §15 and 22) that these σαμαρειτικόν readings seemed to have a special relationship with the Samaritans' Aramaic Targum (ST). Edmund Castell, who collected these in the Polyglot's sixth volume (that providing textual variants), put forth a stricter conclusion: the fragments labeled τὸ σαμαρειτικόν always derived from ST.1 This Walton-Castell hypothesis, namely, that there is a special, largely direct correlation between the Samareitikon and ST, was embraced by Frederick Field in his greatly influential 1875 edition of hexaplaric remains.2 Later, Abraham Geiger in 1876 and Samuel Kohn in 1894 also accepted this hypothesis, though the former believed the σαμ-readings were merely marginal Greek glosses based on ST, while the latter argued them to be the remnants of a complete translation the Samaritans had made of their Aramaic version.3

The Publication of Gießen Fragments
Then in 1911, Alfred Rahlfs and Paul Glaue published two mss from Egypt transmitting fragments of Gen and Deut dated to the 5th-6th century ce.4 In their study, Rahlfs and Glaue argued the fragments of Deut contained remnants of the Samareitikon.5 They based this theory primarily on the reading

Marsh
Vetus Testamentum 70 (2020) 285-297 not be Samaritan because the Samaritans used the Qere šēmå or pronounced ‫יהוה‬ as Ιαβε.17 He also pointed out several readings which disagreed with ST. In the end, though Tov acknowledged it is possible Gie could be a Samaritan revision of lxx, he believes this scenario improbable, instead preferring to view its text simply as a revision, of unknown provenance, akin to the recentiores. 18 Despite the influence of Tov's widely cited article, not all have followed his lead. Key to understanding Tov's analysis is his full and unqualified acceptance of the Walton-Castell hypothesis, specifically Kohn's version. For example, noted Samaritanologist Reinhard Pummer, whose earlier 1987 article argued against insisting every instance of Αργαριζιμ must derive from a Samaritan source,19 later published in 1998 the most detailed analysis of the σαμ-readings to date. Therein, he generally supports Tov's textual construal of Gie but differs on the final analysis. Pummer's investigation, aided by the publication of a reliable critical edition of ST edited by Abraham Tal in the 1980s, leads him to conclude that only those readings agreeing with ms J, the earliest strand of ST, are "potentially significant".20 As a result, the σαμ-group collectively fails to show adequate exclusive correspondence to the earliest strand of ST. Thus, Pummer concludes that "it can no longer be taken for granted that the Samareitikon passages have a close affinity to [ST]".21 Consequently, since Pummer rejects the Walton-Castell hypothesis, he does not preclude the possibility that Gie could be Samaritan.22 More recently, Adrian Schenker contested Tov's negative assessment of Gie's alleged Samaritan provenance vis-à-vis lxx.23 To him, scholars should expect the Samareitikon to have been a specifically Samaritan revision of lxx, parallel to the Jewish recensional activity evinced by the other recentiores.24 For Schenker, Αργαριζιμ represented the key instance where the reviser of 17  Most recently, Jan Joosten has argued against both Tov's categorization of Gie as well as Pummer's denying significance to σαμ-readings agreeing with ST mss other than J. Instead, Joosten accepts a modified version of the Walton-Castell hypothesis contending that the Samaritans revised lxx in accordance with their own evolving Hebrew text and exegetical traditions (i.e., ST and their reading tradition).27 Indeed, he asserts that agreements between Gie and lxx probably reflect an era when the Samaritan community regarded lxx as common property between themselves and "Jerusalem-based Judaism".28 As for their Targum, Joosten believes that "Samaritan traditions, like Jewish ones, are multiple, and sometimes contradictory".29 Thus, σαμ-readings agreeing with ST mss other than J (e.g., Gen 50:19, Exod 13:13, Lev 25:5) are not a problem; further, even those disagreeing with ST possibly represent exegesis the ms tradition no longer preserves (e.g., Exod 3:22; 16:31). Thus, Joosten holds ST was not the Samareitikon's only Vorlage but merely "one of the inputs which went into the production of the Samareitikon".30 Whether one adheres to either Tov's, Pummer's, or Joosten's construal of the Walton-Castell hypothesis, it can be held with confidence that the Samareitikon likely had a meaningful relationship to ST. Determining the nature of this relationship more precisely, however, requires additional evidence.

4
The it would not be unusual if both Aquila and "κατὰ σαμαρειτῶν" adopted the same Greek. Verse 15's occurrence of this same designation is unfortunately not indexed to anything in the running text, nor is there any attending variant reading. The part of v. 15 immediately adjacent to the note (Οὕτως-Ἰακώβ) reads in precise agreement with lxx, and the only variation between MT and SP for that portion is the latter's waw before ‫יצחק‬ ‫.אלהים‬ Kohn attempted to connect the notation to the anonymous ἐπικέκληται recorded in another ms at v. 18. However, his reason for this is unclear; possibly, he misread Field.39 The most challenging aspect of these marginalia is the unique form of the attribution. Strictly the annotation could be read to mean "against the Samaritans" or "according [to the (acc.)] of the Samaritans". Certainly, the situation in 3:6 ostensibly admits the latter possibility. Kohn believed the laconic annotation was shorthand for κατὰ ἑρμηνείαν Σαμαρειτῶν.40 With respect to the designations κατὰ Σαμαρειτῶν and σαμαρειτικόν, however, it is important to remain open to possibilities given the very early age of the ms. With respect to "Samaritan" labels, M itself only records these here and a paltry three other instances as follows: Conspicuously, the fully spelled-out σαμαρειτικόν of later mss is unrepresented. It is thus logical to presume that, to the scribe(s) of M, το Ϲαμ′, το Ϲαμαρ′, and κατα Ϲαμαρειτων were one and the same; this is particularly possible since the more ambiguous designation was provided first and only later made clearer. One would rationally expect that the designation had not yet been standardized in Christian scribal circles.
But are there other possibilities? One might consult other such annotations in the witness before confirming this analysis. In fact, there are two such similar remarks also transmitted in M. The first is at Gen 19:24, where the scribe wrote ση. (= σημείωσαι) κατὰ Ἰουδαίων.41 This notice served not as shorthand indicating a variant reading, but as an indicator flagging up a proof-text which was used to contradict the Judaic rejection of the Trinity.42 The second such occurrence, and more directly related to the present discussion, is the anonymously recorded scholion at Num 19:13 (fol. 111b). This note, which is identified by other sources as Apollinaris of Laodicea's (d. ca. 390 ce), states that this passage argues "very strongly" against the Samaritan denial of the resurrection.43 The Samaritan denial of the resurrection was of course a common subject exploited by patristic writers, often in heresiological or exegetical works.44 For instance, Origen discussed this in his Comm. in Matt. (17.29) where he expands upon Matt 22:32, observing that both the Sadducees and Samaritans rejected the resurrection. Matthew 22:32 quotes Exod 3:6 in providing a proof-text for resurrection, having Jesus add to this "He is God not of the dead, but of the living" (nrsv). This surely is the key to understanding the κατὰ Σαμαρειτῶν note at v. 15. This is why that annotation has no marginal reading or index marker. It was placed next to the lines containing "Thus you shall say to the sons of Israel, the Lord, the God of Abraham, and the God of Isaac, and the God of Jacob", a repetition of that from 3:6, which was parroted as a proof-text (via Matt) for the resurrection against the Samaritans by Origen and others.45 As for the note at Exod 3:6, the case is less clear. Was it meant to refer to the marginal ἀπέκρυψεν or to the running text a few lines above it: "I am the God of Abraham, the God of Isaac, and the God of Jacob" (v. 6a)? If the former, then ἀπέκρυψεν may very well be one of the earliest σαμαρειτικόν readings. If the latter, then it ought to be interpreted just as the annotation at v. 15.46 But if the scribe meant to indicate this, why was κατὰ Σαμαρειτῶν not provided before (or rather above, in the ms) the Aquila reading? It should further be pointed out that no patristic evidence (that I know of) indicates that the reading ἀπέκρυψεν or ἀπέστρεψεν was ever used polemically against the Samaritans. The lemma ἀπέστρεψεν, however, was used polemically against the Jews, namely their having "turned away" from the divinity of Christ in much the same way Moses turned his face from the bush.47 But if this kind of inference was what the scribe meant, why then did the note not also read (καί) κατὰ Ἰουδαίων, including both Jews and Samaritans? Why did he only write κατὰ Σαμαρειτῶν? As the evidence stands, Kohn's interpretation could perhaps be cautiously upheld-but only for 3:6. In light of such ambiguity, can any further evidence shed light on this obscure marginal note?  [MT]".52 As an example of this, they mentioned but one place where Carl 49 agrees with SP (των π̅ ρ̅ ω̅ ν̅ σου).
No further examples of agreement with SP were offered, and it is unclear if the Samaritan textual tradition was taken seriously as a possible base. For instance, the reading γεργεσαιου was compared with the "Syriac text" as opposed to SP.53 Equally interesting are those qualitative readings which differ from lxx provided in Table 2. Concerning Greek renderings, Bülow-Jacobsen and Strange noted certain oddities, especially Carl 49's μη ενγισης [συ]ναρπαγη̣ ι ̣.54 This reading, which they translated "Do not approach hastily", prompted unease, especially as they found that "[Carl 49's] translator is usually very faithful to the Hebrew".55 As a result they struggled mightily to explain it. They attempted to connect the reading to the root ‫,הלם‬ which "like συναρπαγη, has a connotation 52 "P. Carlsberg 49", p. 16. 53 "P. Carlsberg 49", p. 21. This is confusing since the Peshitta provides all the nations in their plural forms without mention of the Gergashites. Presumably then they meant the Syrohexapla which does not really follow. 54 Bülow-Jacobsen and Strange interpret συναρπαγη as a single lexeme (thus LSJsuppl, s.v.), not as two separate words (so Wevers' AppI). 55 "P. Carlsberg 49", p. 20.