Toward a Reading of Proverbs 30:1b: Tracing the Life of the Text in the Versions

Proverbs 30:1b presents one of the most intractable text-critical dilemmas in the HB. Following Ronald Troxel’s suggestion that text criticism be reimagined as “a commentary on the life of the text,” I suggest the way forward in reading Prov 30:1b lies in care-fully engaging with the versions as a window on its history. Emerging from this process, I argue that Prov 30:1b may have once read * ל ָכוּא אֹל ְו י ִתי ִא ָל , “I am weary and pow-erless.” Early on, however, this text was conflated with another textual tradition that read a proper name thus producing a double reading. In time, scribes harmonized this double reading which then calcified in MT. The versions and analogous biblical passages suggest the proposed text, while documented scribal practice and lexical usage support it.


Introduction: The Obscurity of Prov 30:1b
Proverbs 30:1 contains one of the most inscrutable text-critical dilemmas in the HB: The words of Agur son of Jakeh, [man of] Massa; The speech of the man to Ithiel, to Ithiel and Ucal: (JPSV) The simplest approach to the second half of the verse, illustrated by the JPSV, is to read three proper names identifying the addressees of this collection.Whether or not one adopts this reading, nearly all scholars admit the text is corrupt.1Both the two named addressees and the awkward repetition of the first name are unusual since Proverbs, along with instructional material generally, does not identify addressees by name but rather as son(s) of the speaker/ author.2Moreover, although the name Ithiel is attested once (Neh 11:7), the name Ukal is unheard of.Finally, v. 2 opens with a ‫כי‬ clause that compounds the disorienting effect of these indecipherable characters.Most often ‫כי‬ subordinates a clause to what precedes it, suggesting to interpreters the content of the composition began in v. 1.3 Therefore, this string of proper names leaves scholars with the distinct impression that the beginning of Agur's discourse is missing.Darkening the situation, a dearth of Hebrew manuscripts confounds text criticism of Proverbs generally and of Prov 30 in particular.4If one turns to the ancient versions for help, their wildly divergent readings cast disorienting shadows.It is not obvious that the Septuagint (G) and the Vulgate (V) are reading a Hebrew text that bears any resemblance to the Masoretic Text (MT).William McKane despairs: "Where there is hardly a glimmer of light, one feels powerless to make even the first move towards its elucidation."5The text critic who hopes to untangle this knotted palindrome must do so blindfolded.Faced with such impediments, some scholars think it best simply to read with MT.6 Other scholars-emboldened by the desperate textual straits-propose conjectural emendations that rest on little evidence and shaky theological reconstructions.7The majority of commentators, however, take a measured approach in redividing and repointing MT so a meaningful start to Agur's discourse emerges.8 I propose that we can illuminate this obscure verse by deepening our appreciation of the task at hand.In a recent Vetus Testamentum article, Ronald Troxel renewed the call for the collapse of higher and lower criticism.9The task of textual criticism itself might be reconceived, not as a quest for the elusive archetype, but as "a commentary on the life of the text."10This approach treats text criticism as an interpretive task-part of the exegetical processthat persuades to the extent it weaves a narrative explaining the evidence and yields meaningful readings.11Similarly, Hindy Najman urges scholars to think in terms of "traditionary processes that encompass both textual formation and textual interpretation."12Her approach would transform how we see variations within the textual tradition.Once thought of as errors and corruptions, we might now view them "as signs of life-as manifestations of the vitality of the traditions that we are studying."13Returning to Prov 30:1b, because of the nature of the problems and the available evidence, there is little hope for a solution that will win a broad consensus.Yet contained within the versions is a 6 For example, Markus Saur reads proper names and looks for symbolic meanings; "Prophetie," 575.7 Nearly seventy years ago, Charles C. Torrey hypothesized that in order for our text to become as corrupted as it appears to be, a scribe must have intentionally moved to blot out a statement he found theologically reprehensible by translating it into Aramaic ("Proverbs," 94).Torrey's core proposal has been adopted and developed in Scott, Proverbs; Murphy, Proverbs; and now Sandoval, "Texts and Intertexts."While these approaches have produced an ingenious solution to the problem, they have done so without recourse to the textual evidence we actually possess.For example, Sandoval interacts with G and V in passing by citing their readings in footnotes, largely to claim, "the textual traditions had difficulty understanding" their Hebrew Vorlage (" narrative of the struggle to understand this verse.This narrative suggests a difficult text, misunderstanding, and competing interpretive strategies and agendas.Though scholarly proposals for emending the line are legion, still lacking is an attempt to trace the life of the text through the versions, weaving a narrative that holds together all the available evidence in a compelling synthesis.14If a path forward exists, it lies backwards through the versions.
In this paper, I examine the evidence of the versions then survey some recent text-critical approaches to Prov 30:1b.Emerging from this process, I argue that Prov 30:1b may have once read ‫ל*‬ ‫אּוכָ‬ ‫ֹא‬ ‫ל‬ ‫וְ‬ ‫י‬ ִ ‫ית‬ ‫אִ‬ ‫,לָ‬ "I am weary and powerless."This challenging text appears to have been understood to contain a name by at least one reading tradition.This interpretation was then preserved within a MS as a double reading.Eventually, scribes harmonized this double reading which then calcified in MT.I will support this proposal by appealing to documented scribal practice concerning the preservation of doublets, the cognitive process of reading, and the creation of "ghost names" in MSS.By offering a solution as a commentary on the life of the text, not only do I hope to offer an illuminating reading of Prov 30:1b but also to model and commend an approach to working with texts that could shed light on other dark passages.

Narrating the Versions
The readings of the versions do not allow us to establish an original text with any confidence but rather to explore the backstories of the text, i.The second approach, represented by G and V, interprets all of the words understood to be proper names in other witnesses as verbal phrases.I will look more closely at the Greek versions, followed by S and V in turn, before evaluating some modern scholarly approaches to reading MT.Nearly all the major interpretive options are already available in the Greek versions.G's τοῖς πιστεύουσιν θεῷ could represent a periphrastic rendering of ‫71.לאיתיאל‬ Michael Fox suggests that G read ‫י+ל‬ ‫ל+איתַ‬ ("to whom there is God"), with ‫י‬ ‫איתַ‬ construed as the Aramaic particle of existence ("there is," 16 The Masora Parva at ‫ואכל‬ has the annotation ‫וחס‬ ‫,ל‬ indicating that the masoretes believed this form only occurs here written defectively.This notation suggests they associate ‫ואכל‬ with the plene form ‫ל‬ ‫אּוכָ‬ ‫וְ‬ (1cs qal yiqtol ‫)יכל √‬ that occurs in Jer 20:9; Ps 101:5; and Job 31:23 and is always qualified by ‫.לא‬In other words, the masoretes appear to be reading with θ´.Kennicott records 64 out of 200 MSS with the reading ‫ואוכל‬ and an additional three originally contained that reading (Vetus Testamentum, 474).While this evidence cannot be used to suggest anything about the original text, it does suggest there is a strong minority tradition within MT that read ‫ואכל‬ as a verb.17 At first glance, G appears to have no connection to MT.However, it is possible to find correspondences between the two for all the major elements based on etymological exegesis; e.g., HBCE, 379; cf.BHQ.This is an extreme instance of the overall translation character of G Proverbs, which Fox has described as "flexible" for the sake of "control" ("A Profile," 16-17); and Forti has described as "free and even periphrastic" ("Septuagint," 254).It remains a matter of debate, however, to what extent G is working from a different Vorlage (Tov, "Recensional Differences"; Cuppi, "Long Doublets," 92) or carrying out a quasi-authorial agenda (Cook, Septuagint; Waltke, Proverbs, 454 n. 1).In either case, G represents a different recension of the book of Proverbs.G Prov 30:1 does not give a clear indication of what the translator understood ‫לאיתיאל‬ to mean.On the relationship of G Proverbs to MT consider the works above; as well as Clifford, "Observations"; and Aitken and Cuppi, "Proverbs."e.g., Ezra 4:16; 5:17; cf.BH ‫81.)יׁש‬This could plausibly amount to etymological exegesis of the post-exilic name ‫,איתיאל‬ which appears in Neh 11:7, perhaps meaning "God is with me" ‫אל(‬ ‫איתי‬ for ‫אל‬ ‫91.)אתי‬The Greek plural is universalizing: "to the one whom God is with"="to anyone whom God is with."The second element in G, καὶ παύομαι, suggests G read a 1cs qal wayyiqtol/jussive from ‫כלה√‬ (to be complete, finished) for ‫02.ואכל‬The Hexaplaric recensions, α´ and θ´, break from G by reflecting ‫לאיתיאל‬ as a proper name + ‫.ל‬ Aquila's reading, καὶ τέλεσον ("now [you] stop"), agrees with G in finding ‫כלה√‬ in ‫,ואכל‬ although it transposes the form from the first to the second person.21On the other hand, θ´'s καὶ δυνήσομαι interprets ‫ואכל‬ as deriving from ‫יכל √‬ ("to be able, prevail").Theodotian thus supports the vocalization represented in ML and even more so the Masora Parva and the MSS that read ‫22.ואוכל‬Interestingly, both α´ and θ´ use the future tense.23This strongly suggests the translators read yiqtols in their Vorlage(n), but it does not require a different consonantal text than MT.
Like the Greek versions, S registers just one occurrence of ‫,לאיתיאל‬ which it treats as a proper name.Moreover, like θ´, S treats ‫ואכל‬ as a verb from ‫.יכל √‬ However, S differs from θ´, and from G, by apparently conflating the sense of ‫ואכל‬ with ‫גבר‬ earlier in the verse and altering the word order to produce ‫ܘܐܬܡܨܝ‬ ‫ܚܝܠܐ‬ ("and was able with strength"="prevailed").24We can deduce this because S has no other element synonymous with man that might correspond to ‫,גבר‬ of our verse in Midrash Tanhuma, Vaera 5: "The man saith unto Ithiel is written because he [Solomon] would say: 'God is with me (iti-el), and I will be able (ukhal) to withstand temptation.' […] The names Ithiel and Ucal were written in that verse because he said: 'I will multiply wives, but I still will not turn my heart away (Ithiel) from God; and I will multiply the number of my horses; but I will not cause the people to return.'"27This midrashic approach to the names in the verse is reflected in Jerome's translation; he may even be our earliest witness to it.28The etymological exegesis Jerome engages in to translate the names as phrases is not far from the impulse reflected in G's reading.However, G registers only one occurrence of ‫לאיתיאל‬ while V exhibits two equivalent elements: 1. ‫לאיתיאל‬ = cum quo est Deus, 2. ‫לאיתיאל‬ = et qui Deo secum morante.
V appears to witness clearly, albeit at low resolution, to MT.

Evaluating Contemporary Readings
Having thus oriented ourselves to the versions, we can now consider stories modern scholars have told about them.Jan de Waard thinks the longer text represented in MT, T, and V is the earlier one, based on his attribution of haplography to G and S.29 However, I find this unlikely.Why would the older witnesses-G, α´, θ´, and S-all exhibit an element corresponding to only one occurrence of ‫?לאיתיאל‬ Surely the slavish α´ and θ´, which render ‫לאיתיאל‬ as a proper name, would have repeated that name again had it appeared twice in their respective Vorlagen.30This observation is strengthened by the fact that S, α´ and θ´ all read ‫לאיתיאל‬ as a proper name, whereas G reads it as a verbal phrase-the only thing they all agree on is that ‫לאיתיאל‬ appears once.Dittography in the tradition behind MT seems more likely than haplography behind G and S.31 However, a more complex process is at work than a simple scribal slip.Lorenzo Cuppi and Fox redivide and repoint the consonants of MT to arrive at the same suggestion: ‫ל‬ ‫כָ‬ ‫אֵ‬ ‫וָ‬ ‫ל‬ ‫אֵ‬ ‫י‬ ‫יתִ‬ ‫אִ‬ ‫לָ‬ ‫ל‬ ‫אֵ‬ ‫י‬ ‫יתִ‬ ‫אִ‬ ‫,לָ‬ "I became tired, God, I became tired, God, and I may fade away."32Dominique Barthélemy has a slightly different take: ‫ל‬ ‫כָ‬ ‫אֻ‬ ‫וְ‬ ‫ל‬ ‫אֵ‬ ‫י‬ ‫יתִ‬ ‫אִ‬ ‫לָ‬ ‫ל‬ ‫אֵ‬ ‫י‬ ‫יתִ‬ ‫אִ‬ ‫,לָ‬ "Je me suis fatigué, ô Dieu, je me suis fatigué, ô Dieu, pour y réussir."33All three scholars agree on redividing ‫לאיתיאל‬ and identifying it as 1cs qal qatal from ‫לאה√‬ (to be weary, impatient) with ‫אל‬ as a vocative.The vocative of ‫אל‬ tends to be signaled by the article ‫,ה‬ but it is possible to find exceptions (Num 12:13; Ps 83:2).More significantly, however, none of the versions or ancient MSS suggest that this vocative was ever part of the reading tradition.It appears to be the invention of modern scholars.
Where scholars differ is their handling of ‫.ואכל‬Following the vocalization implied by G and α´, Cuppi and Fox repoint ‫ואכל‬ as a 1cs qal wayyiqtol/jussive from ‫.כלה√‬This form does not occur in the HB but is plausible (cf.Exod 39:32; Job 33:21).Fox argues ‫כלה√‬ fits the context well since it "reverberates in the clause 'before I die' in 30:7b."34Drawing connections to Ps 73:22 ‫)כלה(‬ and 2 Sam 23:1 ‫,)נאם(‬ Fox construes Prov 30 as Agur's last words.Cuppi speculates ‫לאיתיאל‬ was understood as a proper name first and this interpretation eventually contributed to ‫ואכל‬ being taken the same way.35But why this text was misread in the first place remains a question.Cuppi tentatively suggests that without vowels an Aramaic speaker might have scanned the consonantal text of MT as ‫ל‬ ‫כָ‬ ‫אֻ‬ ‫וָ‬ ‫ל‬ ‫אֵ‬ ‫י‬ ִ

4
Weary and Powerless: A Proposal for Reading Prov 30:1b Taking all this evidence and scholarly ingenuity under advice, what can we say with any confidence?Lines of correspondence connect elements of G, V, and MT despite the meaning being changed systematically.This indicates the periphrastic interpretation in G and V is intentional and exegetical.The midrashic tradition latched on to these words.Because interpretive traditions are in play, and G's translation style is flexible overall, retroversion is challenging.Still, first and foremost, it is probable that only one instance of ‫לאיתיאל‬ is original and ‫ואכל‬ began life as a verb.This is the majority opinion of the ancient versions, including all of the oldest witnesses.I cannot see how ‫ואכל‬ could have originated either as a second addressee or hanging off the end of the title line as a lonely verb, stranded from the body of the poetic composition as in G.If ‫ואכל‬ was originally a verb then ‫לאיתיאל‬ probably represented a verbal phrase as well.We already have two early interpretations represented in Greek.G takes ‫לאיתיאל‬ as a verbal phrase ("to those believing in God") while α´ and θ´ take it as a proper name ("to Ithiel").But these traditions both interpret this phrase as indicating the addressee through the use of the preposition ‫.ל‬ Perhaps this was the inciting incident.If ‫לאיתיאל‬ originally contained a verbal phrase, then it is easy to imagine how a ‫ל‬ on a strange lexeme in the superscript of a discrete collection could have triggered a scribe to think in terms of an addressee.Perhaps eventually both the verbal phrase and the tradition representing an addressee were included in one manuscript by a scribe who did not know which was the correct reading.This scribe may have inserted a double reading that differed by as little as a word division or the transposition of two letters.Over time, MT's baffling reading calcified as the distinctions between these two interpretive options were lost.But how might this have occurred and what could the original text have been?Building on Barthélemy's approach, I will suggest a reading for Prov 30:1b that produces a plausible history of the transformation of the text grounded in both the evidence of the versions and emerging text-critical methodology.Recall that Barthélemy connects ‫ואכל‬ to ‫יכל √‬ and observes that ‫לאה‬ is collocated with ‫יכל‬ in Isa 16:12; Jer 20:9; Job 4:2.The verb ‫יכל‬ occurs just thirty-four times in the qal.Twenty-seven of these are negated.So it will be when he appears that Moab has wearied himself on the high place, and he will enter his sanctuary to pray but he will not be able.
Apart from the optional ‫,ו‬ this phrase contains identical consonants to MT, differing only in word division and the transposition of letters.39In contrast to other text-critical solutions, a compelling case can be made that such a primary text produced the readings in the versions.First, note that ‫לאה√‬ is a relatively rare verb with just nineteen occurrences and of those only three are in the qal.This form would be unique in the HB although it is quite plausible.Second, remember that ‫אכל‬ potentially derives from several different roots when written defectively.Third, consider the density of lameds (3x), alephs (3x), and yods and waws (3x) in the proposed text.These four graphemes account for nine out of eleven letters.Considering the semantic and graphic difficulty in this clause, there are many things that could go wrong in transcription.All it would take to trigger a process of transformation is one of several common scribal errors.This is the first stage in the process.By simply misdividing the words and running them together the scribe may have produced ‫.לאיתיולא*‬ Graphic confusion between ‫ו‬ and ‫י‬ would In my opinion, such errors need not represent stages in manuscript transmission.Rather, they probably occurred as simultaneous elements of the translator-scribe's natural mental process of reading.As John Screnock argues in Traductor Scriptor, scribes and translators both construct mental versions of texts based on but not identical to the physical Vorlagen in front of them.40By using resources from translation studies, particularly intralingual translation, Screnock argues that translation and transmission of MSS involve fundamentally similar cognitive processes.41He explains, "the translation process does not involve one single move from the physical Vorlage directly to the physical text of the translation; rather, there are additional intermediary stages in the translator's mind, appropriately conceived of as texts, through which this move is channeled."42Through a process of working memory and phonological loops, the translator-scribes produce physical copies of their Vorlage from their mental text.43One of the implications of Screnock's study is that many variants in the textual tradition stem from the decoding (i.e., reading) process within the translator's mind rather than errors of the eye and ear or the Vorlagen.44To recall Emanuel Tov's dictum, many reliable variants never actually existed in writing.45I propose that a reading glitch triggered a scribe to decode ‫ולא‬ ‫לאיתי‬ as a proper name.While it is possible one of the classic errors of textual criticism was at work (e.g., dittography, haplography, or metathesis as described above), it is equally likely that the scribe simply misread a difficult string of letters in a fraught context.For example, in the Antiochene recension of G, Natalio Fernández Marcos has identified a phenomenon he dubbed "ghost names."46Ghost names appear when a challenging string of letters in a Hebrew Vorlage is interpreted as a proper name and then translated as such.47This device creates a meaning in the target language that did not exist in Hebrew.48For example, in 1 Kgs 15:22 the Hebrew phrase ‫נקי‬ ‫אין‬ ("without exception") emerges in the Antiochene recension as Ἑνακείμ (cf.G: Αινακιμ).49Or, in a different kind of example, 1 Sam 14:33 has the verb ‫ם‬ ‫ּתֶ‬ ְ ‫ד‬ ‫גַ‬ ‫ּבְ‬ ("you acted treacherously") represented in the G as a place name Γεθθεμ.The Antiochene recension, however, represents it as a place name twice; once as in G and again as Ημάρτετε-still a proper name but translated "ad sensum with recourse to the verb ἁμαρτάνειν."50Fernández Marcos emphasizes the vast majority of ghost names are "attested in sequences of double readings,"51 and "occur especially in the genealogical material at the start of 1 Chronicles where the absence of a meaningful context cause [sic] major confusion of similar letters in Hebrew and throughout the Greek tradition."52He cautions scholars to attend to proper names because they often convey semantic information: "[…] they have been incorporated by the scribes to the narrative, circulated for centuries as part of the official biblical text for a community, and gave rise to new meanings, exegesis and commentaries.And in a few cases they preserve very ancient, alternative variants that may go back to a Hebrew text different from the Masoretic one."53 The phenomena analyzed by Screnock and Fernández Marcos help explain the process I believe took place in the transmission of Prov 30:1b.While transcribing the superscript of a new collection within Proverbs, a scribe sees a ‫ל‬ followed by a string of letters ‫.)לאיתי(‬In context, he processes this relatively rare set of letters as the name Ithiel (Neh 11:7).Once ‫ולא*‬ ‫לאיתי‬ was transcribed into a physical MS as a name, the traditions diverge-the earlier verbal phrase At this point-if they had not already done so-the traditions represented by the Greek and Syriac versions diverge from MT. Orthography-the optional ‫ו‬ in ‫-אכל‬could have been the factor differentiating α´ and θ´'s readings.In my opinion, G probably also witnesses to example 2, although it interpreted the name by translating its constitutive elements.54A conjunction on ‫אכל‬ may have been inserted at any stage in the process after ‫לאיתיאל‬ was interpreted as a proper name in order to make sense of the verb and smooth out the reading.Alternatively, if ‫אוכל‬ was in the Vorlage, the scribe could have engaged in exegetical metathesis of the ‫ו‬ and the ‫,א‬ whether intentionally or unintentionally, in order to make sense of his text.55 Years pass.Faced with alternative readings in competing Vorlagen, a scribe is unable to adjudicate between them and represents both side by side in a fresh MS, presenting them as complementary options for the interpreter.56This is stage three and the text may have looked like this: 3. ‫א(ו)כל*‬ ‫ולא‬ ‫לאיתי‬ ‫75 לאיתיאל‬ To Ithiel, I am weary and powerless.
Despite the lack of direct evidence for this stage, a double reading seems the most plausible explanation for how ‫לאיתיאל‬ came to be duplicated in MT First, the versions testify that there was a time before the double reading entered the MS tradition.Second, ‫לאיתיאל‬ has not always been a name but has sometimes been understood as a verbal phrase, and the midrashic approach of G and V suggests it was originally a verbal phrase different from those that have been preserved.Third, the phenomenon of incorporating alternative textual traditions into one MS as a double reading is a known practice whereby texts grew and preserved "alternative wordings of the same texts."59In the fourth and final stage of transmission, the traditions preserved in the double reading-still graphically and semantically challenging-were harmonized toward one another.This could have been accomplished through a similar set of scribal errors described above or through a simpler process of homeoarchy resulting in dittography of ‫לאיתיאל‬ triggered by the reduplication of no less than five consonants in the postulated Vorlage ‫.)לאיתי(‬ Once both readings were harmonized as this double addressee, we arrive at the text of MT and at T and V, which witness to it.The move to interpret ‫ואכל‬ as a name only comes at this stage.Strictly speaking, "Ukal" is not part of the textual tradition, since MT can still be read as a verb, especially if one adopts the minority reading of the Kennicott MSS.The same is true of T (cf.Dan 2:10; 5:16).Only V, in reflecting the midrashic tradition, demands to be read as a name.
If my proposal is compelling, Agur's ‫כי‬ clause in v. 2 now depends on a line that confesses his weakness and ineptitude in a striking manner.Both ‫לאה‬ and ‫יכל‬ have clear uses where the emphasis is on an acute mental weariness and emotional exasperation over sin (Exod 7:18; Isa 1:14; 7:13; 47:13).In Job 4:2 and 5, Eliphaz uses ‫לאה‬ twice to describe Job as edgy, impatient, easily offended.Similarly, in certain distinctive phrases ‫יכל‬ means "to endure" within the spiritual/psychological realm (Ps 101:5; Job 31:23).In Ps 139:6, ‫יכל‬ + ‫לא‬ describes the psalmist's inability to attain to the knowledge of God.Notably, when ‫לאה‬ and ‫יכל‬ are used in close quarters they focus on spiritual exasperation.In Isa 1:13-14, YHWH denounces Israel, saying he cannot endure ‫)לא-אוכל(‬ their iniquity and assemblies; and he grows weary of bearing ‫נׂשא(‬ ‫)נלאיתי‬ their feasts (cf.Jer 20:9 above).Agur's combination of these two lexemes in a context where he is confessing his intellectual limits suggests his weariness is not necessarily physical but acutely psychological and even spiritual.Agur opens on a note of profound exasperation.60distinct semantically.Similarly, Waltke deduces the purpose for which I believe the doublet was originally preserved by transliterating the first instance as "Ithiel" and translating the second instance, "I am weary, O God" ‫ל(‬ ‫אֵ‬ ‫י‬ ‫יתִ‬ ‫אִ‬ ‫)לָ‬ (Proverbs,455) The approach modeled here is what I take Troxel's collapse of higher and lower criticism to entail.Tracing the life of a text back through the versions can suggest a reading that might plausibly lie behind all of them, particularly where we lack sufficient MSS evidence to compare, count, and weigh readings.Counter-intuitively, the more cautious approach of redividing and repointing MT probably creates a text that never actually existed in the history of transmission.It is true, as Moshe Goshen-Gottstein has said, that "there is no retroversion without a residue of doubt."62However, it is equally true that every reading in the versions contains a fragment of the history of wrestling with and transforming the text.As text critics we are interpreters who concern ourselves with that history.We bring the most light to bear on the text and offer the richest interpretations when we judiciously engage all the evidence the textual traditions have preserved.6361 Troxel,"Writing Commentary," 128;cf. ibid.,106 n. 6. 62 Goshen-Gottstein, "Theory and Practice," 132.63 Goshen-Gottstein argues that we would do well to identify doublets and preserve both readings in the apparatus of our critical editions because, on principle, we cannot know which readings preserve true variants and which readings do not.Recently, Screnock ("A New Approach") has proposed updating Tov's methodology for retroverting the Septuagint in order to admit more data into the text-critical process.
‫לאיתיאל‬ and also allow for recognizing ‫ואכל‬ as a proper name, although nothing requires one to interpret it that way.16Aquila, θ´, and S register only one occurrence of ‫לאיתיאל‬ and treat ‫ואכל‬ as a verb.
Barr makes the salient point that while most variable spellings were inconsequential semantically, occasionally, especially with I-‫י‬ verbs, a vowel letter could distinguish one root from another(Variable Spellings,(93)(94).So, while ‫ואכל‬ could represent either ‫כלה√‬ or ‫,יכל √‬ ‫ואוכל‬ could only represent ‫.יכל √‬ Thus it would seem that α´ was reading a MS that had ‫,ואכל‬ while θ´ could have been reading either ‫ואכל‬ or ‫,ואוכל‬ although it seems most likely his Vorlage had the plene spelling.23ForG, Rahlfs and Hanhart, Swete, and Holmes and Parsons all record a variant with the future: παύσομαι.
Two of the three examples where ‫לאה‬ is followed by ‫יכל‬ fall into this category: ‫ה‬ ‫מָ‬ ‫ּבָ‬ ‫ל-הַ‬ ‫עַ‬ ‫ב‬ ‫מֹואָ‬ ‫ה‬ ‫אָ‬ ‫לְ‬ ‫י-נִ‬ ‫ּכִ‬ ‫ה‬ ‫אָ‬ ְ ‫ר‬ ‫י-נִ‬ ‫כִ‬ ‫ה‬ ‫יָ‬ ‫הָ‬ ‫וְ‬ ‫ל׃‬ ‫יּוכָ‬ ‫ֹא‬ ‫ל‬ ‫וְ‬ ‫ל‬ ‫ּלֵ‬ ‫ּפַ‬ ְ ‫ת‬ ‫הִ‬ ‫לְ‬ ‫ׁשֹו‬ ָ ‫ּד‬ ‫קְ‬ ‫ל-מִ‬ ‫אֶ‬ ‫א‬ ‫ּובָ‬ ‫ל‬ ‫אּוכָ‬ ‫ֹא‬ ‫ל‬ ‫וְ‬ ‫ל‬ ‫כֵ‬ ‫לְ‬ ‫ּכַ‬ ‫י‬ ִ ‫ית‬ ‫אֵ‬ ‫לְ‬ ‫נִ‬ ‫וְ‬ I am weary from containing [it] and I am no longer able.Downloaded from Brill.com05/30/2022 12:25:58PM via free access Isa 16:12 38In 1891, Gustav Bickell proposed something similar: ‫אל‬ ‫לאתו‬ / ‫ו‬ ‫אל‬ ‫]לא[לאתי‬ ‫אכל‬ (Kritische Bearbeitung, 293).He translates, "Ausspruch des Mannes, der sich um Gott abgemüht hat: ich habe mich um Gott abgemüht und es nicht vermocht."Unfortunately, Bickell's project concerns justifying a theory of prosody and he provides little explanation for his proposal, but he anticipated two important aspects of mine.He discerned that the repetitions of ‫לאיתיאל‬ originally preserved different readings and that ‫אכל‬ was likely negated.39 I represent the ‫ו‬ in ‫א(ו)כל‬ in parentheses to indicate it may or may not have been present in different MSS and stages of the transmission process.If one considers plene and defective spellings, various orthographic options emerge.My argument is not married to one spelling, but is aided by the fact that orthography was not standardized in ancient texts.See Barr, Variable Spelling, 7-11.yield ‫ולא*‬ ‫לאיתו‬ or ‫ילא*‬ ‫.לאיתי‬These readings are primed for elision, producing ‫לאיתילא*‬ or ‫.לאיתולא*‬Likewise, the back-to-back ‫א‬s could easily have been elided.Finally, metathesis of the ‫ל‬ and the ‫א‬ would produce ‫ואל*‬ ‫.לאיתי‬ Once one of these shifts occurred a second would be even more likely if only to make sense of the first.For example, the elision of an ‫א‬ could have produced ‫לאיתי*‬ ‫וכל‬ ‫.ולא‬ Searching for a meaning for ‫וכל‬ a scribe may have vocalized it as the imperative ‫ה(‬ ‫לֵ‬ ‫)ּכְ‬ and then overcorrected by adding a ‫.ה‬ Once ‫וכלה‬ is read as an imperative there is no meaningful role for ‫,לא‬ which could have been absorbed into ‫לאיתי‬ to create ‫וכלה*‬ ‫‪-a‬לאיתיאל‬ reliable retroversion of α´ is born.It is easier to arrive at a retroversion of θ´.The letters of ‫ולא‬ simply have to be reversed and redivided and ‫אוכל*‬ ‫ולא‬ ‫לאיתי‬ becomes ‫ואוכל*‬ ‫.לאיתיאל‬

Text Criticism, Interpretation, and the Life of the Text Proverbs
30:1b was always a challenging string of letters and it has been understood and interpreted in various ways throughout its history.Early on, at least one scribe expected an addressee and read a proper name.The resulting texts, one with a proper name and one without, were eventually conflated producing a double reading.Midrashic traditions came into play to explain the text.In time, various readings calcified.If I were editing a critical edition of Prov 30, I would probably offer a of θ´ as the "the earliest recoverable Hebrew text" of v. 1b, e.g., ‫ואוכל*‬ ‫.לאיתיאל‬But I remain convinced this "earliest recoverable Hebrew text" is not the "earliest inferable textual state."61I believe the text of Prov 30:1b could have once read ‫אוכל*‬ ‫ולא‬ ‫,לאיתי‬ "I am weary and powerless."But this reconstruction still necessitates interpretation.I have argued for reading ‫ואכל‬ as 1cs qal yiqtol from ‫יכל√‬ with support from θ´, S, the midrashic tradition, and ‫ואוכל‬ in the Kennicott MSS.Reading ‫יכל√‬ then supports the contextual reading of ‫לאיתיאל‬ on analogy with the collocation of ‫לאה√‬ + negator + ‫יכל√‬ in Isa 16:12 and Jer 20:9.As an opening to Agur's discourse, "I am weary and powerless" offers intriguing possibilities by connecting to Agur's core themes of human finitude and the location of wisdom.