1 Like a Lion (כ+ארי) […] My Hands and My Feet

In Ps 22:17 the Masoretes marked כְּלָבִים with a disjunctive ʿoleh weyored ( ֥ ֫ ) and placed a disjunctive ʾathnaḥ ( ֑ ) under הִקִּיפוּנִי, producing three distinct cola:³

The first two cola are written in synonymous parallelism and form an A–B/B’–A’ chiastic structure:

View Full Size

• Download Figure
• Download figure as PowerPoint slide

The last colon can be translated literally as, “like a lion my hands and my feet”.

The presumed advantages of this reading are that it follows the overwhelming majority of extant Hebrew manuscripts and coheres thematically with vv. 14 and 22 where similes involving a lion also appear.

Against this reading, the third colon is syntactically difficult because it lacks a verb to explain the relationship between the lion and the psalmist’s hands and feet, a relationship that is not immediately obvious from the broader context. Commentators who accept the MT have usually tried to address this problem by positing that a verb was either omitted through faulty transmission or elided (i.e. purposely omitted as unnecessary and disruptive to the poetic structure) by the psalmist and so must now be inferred by the reader.

Abraham Cohen opted for the first of these options, suggesting that the original verb resembled כארי and was omitted via haplography.⁴ He hazarded no guess as to what this verb had been but cited the targum for consideration: נכתין היך כאריא אידי ורגלי (“Like a lion they have bitten my hands and feet”). It seems more likely, however, that the targumist simply added the verb in order to convey what he thought the psalmist was trying to say. For the verb to have dropped out by haplography we would expect for it to have resembled כארי. But נשכו, the Hebrew equivalent of the targumist’s נכתו, looks nothing like כארי in any of the ancient scripts; the same is true of the qal participle, נשכים, which would be the precise equivalent of נכתין.‪⁵‬ A similar criticism could be made of the Jewish Publication Society (JPS) Tanakh translation of 1985: “Like lions [they maul] my hands and feet.” The words “they maul” appear to translate a conjectured טרפו (“they have rent, torn”) (cf. v. 14 where the verb occurs in connection with a lion).⁶ But again, since this verb would not have resembled כארי in any of the ancient scripts it seems unlikely to have dropped out via haplography. אכלו (“they have eaten”) would be an equally unsatisfying conjecture, both for the reason just mentioned and because lions do not typically eat the hands and feet of their prey.⁷ There seem to be no presumptive grounds, then, for conjecturing that a verb was accidentally omitted.

If the psalmist elided the verb, the most reasonable assumption would be that the last-mentioned verb (הקיפו) or a copulative verb (היו, “they are”) was meant to be read into the text. Yet neither of these options seems viable here. An individual lion surrounding or circling the psalmist’s hands and feet would make for a strange image.⁸ Nor is it possible to read a copulative verb into the statement, despite the apparent attempt to do so in the JPS Tanakh translation of 1917: “Like a lion, they are at my hands and my feet” (my emphasis). The word “at” cannot be derived from the Hebrew text without emending it. On the basis of Isa 38:13, Rashi suggested that the verb “crush” (שׁבר piel) can be read into the text while the New English Translation (NET) renders, “like a lion they pin my hands and my feet.”⁹ These are only guesses, however. One could insert other verbs into the text with equal justification.¹⁰ From a text-critical standpoint, this kind of approach is useless. Moreover, lions do not attack the hands and feet of their prey as this is an ineffective way to kill.¹¹ They certainly do not crush, much less pin, the hands and feet of their prey.

Although כארי is the most widely attested reading among the preserved manuscripts, its earliest witnesses are relatively late, namely, the sixth century Cairo Geniza palimpsest of the Hexapla (ὡς λέων) and the Psalms Targum (כאריא) which is probably later.¹² The oldest preserved reading is כ(א)רו, the evidence for which shall be discussed under no. 3 below.

From a structural standpoint, the two mouth+lion references in vv. 14 and 22 form an inclusio, while a chiasmus can also be detected:¹³

In support of כארי it could be argued that a lion reference in v. 17c would result in a central D section. However, there are at least two reasons to think that v. 20 is at the center of this poetic structure: (1) this is where the psalm shifts from complaints (vv. 13–19) to petitions (vv. 20–22); (2) ואתה signals a thematic emphasis, as in v. 4, which stands at the center of another poetic structure in vv. 2–6:

Although lion similes occur in vv. 14 and 22, the spelling for “lion” is not ארי but אריה and that is also how the term is consistently spelled elsewhere in the book of Psalms. One could possibly argue that in Ps 22:17 the author chose a different spelling in order to create some stylistic variation. But other terms for “lion” could have easily been chosen if variation were the concern: e.g. כפיר (Ps 34:11; 35:17); לביא (Gen 49:9; Num 23:24;); שחל (Job 4:10; Ps 91:13). Given the preserved variants of כארו and כרו (see no. 3 below), as well as the syntactical problems that result when כָּאֲרִי is read, the alternate spelling here can be arguably viewed as further indication that the MT is corrupt. Indeed, if the psalmist had written a verb that was morphologically similar to כארי (see no.’s 3, 4, 6, 7, and 11 below) the other lion references in the psalm might well have encouraged copyists to misread this verb as כ+ארי.

2 Like a Lion (כ+ארי) They Circumscribe My Hands and My Feet

Kristin Swenson¹⁴ rejects the MT’s tripartite structure, proposing instead that the verse be arranged into two cola, the first ending with מרעים:

View Full Size

• Download Figure
• Download figure as PowerPoint slide

In her second colon Swenson understands the psalmist to be saying not simply that his enemies had encircled or surrounded him but that they had circumscribed his hands and feet – i.e., restricted or constricted their movements/ability. She then understands “my hands and feet” metaphorically, as referring to the psalmist’s ability to fight or flee. A lion can strike its prey with such gripping fear that it can neither fight nor flee. That, Swenson suggests, is what the psalmist had in mind here when speaking about being “circumscribed” in his hands and feet.

Swenson’s analysis of this verse attempts to deal with the syntactical difficulty caused by the missing verb in the third colon of the MT. Structurally, it might also seem to cohere better with the rest of the psalm which consists mostly of bicola (but cf. vv. 8, 16, 24, 27).

But several objections can be offered against this analysis:

• It is lexically dubious that the hiphil of נקף could mean “constrict, restrict”.¹⁶ In the Hebrew Bible and the Qumran scrolls the qal or hiphil of נקף can mean the following: “encircle” or “surround” (1 Kgs 7:24; Isa 15:8; Lam 3:5; 1QpHab 4:7); “circumambulate, complete a circle/cycle” (e.g. Josh 6:3; Ps 48:13; Isa 29:1); and perhaps “round off” or “curl” (Lev 19:27).¹⁷ If the psalmist had wanted to say that he felt constrained or restricted, several other verbs would have been available: e.g.עצר, “restrain, detain”;¹⁸ גבל, “set bounds/boundaries”;¹⁹ סגר, “shut, enclose”; ‪רפה‬²⁰, “drop, sink, weaken”.²¹ The evidence that Swenson adduces to show that נקף could mean “constrain, restrict” is neither extensive nor compelling: Josh 6:3, 11; Lam 3:5.²² When used in Josh 6:3, 11, the verb speaks of a religious procession around Jericho which cannot be equated with a constraint/restriction of movement. Lam 3:5 is perhaps more apposite since the verb is used in synonymous parallelism with בנה עלי (“he has built [siegeworks] against me”), which implies that the speaker feels confined/restricted. The text is possibly corrupt, however.²³ Swenson might have also cited 2 Kgs 6:14, where soldiers are said to have encircled (יקפו) the city of Dothan, and 1QpHab IV:7, where the “Kittim” are said to encircle (יקיפום) “the fortresses of the peoples” with a great army in order to capture those inside. Restricted movement is certainly implied here on the part of those holed up in the city/fortresses. Yet this idea of re-/constricted movement is something one can only infer from the distinctly military context in which נקף is used. In Ps 22:17 the context is very different. Here it is not a city being surrounded by an army but an individual whose hands and feet are being surrounded/circumambulated by his enemies “like a (single) lion” (assuming here that: Swenson is correct in her construal of the syntax; כארי was original; the verse is a bicola). Swenson’s argument that ידי ורגלי are metaphors for the psalmist’s ability to fight or flee mitigates this objection somewhat. But this interpretation is highly metaphorical (see the final criticism below). The psalmist seems rather to be employing imagery, not metaphor.

• The pronominal suffix on הקיפוני indicates that the action is not directed against “my hands and my feet” but against “me”. Swenson tries to address this problem by appealing to texts where the verb’s object is first indicated by a pronominal suffix and then by a separate noun (e.g. Exod 2:6: ותראהוּ את הילד; ‪35:5:‬ יביאֶהָ את-תרומה; ‪1 Kgs 21:13:‬ ויעדהוּ … את-נבות; ‪Prov 5:22:‬ ילכדנוֹ את-הרשע).‪²⁴‬ In such cases the pleonastic noun stands in apposition with the pronominal suffix (e.g. “And she saw him, [namely,] the boy”, etc.). Swenson admits that the pleonastic noun is typically preceded by את but hastens to add that this is not always the case, citing Ezek 3:21 (הזהרתוֹ צדיק) as proof. Even so, the fact that she must appeal to an exceptional grammatical phenomenon weakens her argument. Moreover, none of her examples really corresponds to the grammatical situation in Ps 22:17 because “hands” and “feet” do not agree in person, number, or gender with the pronominal suffix on the verb. Swenson contends that while the agreement is not strictly grammatical it makes sense conceptually since “my hands and feet” can be understood as a merismus, referring to the psalmist’s entire being, not his hands and feet only. Yet there are no other instances in the Hebrew Bible or the Qumran scrolls where the phrase “hands and feet” is used in this way. Even if one could adduce such an instance, perhaps from another Semitic language, a merismus in our text would seem unlikely since lions do not encircle or surround their prey.²⁵ All things considered, then, Swenson’s proposed second colon ought to have been translated: “They surround me like a lion my hands and my feet.” This is grammatically nonsensical.

• By joining כארי with הקיפוני Swenson undermines the parallelism and the A–B/B’–A’ chiastic structure that would otherwise result if the verse were structured into three cola rather than two.

• Swenson’s interpretation is highly abstract, metaphorical, and elliptical, even for poetry: “Dogs have surrounded – a congregation of evildoers (i.e. a ‘pack’ of dogs) has encircled me (constraining me, as when an army surrounds a city or fortress and effectively barricades those inside), like a lion that is constricting (by means of fear) my hands and feet (i.e. my entire person and/or my ability to fight/flee).” One suspects that this interpretation takes too many liberties and reads more into the verse than is warranted.

3 They Have Dug (כארו, from כרה) My Hands and My Feet

The Septuagint of our verse (= 21:17) reads as follows:

View Full Size

• Download Figure
• Download figure as PowerPoint slide

In agreement with the MT, the Greek translator appears to have identified three cola in this verse (note that there are three verbs and that περιέσχον agrees in number with συναγωγή but not with κύνες). Unlike the MT, however, the translator did not read כארי (“like a lion”) in his Hebrew Vorlage but a third person plural verb: כארו or כרו. There is no כאר root in Hebrew. Hence, if the translator read כארו he must have understood the ʾaleph as a mater lectionis (כָּארוּ) and derived the verb from the third-weak root כרה (“to dig”).²⁶ Tertullian (Marc. 3.19) and Jerome’s Vulgate (ad loc.) likewise read foderunt (“they dug”); but since they were likely following the Septuagint here they cannot be cited as independent support for a Hebrew manuscript tradition attesting כ(א)רו. The Syriac Old Testament translator also reads a verb – bzʿw – and this can, perhaps, be rendered into English as “they pierced”.²⁷ But again, it is uncertain whether the translator was working with only a Hebrew Vorlage or was being influenced by the Septuagint.²⁸ Aside from the Septuagint, the best support for “they dug” comes from surviving Hebrew manuscripts. Several medieval manuscripts attest כארו.‪²⁹‬ Even more significantly, this reading is now attested in 5/6ḤevPs XI 9, which the editors have assigned “a late Herodian date (c. 50–68 CE)”.³⁰ Additionally, כרו is attested in two medieval manuscripts and is thought to be attested in 4QPs^f [4Q88] (ca. 50 BCE).³¹ In an article from 1897 Henri Lesêtre observed that although Justin Martyr quotes Ps 22 for his Jewish interlocutor Trypho and appeals to it as a proof-text for Christ’s crucifixion, he never pauses to consider Jewish objections to the Septuagint rendering of v. 17. Since Justin is aware of other Jewish objections to Septuagint renderings – as in Dial. 67 where the term παρθένος in Isa 7:14 is discussed at length – Lesêtre hypothesized that כארו was still the established reading in the mid-second century.³² Gregory Vall characterizes this as an argumentum ex silentio but the silence is all the more striking now, given the readings from the Dead Sea Scrolls (which were yet unknown to Lesêtre).

The Septuagint’s handling of our disputed text has several merits:

• Unlike some of the proposals examined below, it does not require any emendations to the text or appeals to otherwise unattested Hebrew verbal roots. As was just mentioned, כארו and כרו are both attested in extant Hebrew manuscripts and both readings are easily taken as derivatives of the third-weak root כרה, which translates as “dig”.³³ There can be little doubt, therefore, that the Septuagint translator was using a Hebrew Vorlage that attested either כארו or כרו. Surviving Hebrew manuscripts attesting these two readings confirm that the Septuagint Vorlage was not idiosyncratic but representative of an established text tradition.

• By employing a verb (rather than the noun ארי) the Septuagint avoids the syntactical difficulties posed by no.’s 1 and 2 above.

• ὤρυξαν/כ(א)רו fits with the psalm’s broader context, at least if this verb can be understood to mean “they pierced”.³⁴ The psalmist has already spoken in vv. 13, 17a–b about being surrounded by his enemies, and his capture is presumed to have already taken place in vv. 18–19; he also complains of being at the point of death in vv. 15 and 16, and of being surrounded/threatened by savage beasts and “the sword” (a metaphor for violent death?) in vv. 13–14, 17a, 21–22. A death by crucifixion would not be inappropriate to this context.³⁵ Alternatively, the psalmist could be seen as returning to his earlier imagery of dogs, which are known for biting (i.e. digging/piercing with their teeth) the extremities when attacking.³⁶

• If one posits an original כרו (read: כָּרוּ), it is easy to account for the emergence of the known variants. Without vocalization, כארי and כארו were morphologically very similar since waws and yods were often indistinguishable and easily confused. Moreover, the other references to lions in the psalm might have encouraged a copyist to misread כארו as כארי. As we have seen, the כארי reading deprives the colon of a verb and this omission is very difficult to explain text-critically; hence, כארי is better seen as a corruption of כארו, and not vice versa. כרו could then be seen as the original orthography for this verb, since it does not employ the mater lectionis (a later convention) and can explain the other readings: כרו ‪>‬ כארו ‪>‬ כארי.

But there are problems with the Septuagint’s reading as well. As we shall see below, interpreters who postulate an original כרו do not necessarily derive this verb from כרה, or from a root that means “dig”. Also, the remark, “They dug my hands and my feet”, is a bit baffling. Above it was suggested that the psalmist was thinking of a crucifixion. English translators of Ps 22:17 often render כארו (or ὤρυξαν in the Septuagint) as “they pierced” (e.g. Brenton’s LXX translation, KJV, NIV, ESV, NASB). This translation evokes the image of Jesus’ crucifixion and has therefore seemed agreeable to many Christians.³⁷ There would have been more obvious Hebrew words for “pierce” than כרה, however: e.g. דקר (e.g. Zech 12:10), נקב (e.g. Isa 36:6), חלל (poel), and טען (pual). The alternative interpretation mentioned above (about the psalmist thinking here about dogs biting his hands and feet) poses difficulties as well. The imagery of dogs “digging” at their prey with their teeth would be unparalleled, at least in the Hebrew Bible. Dogs do not dig with their teeth but with their claws. More obvious words for “bite” would have been נשך or אכל.

4 They Have Made My Hands and My Feet Repugnant (כֹּאֲרוּ, from כאר)

Franz Delitzsch suggested that Aquila, in his original translation of this verse, derived ᾐσχύναν (“they shamed, rendered contemptible”) from the Aramaic loan word כאר, “to soil, make repugnant”.³⁸ Bernhard Duhm, in his commentary on in Ps 22:17, took his inspiration from Aquila and vocalized the verb as a poel, כֹּאֲרוּ, translating: “Entstellt sind meine Hände und Füsse” (“My hands and feet are disfigured”).³⁹

This option is attractive because it relies on an extant reading (כארו) and hence does not require an emendation.⁴⁰ True, in Aramaic, the term for “make repugnant” is כער, with an ʿayin rather than an ʾaleph (e.g. Tg. Nah on 3:6); but there is evidence that in Hebrew the root could be alternatively spelled with an ʾaleph.⁴¹

Among the difficulties with this rendering are, first, that the poel is an infrequent stem formation that normally goes with geminate verbs.⁴² Second, כאר II (“make repugnant”) seems to have been an infrequently used root. More common ways of expressing repugnance, abhorrence, shamefulness, etc. were with roots like שקץ, ‪בזה‬, חלל, ‪בוש‬, חפר (cf. also the noun חֶרְפָּה), and כלם (hiphil or the noun כְּלִמָּה). Third, if any copyists had understood the verb as a poel we would expect to find in the manuscript record the insertion of a waw (mater lectionis) after the kaph in order to assist readers with the verb’s proper pronunciation and meaning. Fourth, while Duhm’s “entstellt sind” may be translationally satisfying it is philologically dubious. A more literal translation would be: “they rendered as repugnant/disgusting my hands and feet”. No explanation is given as to how or why the psalmist’s “hands and feet” had been rendered thus by his enemies. Fifth, if indeed Aquila read the verb as כֹּאֲרוּ he may have been more indebted to the Aramaic of his own day than the psalmist’s original meaning in Hebrew. Sixth, כֹּאֲרוּ would be a more convincing reading here if a Semitic parallel could be adduced showing an individual’s hands and feet being “rendered repugnant, disgusting, etc.” in a context similar to what we have in Ps 22:17. No such parallel has so far been adduced.⁴³

5 They Have Bound (Emend to אסרו, from אסר) My Hands and My Feet

Gregory Vall⁴⁴ argued that the psalmist originally wrote אסרו (“they bound”) but this was corrupted, via metathesis, to סארו. Since the corrupted term was nonsensical, scribes “corrected” it to כארו, ‪כארי‬, or כרו.‪⁴⁵‬ Although no extant Hebrew manuscript attests אסרו or סארו, Vall cites as an ancient and independent witness the Greek verb ἐπέδησαν (“they have bound”), which appears in Aquila’s revised – though not his original – translation. Vall also cites Symmachus’ ὡς ζητοῦντες δῆσαι (“like those seeking to bind”) and Jerome’s vinxerunt (“they have bound”), which appears in his Psalterium iuxta Hebraeos.

Vall’s proposal here has at least three strong points. First, it is not purely conjectural but relies on three ancient versions. Second, unlike the two other proposals considered below (no.’s 6–7), Vall’s makes use of a well-known Hebrew word for “bind” (see, e.g., Gen 42:24; 49:11; Judg 15:10ff.). Third, Vall’s proposal fits with Ps 22’s context. The psalmist appears to envisage himself as someone who has been, or is about to be, taken captive by his enemies (cf. vv. 5, 6, 9, 13, 21, 22). The act of binding a war captive’s hands and feet is alluded to in 2 Sam 3:34. A comment by the psalmist about his hands and feet being bound by an assembly of evildoers would have therefore made for some apt and poignant imagery.

On the other hand, Vall’s hypothesis is quite speculative since neither אסרו nor סארו is attested in any surviving Hebrew manuscripts and since it requires a sequence of two copyist errors – metathesis followed by an adjustment of ס to כ – in order to account for the existence of כ(א)רו and כארי in the manuscript tradition. The plausibility of Vall’s hypothesis is further reduced because “samekh and kaf do not look especially alike in the ancient scripts”, as Vall himself acknowledges.⁴⁶ It seems dubious that the translations by Symmachus and Jerome represent independent witnesses of a Hebrew Vorlage. Aquila’s revised translation was known and used by both translators so both men may have simply deferred to Aquila. Of course, Aquila himself could have accessed a Hebrew Vorlage attesting אסרו. But he might have had other reasons for using ἐπέδησαν:

• He read כארו or כרו and derived this from another root that meant “bind” (see no.’s 6 and 7 below).

• He did not know how to interpret כ(א)רו and chose ἐπέδησαν, first, because it went well with ידי ורגלי and, second, because it deprived Christians of a favorite prophecy about their Messiah’s crucifixion. The Septuagint translation (ὤρυξαν χεῖράς μου καὶ πόδας) had certainly been used by early Christians for apologetic purposes.⁴⁷ And Aquila’s own earlier translation (ᾐσχύναν, “they have dishonored, disfigured”) had probably also been used in this way, much to that translator’s chagrin.

6 They Have Bound (כארו, from כור/כאר) My Hands and My Feet

John Kaltner⁴⁸ accepts the minority כארו reading and agrees with Vall that the psalmist had originally written something about his hands and feet being bound. He rejects Vall’s hypothesis about scribal corruption, however, positing instead the existence of a middle-weak Hebrew root כור/כאר with the meaning of “bind, tie.”⁴⁹ As support for this conjectured root he cites the Arabic kwr.

The advantage of Kaltner’s proposal, over against Vall’s, is that while it also fits nicely with the psalm’s context and harmonizes with the readings of Aquila, Symmachus, and Jerome, it does not require a sequence of two or more scribal errors to account for the extant Hebrew readings.

On the other hand, conjecturing a Hebrew root on the basis of an Arabic one is precarious. This is especially true in the present case since Hebrew already had a perfectly suitable and frequently used word for “bind”: אסר.‪⁵⁰‬ Furthermore, the philological evidence that Kaltner cites from Arabic does not support a meaning of “bind, tie” but of “wrap, wind” (as one wraps/winds a turban), “increase” (as one increases a turban’s size by adding wrapping), or “gird” (as one girds a camel with a saddle).⁵¹ If the psalmist had originally spoken about his hands and feet being bound by evildoers he would have presumably meant that he felt constricted, unable to escape. But turbans were not wrapped/wound and saddles were not girded in order to constrict movement but to facilitate it. Kaltner’s argument might be more convincing if he could show that kwr had been used with reference to one’s hands/feet, or in connection with shackles, ropes, or other restrictive items. Kaltner himself finds another problem with his conjectured root: the middle radicals in the terms כאר and kwr are not equivalents since the Hebrew ʾaleph corresponds to the Arabic ʾaliph, not wāw. Kaltner downplays this problem, commenting that “this type of interchange, particularly with middle radicals, is a well-documented phenomenon in Semitic languages.”⁵² Swenson finds this argument moot, noting that “א does not behave as י and ו do in the second position of the triliteral root”.⁵³ What she means, presumably, is that in hollow verbs the second character of the lexical form – whether י or ו – is not even used in the qal perfect. Kaltner could still argue, however, that the aleph is a mater lectionis which was used here to indicate the long a (qameṣ) vowel (cf., e.g. Hos 10:14).

7 They Have Bound (כארו, from כרה) My Hands and My Feet

G. R. Driver also accepted that a third person plural verb had been used by the author of our disputed text. He rejected the Septuagint’s rendering, however, supposing instead that the psalmist was claiming his hands and feet to have been bound. But unlike the earlier proposals deriving the verb from אסר (Graetz/Vall) and כור/כאר (Paulus/Kaltner), he suggested that it derived from the third-weak root כרה, pointing to evidence from cognate languages: e.g., the Akkadian verb karāru (“tie, be twisted”), the Arabic roots krr (“wind”) and kwr (“wrap, wind”), and the Syriac kār (“press together”).⁵⁴ As circumstantial evidence from the Hebrew Bible itself, he argued that in Ezek 16:4 the Septuagint translator misread כָּרַּת (“it was cut”) as כָּרוּ and translated it as ἔδησαν (“they bound”). Driver further argued that since the psalmist has already spoken of his own death in v. 16c a reference to the binding of his hands and feet in v. 17 would seem anticlimactic and out of place. He therefore followed Sigmund Mowinckel in re-locating v. 16c (ולעפר-מות תשפתמי) immediately after v. 17.⁵⁵ He then translated אספר כל־עצמותי in v. 18 as, “I recount all my great sufferings” (cf. Job 7:15; Ps 40:6) and emended תִּשְׁפְּתֵנִי (“you have cast me”) in v. 16 to a defective 3 pl. שְפָתֻנִי (“they have cast me”). Thus:

Against Driver, אסר is the normal root for describing the binding of one’s hands and feet. Also, J. J. M. Roberts observes that Driver’s supposed Akkadian cognate karāru “is really q/garāru, ‘to wind, coil’ (von Soden, Attw, 902–3) or ‘to turn, roll over’ (CAD g, 47–48 [2nd ed.: 285])” whereas “karāru means ‘to put in place’ (CAD k, 207 [148]).”⁵⁶ Nor is Driver correct in claiming that the Arabic verb krr means “wind”. It means, “turn against, return, withdraw”.⁵⁷ Kwr, the other Arabic root that Driver cited, is likewise unhelpful, not only for the semantic reasons that were mentioned in the above critique of Kaltner (no. 6), but because it would correspond to the hollow verb כור, not the third-weak כרה. Finally, Driver’s rearrangement of v. 16c is highly arbitrary, as is his emendation of תִּשְׁפְּתֵנִי.

8 My Hands and My Feet Have Been Torn (Emend to קָרֲעוּ, from קרע)

Hans Schmidt⁵⁸ suggested that the original verb in Ps 22:17 was קָרֲעוּ (“they have been torn/rent”) but a copyist misheard it as כָּארוּ (“they have dug”). On this conjecture we can perhaps imagine that the psalmist was envisaging his enemies as dogs who were tearing or ripping his hands and feet apart (cp. no.’s 11–12 below).

Schmidt’s proposal can be challenged on several counts. First, there is no external evidence for קרעו. Second, כָּ(א)רוּ could have only been audibly confused for קָרֲעוּ at a somewhat late period, after ʿayin had become quiescent; hence, if קָרֲעוּ had been original its lack of attestation in surviving witnesses seems all the more surprising. Third, dictation was typically used when multiple copies were needed; and while the Hebrew psalms would eventually be “mass produced”, this can probably not explain the early readings of כָּ(א)רוּ by the Septuagint translator or the כארו reading in 5/6ḤevPs XI 9 and the (apparent) כרו reading in 4QPs^f. Fourth, I am not aware of any relevant parallels in which dogs are described as tearing or ripping at one’s hands and feet; we should probably have expected “they bit”.

9 My Hands and My Feet Are Wasted Away/Exhausted (Emend to כלו, from כלה)

E. J. Kissane conjectures an original כלו (from the third-weak כלה root), noting that lameds were sometimes miscopied as reshes.⁵⁹ This commonly used root connotes the idea of completion or bringing something to an end. Kissane also departs from the MT by reading the third colon in v. 17 with the first one in v. 18:

Thus arranged, v. 17c and 18a would suggest that the speaker is languishing in the envisioned situation. As key support for his proposed emendation and restructuring of the cola in the MT, Kissane points to a possible parallel in Job 33:21:

View Full Size

• Download Figure
• Download figure as PowerPoint slide

The second colon here is nicely parallel with Ps 22:18a. Kissane points out that the first colon would also parallel Ps 22:17c if only the resh in כארו could be emended to a lamed.

Peter C. Craigie follows Kissane’s suggested emendation but translates, “My hands and my feet were exhausted”.⁶¹

כ(א)לו is among the more convincing conjectural emendations that have been proposed for this verse. Not only does it have an impressive parallel in Job 33:21 but it makes good sense structurally since it meaningfully connects the otherwise dangling third colon in v. 17 with the first one in v. 18.

Nevertheless, the fact that כ(א)לו is not supported by any Hebrew manuscripts or corroborated by any of the ancient versions makes it suspect. Indeed, since a resh would only have been confused with a lamed in later scripts one would expect to find some evidence of its prior existence among surviving witnesses.⁶² Moreover, on this reading one would have to wonder why the psalmist is offering this kind of lament. Is he complaining about his generally emaciated condition? If so, should we not have expected him to speak of his entire body or at least some other part of it? Hand and foot bones are among the most visible in the human body, regardless of how well-fed and healthy a person is; hence, it would seem odd for the psalmist to emphasize the visibility of those bones. In the Hebrew Bible the root כלה is used most frequently of the eyes (e.g. Job 17:5; Ps 69:4; Lam 2:11; 4:17). It is also used of a person’s “reigns” (Job 19:27); “flesh” (Job 33:21); “life” (Ps 31:11); “heart and flesh” (Ps 73:26); “soul” (Ps 84:3; 119:81); “spirit” (Ps 143:7); “flesh and body” (Prov 5:11); entire being (Ps 37:20; 39:11; 119:87; Mal 3:6). It is never used of either the hands or feet, or of both together. Neither Kissane’s nor Craigie’s translations seem to reflect a known idiom, therefore, at least in biblical Hebrew. Perhaps someone will one day adduce an ancient text that refers to a person’s hands and feet wasting away or being exhausted due to hunger or some other situation that would fit the context of Ps 22. Until then this proposed emendation will require philological support.

10 My Hands and My Feet Hurt/Are Pained (Emend to כאבו, from כאב)

Another proposed conjectural emendation regards כארו as a corruption of כָּאֲבוּ (“they hurt, are pained”).⁶³ This proposal has some plausibility because the letters beth and resh are morphologically similar in several ancient scripts. It must be treated with some suspicion, however, since it is not supported by any surviving witnesses (but cf. n. 27 for Syriac). Moreover, one would have to explain why the psalmist is calling special attention to the pain in his hands and feet. Perhaps we can imagine the dogs at the beginning of the verse biting at these extremities. But a connection with these dogs is difficult to infer here because of the intervening colon and because the dogs were not said to have been biting. Hence, we should have expected the third colon to make this connection clearer: e.g. “My hands and feet hurt because dogs are biting them”. Without any such clarification, the assertion, “My hands and feet hurt”, would read like a random or undeveloped thought. Another possibility is that the psalmist was imagining his hands and feet being amputated by his enemies (cf. 2 Sam 4:12; 2 Macc 7:4). But if the psalmist wanted to evoke this imagery he would have presumably used a verb like קצץ or כרת, not the vague כאב. A more basic problem here is that pain cannot be experienced in one’s hands and feet once they have been amputated.

11 Because They Picked Clean or as If to Rip Apart (כארי\ו, from ארה) My Hands and My Feet

Mitchell Dahood accepts the כארי reading and thinks that it derives from כִּ+אָרְיֻ, that is, an irregularly formed causal conjunction and an irregular 3 pl. perf. ending of the root ארה (“to pluck, hack in pieces”).⁶⁴ Like Kissane, he also joins v. 17c with v. 18a in order to form a distich:

R. Tournay⁶⁶ reads כֶּאֱרוֹ, that is, a comparative kaph prefixed to an irregularly formed infinitive construct of the root ארה. He translates, “comme pour déchiqueter (mes mains et mes pieds)” [“As if to rip apart (my hands and my feet).”], and then paraphrases:

A pack of dogs, a figurative description of the villains, encircles and surrounds the unfortunate, as if they wanted to rend him with claws and teeth, tearing all his limbs into pieces, so much so that he is able “to count all his bones” (v. 18).⁶⁷

Deriving כארי or כארו from the root ארה has two advantages: it requires no emendation to the text and no conjectures about otherwise unattested Hebrew roots.

But the proposed readings/interpretations of Dahood and Tournay are problematic because they:

• rely on orthographies that are highly irregular or unparalleled in classical Hebrew: כִּ rather than ארְָי ;ֻכִּי rather than אָרוּ; ‪אֱרוֹ‬ rather than אֱרוֹת);

• rely on lexically dubious translations: ארה was an agricultural term, referring to the gathering or plucking of fruits and myrrh from a tree, bush, etc. (Ps 80:13; Song 5:1; m. Šeb. 1.2); it is never used to speak of the plucking clean/picking at meat from the bones, or the ripping of a carcass to pieces;

• produce a difficult sense: why would the “assembly of evildoers” be picking clean the psalmist’s hands and feet like cannibals eager to consume the last bit of his flesh?

Tourney’s approach mitigates the last of these problems by attributing the flesh-eating act to the dogs in the first colon. But this requires some questionable assumptions. Tourney understands עדת מרעים in the second colon as the implicit subject of כֶּאֱרוֹ. This phrase does not seem to be an equivalent for כלבים, however, despite the parallel structure. Parallel expressions occur elsewhere in biblical Hebrew and these always refer to humans (e.g. Ps 26:5: קהל מרעים; ‪86:14:‬ עדת עריצים; ‪Job 15:34:‬ עדת חנף; ‪Isa 31:1:‬ בית מרעים). Hence, it would be exegetically difficult to connect the imagery about the dogs in the first colon to the third colon.

12 Because They Picked Clean (= No. 11) and Like a Lion (= No. 1)

James R. Linville⁶⁸ proposes a “very complex” exposition according to which double meanings can be detected in both הקיפוני (“they have encircled me” and “they have dismembered me”) and כארי (“they have picked at” and “like a lion”), and “multiple meanings” in ידי ורגלי.

Linville’s approach mitigates the first and last of the three problems that we mentioned under the last heading.⁶⁹ He argues that the psalmist was playing fast-and-loose with the rules of orthography in order to create word plays. Since כארי can be read both as “they picked clean” and “like a lion”, the flesh-eating act can be appropriately connected with the lion rather than the עדת מרעים (which would presumably refer to humans).

Linville may be suspected here of wanting to “have his (exegetical) cake and eat it too.” One can find lexical and syntactical ambiguities in the works of almost any author, regardless of the language or the era when he/she wrote. But unless one has good reason to suspect that these ambiguities were deliberate it is safest to treat them as coincidental, otherwise one will have to adopt the absurd position that almost any ambiguously worded text was intended to have multiple meanings. Ambiguities are most often the result of careless writing. As support for his interpretation, however, Linville appeals to an article by Paul R. Raabe, which argues that the psalmists deliberately use multivalent words and phrases.⁷⁰ Raabe may have identified a real phenomenon but the examples that even he adduces are not all beyond dispute (e.g. Ps 4:5). More importantly, Linville’s examples of deliberate ambiguity in Ps 22:17 are unconvincing.

To elaborate on this last point, it is not enough to note that כארי can be read either as a noun (“lion”) or as a form of ארה (“pick at”), especially when the orthography for the verbal form (כִּ+אָרְיֻ) is so irregular and when כארי is already textually suspect due to the variant readings and the syntactical difficulties that it engenders. Linville therefore appropriately looks for other evidence in this verse to confirm his argument about deliberate ambiguity. He focuses on הקיפוני, citing Lev 19:27 and Ps 17:9 as proof that the root נקף could sometimes connote an act of violence and not just of circling and surrounding. If true, this would fit better with a lion metaphor since lions are obviously violent but are not known to circle or surround their prey.⁷¹ Linville notes that the prohibition in the first half of Lev 19:27, where the root נקף is used, stands in synonymous parallelism with a prohibition in the second half of this verse, where the root שחת (“destroy”) is used. This does not prove, however, that נקף could connote violence; for the “destruction” referred to here – i.e. the shaving one’s beard – hardly constitutes an act of violence. Ps 22:17 is not concerned with shaving in any case. Furthermore, in Lev 19:27 the words לא תקפו פאת ראשכם are somewhat ambiguous (cf. also Jer 9:25; 25:23; 49:32). Traditionally, they have been taken to prohibit the shaving of one’s sideburns and the cutting of one’s hair so that the hairline forms a ring around the head. The Septuagint translator, however, reads: “You shall not make a curl (οὐ ποιήσετε σισόην) from the hairs of your head”. If this is an apt rendering, the verse would contain a synthetic rather than a synonymous parallelism and Linville’s entire line of reasoning would be misguided. With regard to Ps 17:9, while נקף and שדד are used here in synonymous cola this does not necessarily mean that the verbs in these cola are synonymous, or even that they carry similar connotations. To be encircled by one’s enemies could, of course, be a threatening experience and could even be a prelude to violence, but it is not itself a violent act, much less an act of dismemberment.

13 My Hands and My Feet Are (a.) Shriveled Up (b.), Gone Lame, (c.) Contracted (כרי\ו, from כרה)

J. J. M. Roberts accepts either כרו or כארו as original. But unlike the Septuagint translator, Kaltner, and Driver, he does not think that the psalmist was speaking about his hands and feet being either “dug” or “bound”. Instead, he conjectures an otherwise unattested third-weak Hebrew root כָּרָה with a meaning similar to the Akkadian root karû, “to be short”. He then construes ורגלי ידי not as the objects but as the subjects of כרו, translating: “My hands and my feet are shriveled up, I can tell [i.e. count] all my bones”.⁷²

Roberts’ interpretation invites the same objection that was raised against Kissane’s: if the psalmist were describing his emaciated condition it seems odd for him to focus only on the visibility of the bones in his hands and feet, given that these are normally visible in a healthy body. Doubts can also be raised as to whether Roberts’ conjectured Hebrew cognate would have meant, “to shrivel up”. Roberts alleges that in Akkadian and Syriac there is a matching triconsonantal root which could, “in certain contexts”, be used “to indicate physical or mental infirmities”. But Vall observes that Roberts has inadvertently conflated two different Syriac roots: krʾ (third-weak), “to be short, sorry”, and krh (third strong), “to suffer, be sick”. “A Hebrew cognate for the latter”, Vall notes, “would require the consonantal ה”.‪⁷³‬ As for Roberts’ Akkadian root karû, this is standardly defined as follows:

Karû(m) II “to be(come) short” Bab., M/NA G (u/u, also i/i) [LÚGUD(.DA)] of space, of time; of heart, life “be(come) diminished” Gtn [LÚGUD.MEŠ] “contract repeatedly” of parts of body; of life “constantly become diminished” D “shorten” time, “diminish” s.o.’s life, “put in desperate straits”; stat. “(are) very short”, of parts of body etc.; MA “deduct, subtract”? Š “shorten” days; > kurû, kurītu^?; kurru; → tagrītum.⁷⁴

Thus, if the Hebrew verb in Ps 22:17 was related to the Akkadian root karû the colon should probably be rendered, “My hands and feet are shortened, contracted”, not “shriveled up”.⁷⁵ Moreover, the alleged parallel to Ps 22:17, which Roberts adduces from an Akkadian medical diagnostic text, does not seem apposite: “If in his sickness his mouth is paralyzed and his hands and his feet are shrunken (qātāšu u šēpāšu iktarâ), it is not a stroke, his sickness will pass”.⁷⁶ While the precise medical condition envisioned here is uncertain, it does not seem to deal with emaciation, as the context of our psalm would presumably require if the author had declared that his hands and feet were “shriveled up”. The Akkadian text may refer to muscular cramping or some bone disease, either condition being inappropriate to the context of Ps 22.

Michael L. Barré accepts Roberts’ proposed Akkadian background for כ(א)רי\ו but regards it as a medical loanword rather than a cognate. He appeals to the same Akkadian medical diagnostic text cited by Roberts, arguing that the hands and feet here are not “shriveled up” but immobile/non-functional, as when one experiences a stroke: the muscles in the person’s hands and feet are envisioned as contracting or cramping spasmodically. The author of Ps 22 utilized this medical terminology to describe the paralyzing effects of his enemies’ attacks. Barré translates v. 17 as, “My hands and my feet have gone lame.”

While understandably discarding Roberts’ rendering (“they are shriveled up”), Barré’s translation/interpretation poses its own difficulties. His rendering of כ(א)רי\ו as “gone lame” plays on an English expression which has little to do with the hands and feet being short(ened) – as the Akkadian root karû connotes – but with their being limp or paralyzed.

Barré’s alternate rendering, “are contracted”, is more justifiable philologically. Its validity would depend on: (1) his interpretations of two Akkadian medical diagnostic texts; (2) his suggestion that the psalmist was invoking an Akkadian loan word. His interpretations of the diagnostic texts seem plausible enough. But his appeal to an alleged medical loanword, which is itself attested in only two Akkadian texts and is unattested in Hebrew, is speculative.

Barré also emends אספר to ספד in v. 18a. Appealing to both a variant in the Septuagint (ἐξηρίθμησαν) and the standard reading of the Syriac (ʾyllw, “they wailed”) he proposes that ספד was either intended as an irregularly formed 3 m. pl. perfect (סָפְדֻ) or “an infinitive absolute serving as a finite verb (סָפֹד).”⁷⁷ This, he theorizes, was miscopied as ספר and read as an infinitive absolute. Afterwards, it was deliberately altered (“modernized”) to read אספר. He proposes that v. 18a originally meant something like, “All my bones have intoned my funeral dirge.”⁷⁸ The imagery here seems rather odd and unparalleled, however. In sum, while Barré’s suggested linguistic background and emendations are ingenious, demonstrating an impressive knowledge of the ancient languages, they are highly conjectural and add further complications to his exegesis.

Conclusion

It will be helpful here to list the thirteen proposals that have been discussed:

View Full Size

• Download Figure
• Download figure as PowerPoint slide

As was stated at the outset, the goal of this article has not been to offer a new proposal that solves the textual and interpretive cruxes in Ps 22:17; rather, the goal has been to survey and evaluate several proposals that have already been offered.⁷⁹ This kind of exercise is usually the first step in moving beyond a scholarly impasse since it helps to identify which of current proposals are least viable and how better to judge between the others.

None of the textual readings and interpretations examined in this article is without significant problems; but some are worse offenders than others. For reasons that need not be repeated, no.’s 2, 4, 7, 8, 10, 11, 12, and 13a–b (“are shriveled”, “are gone lame”) do not seem like viable options and should probably be excluded from future discussions.

The remaining six options – no.’s 1, 3, 5, 6, 9, and 13c (“are contracted”) – may be quickly reviewed. If כארי (“like a lion”, no. 1) is original we shall have to assume that a verb has dropped out, although such a corruption would be difficult to explain text-critically. The proposed emendation of אסרו (no. 5) is enticing but the process by which its allegedly corrupted forms – כ(א)רו and כארי – emerged, as well as the complete disappearance of אסרו from the Hebrew manuscript tradition, would, once again, be difficult to explain text-critically. The proposed emendation of כלו (no. 9) is also enticing for contextual reasons but is purely conjectural. If כארו is original it would be simplest to derive the verb from the commonly used third-he root כרה, “dig” (no. 3), although this seems to leave us with an awkward metaphor or one that is at least difficult to explain philologically (but cf. n. 34). It would be more speculative to derive the word from roots or definitions that are unattested in Hebrew and can therefore only be inferred from (supposed) cognates in other languages (no.’s 6, 7, 13c).

There seems no way of deciding between these remaining options without becoming quite arbitrary, given that they can only be evaluated using incommensurable and subjective standards. There is no obvious way, for example, to judge between no. 9, on the one hand, which seems contextually appropriate and finds an interesting parallel in the book of Job but has no manuscript support, and no. 13c, on the other hand, which may also be contextually appropriate, requires no emendation, and is supported by cognates in Akkadian but has no parallels in Hebrew writings. Again, there is no easy way to judge between no. 1, which is widely attested in the manuscript tradition and fits the psalm’s lion imagery but is grammatically nonsensical, and no. 3, which has the oldest textual support but creates an odd metaphor. Again, there is no obvious way to judge between no. 5, which fits the context and relies on a commonly used Hebrew term but requires a difficult emendation, and no. 6, which fits the context just as well and requires no emendation but relies on a dubious Arabic cognate.

The arbitrariness with which one judges these remaining options will likely be compounded by ideological concerns which scholars tend not to divulge and which often affect them at a largely unconscious level. For example, an orthodox Jew may find כארי (no. 1) more appealing than כ(א)רו (no. 3) because it coheres with tradition and deprives Christians of an important proof-text for their Messiah. A Christian, by contrast, may find כ(א)רו (no. 3) more appealing because it allows the text to be read as a prophecy about Jesus. Someone taking a secular approach may find one of the other options appealing since it is more detached from the Jewish-Christian debate and allows him or her to appear more neutral and dispassionate.

By highlighting the weaknesses of certain proposals, the present study has allowed us to narrow our options significantly. Given the difficulties in deciding between the remaining options, however, it seems unlikely, barring the discovery of new evidence, that we will be able to move much further beyond this. Studying the text is likely to tell us more about ourselves than what the psalmist originally wrote or meant.