Non-Roman fonts used: SPTiberian, SPIonic

A Proto-Masoretic Gloss in the Prayer for the King (Psalm 61:7-8)?

David Marcus

Jewish Theological Seminary

ABSTRACT

The form Nma in the prayer for the king in Psalm 61 is an old crux interpretum. It has traditionally been taken as a pi'el apocopated imperative from the verb hnFmf meaning "appoint!" The ancient versions took it as akin to the Aramaic interrogative Nma and translated it as "who?" Because the form interrupts the natural grammatical and poetic sense of the verse, many suggestions have been made to emend it, and the form has often been regarded as a gloss without which the verse would read more smoothly. In this paper we will examine the tantalizing suggestion that Nma represents a proto-Masoretic gloss standing for Nw%n )l'mf to indicate that the following form w%hrUc;n:yI is written with the letter nûn. We raise a number of objections to this proposal and demonstrate that if Nma represents a Masoretic note it would be the first example of its type.

1. Psalm 61 is a short psalm of nine verses and generally acknowledged to be an individual psalm of lament (Bellinger 1995: 37). However, two of its verses consist of a prayer for the king. These verses 7 and 8 read:

.rdowF rdo-wOmk@; wytfwOn#$; PysiwOt@ K7leme-ym'y:-l(a MymiyF
.w%hrUc;n:yI Nma tme)vwE dsexe Myhilo)v yn"p;li MlfwO( b#$'y"

Most modern translations render these lines similarly to the NRSV as:

Prolong the life of the king; may his years endure to all generations! may he be enthroned forever before God; appoint steadfast love and faithfulness to watch over him!

2. Taken by itself this prayer is not remarkable, as it consists of themes, language, and structures found elsewhere in the Hebrew Bible. Prayers for the life of the king, for granting him divine protection, and for imbuing him with the divine attributes of faithfulness and loyalty (tme)vwE dsexe) are quite common.¹ Our prayer is composed in good Hebrew parallelistic form: there is the synonymous parallelism of MymiyF and wytfwOn#$;, and the reverse parallelism of rdowF rdo and MlfwO(. The one verb in v. 7 (PysiwOt@) does double duty for both stichs in that verse, as is usual in biblical Hebrew poetry. But what is a little unusual in our couplet is the preponderance of verbs in v. 8. Here there is not one verb doing double duty nor two verbs in synonymous parallelism, but three verbs: b#$'y" "may he be enthroned," Nma "appoint!" and w%hrUc;n:yI "to watch over him."

3. While this traditional rendering is possible, it is awkward for a number of reasons. First, it is awkward because it means that the second half of the verse has two verbs in close proximity, one an imperative Nma ("appoint!") and the other w%hrUc;n:yI, which serves as a purpose clause ("to watch over him"). Second, the verb Nma not only interrupts the sequence of imperfects PysiwOt@, b#$'y", and w%hrUc;n:yI, but it is also a hapax legomenon, usually taken as an apocopated piel imperative from the verb hnm meaning "to appoint" (GKC, §75cc). A third difficulty is that it is extremely rare for an object, here tme)vwE dsexe "steadfast love and faithfulness," to come before an apocopated imperative, and Nma would represent the only example of this phenomenon in the Psalms. Most of the ancient versions took the form Nma as akin to the Aramaic interrogative pronoun "who?" Thus the Septuagint (Rahlfs 1979: II, 62), Gallican edition of the Vulgate (Weber 1994: 842), and the Peshitta (OTS: 67) read: "who (ti/j/quis/wnm) will seek out (his) mercy and truth?"² The Targum took the form Nma as the preposition Nmi "from" and added the divine source from which the king's qualities derive, "goodness and truth from (Nmi) the Lord of the Universe will protect him."³

4. A number of ancient traditions simply omitted the form Nma altogether. Aquila and Symmachus (Field 1875) leave it out, as does Jerome's "Hebrew" Psalter, which reads: misericordia et veritas servabunt eum "mercy and truth will serve him" (Weber 1994: 843). Some modern translations also omit the form Nma as, for example, the NEB, which renders: "may true and constant love preserve him." Omitting the form Nma leaves the stich with the identical phraseology as that which is said of the king in Prov 20:28: K7leme-w%rc@;yI tme)vwE dsexe "faithfulness and loyalty protect the king." Clearly a text without the Nma would seem to be a more preferable one from the point of view of grammar, poetry, and theme. Indeed, Michael Fishbane (Fishbane 1985: 64) has remarked that removal of Nma "restores an immediate and natural grammatical sense to the passage."

5. If the text reads better without the Nma, how can its presence be explained? There are a number of scholars who believe that the occurrence of Nma is due to a corruption in the original text. Suggestions have been made as to what the original text might have been. Briggs and Briggs (Briggs and Briggs 1906-1907: 2:67) proposed that Nma was originally NymiyF "on the right hand" (for protection), on the analogy of similar thoughts in Pss 45:10; 109:31; 110:1, 5. Rössler (Rössler 1962: 138-39) suggested, on the basis of Ps 40:12 ynIw%rc@;yI dymit@f K1t@;mi)jwA K1d@:s;xa, that Nma originally may have read dymit@f,⁴ thus providing a synonym to the parallel MlfwO( of the first stich.⁵

6. One of the most tantalizing suggestions made is that Nma represents a proto-Masoretic gloss. This suggestion was included in the Hebrew and Aramaic lexicon of Julius Fürst (Fürst 1867: 831). Discussing the form under the entry hnm, Fürst wrote: "as Nma is superfluous here and not necesary to the sense, some take it for a Masoretic gloss (n~m i.e. N )l'mf) to w%hrUc;n:yI."⁶ In other words, some suggest that the form Nma represents a proto-Masoretic note standing for Nw%n )l'mf to indicate that the following form w%hrUc;n:yI is written with the letter nûn. Normally in the imperfect of an initial nûn verb a voweless nûn at the end of a syllable followed by a consonant will assimilate to that consonant as, for example, in the form w%rc@;yI (Prov 20:28). The form Nma therefore is a proto-Masoretic note, originally written over the line, to indicate that the nûn in the following form w%hrUc;n:yI was not assimilated. It was mistakenly copied into the text by a later copyist and now represents a gloss. This fascinating suggestion was picked up again by Felix Perles (Perles 1895: 20)⁷ and remained dormant until is was revived by Jacob Weingreen (Weingreen 1957: 160) in an article on Rabbinic-type glosses and again in his book on textual criticism (Weingreen 1982: 88-89). The only other modern scholar to notice and adopt this suggestion is Michael Fishbane, who included it in his article "Abbreviatons, Hebrew Texts" for the Supplementary Volume of The Interpreter's Dictionary of the Bible (Fishbane 1976: 4) and then as an example of scribal conservatism in his book Biblical Interpretation in Ancient Israel (Fishbane 1985: 63-64).

7. While the suggestion of a proto-Masoretic gloss is most appealing--especially since some glosses, acronyms, and abbreviations can be demonstrated in parts of the Hebrew Bible⁸--this suggestion has to overcome several reservations. The first has to do with the fact that this note would have had to have found its way into the text at a very early stage in the textual process. Because the Septuagint has preserved a rendering of Nma in its translation (as the interrogative ti/j "who?"), this note would have to go back before to the time of the translation of the Psalms into Greek, that is, to the middle of second century B.C.E., the most probable date for the translation of the Psalms into Greek (Swete 1914: 25). Weingreen (Weingreen 1982: 88-89) argues that this fact indicates the existence of a kind of Masoretic activity going back at least to the third or second century B.C.E. But we have no evidence whatsover that Masoretic activity was at work so early. It is true that the date of the Masoretic notes of the type with which we are familiar in later Hebrew manuscripts is not known, but there is no evidence even in the Talmudic period of the existence of such notes (Yeivin 1980: 135). The only evidence we have in Talmudic times in this regard is that there are discussions in parts of the Talmud which deal with the substance of what would be included in later Masoretic notes. The earliest we can date Masoretic notations is to the 6th century C.E., so the likelihood of such a pre-Masoretic notation going back to before the time of the Greek translation of the Bible is pretty slim.

8. Our second reservation has to do with the fact that this putative note assumes that the Masoretes or their predecessors (the proto-Masoretes) would have commented on the presence or absence of a nûn in initial nûn verbs. But from what we know of Masoretic conventions, this assumption is not borne out. The Masoretes did not comment on this phenomenon. In none of the other forms of the imperfect of initial nûn where the nûn is retained is there any Masoretic note commenting on the retention of the nûn. This is true for all forms of rcn wherever this phenomenon occurs (at Deut 33:9; Pss 78:7; 105:45; 140:2, 5; Prov 5:2), as well as for other initial nûn verbs such as #$gn, (Isa 58:3), Pdn (Ps 68:3), and Pqn (Isa 29:1).⁹ Indeed, on some of these forms the Masoretes make different observations not related to the presence or absence of a nûn. For example, there is a Masoretic note on the form hk@frEc;n:ti in Prov 2:11, but the note concerns the hê at the end, not the nûn, since this form is one of twenty written with the letter hê at the end of the word (Weil 1971: §964).

9. Our third reservation has to do with the fact that the Masoretes themselves seem to be unaware that the form Nma represented an original Masoretic type note. In fact there is a Masoretic note on Nma which reads #dq Nw#l lO "this form occurs only once in Biblical Hebrew." The unstated corollary is that this form does occur in Biblical Aramaic, as, for example, at Ezra 5:3, 4, 9. So we see that, by the time of this Masorah parva note, any tradition that the word Nma was itself a note had been lost to the Masoretes.

10. Our fourth reservation that Nma represents a pre-Masoretic note has to do with the form of the alleged note itself. The proposed suggestion is that Nma stands for Nw%n )l'mf, but the note )l'mf nearly always indicates that a vowel letter, usually waw or yôd, but occasionally hê¹⁰ or alep,¹¹ is written as a mater lectionis. The actual vowel which is the subject of the note, and provides the impetus for the note, is usually not written in the note but is assumed. Thus the note does not write the vowel waw, yôd, hê, or alep, only the number of times the form occurs followed by the abbreviation for )l'mf (lOm or mO),¹² such as lOm gO "three times plene." The term )l'mf is never used to indicate the presence of a consonant such as a nûn. When a note is necessary to indicate the presence or absence of a nûn this is done in the Masorah parva via a qere note. Thus the Masorah has a qere note at 2 Sam 21:6 to indicate that the qere Nt@ayU is preferred to the ketiv reading with a nûn of Ntny. At 1 Kgs 17:14 there is a qere note stating the preference of tt@' to the ketiv form with nûn Ntt. Similarly, in cases illustrating the reverse, where the Masoretes indicate that a nûn should be read in a word, the qere reading is employed. For example, there is a qere note at Prov 3:15 to indicate that the form Myynpm should be read with another nûn as MynIynIp@;mi. At 2 Sam 21:4 there is a qere note to indicate that the form yl should be read with a nûn as w%nlf. All examples of notes remarking on the existence or omission of the letter nûn are formulated as ketiv and qere notes. In none of these notes is there any formulation of the type Nw%n )l'mf, and I have not been able to find such a Masoretic note in any of the Masorah parva notes to the Leningrad Codex.

11. To sum up, the form Nma in the prayer for the king in Psalm 61:8 is an old crux interpretum. It has traditionally been taken as a piel apocopated imperative from the verb hnm meaning "appoint!" The ancient versions took it as akin to the Aramaic interrogative Nma and translated it as "who?" Because it interrupts the natural grammatical and poetic sense of the verse, many suggestions have been made either to emend it or omit it as a gloss. One tantalizing suggestion, revived in the last fifty years by Weingreen and Fishbane, is that Nma represents a proto-Masoretic gloss standing for Nw%n )l'mf to indicate that the following form w%hrUc;n:yI is written with the letter nûn. We raised a number of objections to this suggestion. The first is that the earliest Masoretic notations can be dated only to the 6th century C.E., so the likelihood of such a pre-Masoretic notation going back to before the time of the Greek translation of the Bible is most unlikely. The second is that the Masoretes in their Masorah parva notes do not comment on the presence or absence of a nûn in initial nûn verbs. The third objection is that in our extant Masorahs, not only are the Masoretes themselves unaware that the form Nma represented an original Masoretic type note but they themselves apply a different note to Nma. The final objection is that the form of this putative note does not conform to any other similar Masoretic note, since the term )l'mf is never used to indicate the presence of a consonant but only of a vowel; when the Masorah wishes to indicate a superfluous nûn, a qere notation is used instead. Thus if Nma represents a Masoretic note it would be the first example of its type.

© TC: A Journal of Biblical Textual Criticism, 2003.

Endnotes

¹See, for example, Pss 21:5; 25:10; 40:12; 57:4; 89:15; Prov 3:2-3; 9:11 etc.

²The Vaticanus ms has a variant adding the genitive pronoun au0tw~n "of them," as it reads "which (ti/j) of them (au0tw~n) will seek out his mercy and truth?"; see Rahlfs 1979: II, 62.

³The Targum to the Psalms may be found in any Rabbinic Bible (twOlwOdg%: twO)rFq:mi)

⁴This reading would represent a haplography of the last letter of the preceding word tme)vwE and a supposed interchange of nûn and dalet.

⁵Other suggestions include that of Mitchell Dahood (Dahood 1968: 88), that Nma is a precative perfect qal passive, "may kindness and fidelity be appointed to safeguard him."

⁶This note is also found in the first German edition (Fürst 1857-1861: 753). However, in his earlier work (Fürst 1842: 314), Fürst offered the usual classification of Nma as a piel imperative from hnm.

⁷Perles believed that the suggestion that Nma was a Masoretic note was originally made by Gesenius, but I have been unable to verify this. Every edition of Gesenius' Lexicon that I consulted starting from 1810-1812 lists Nma as the piel of hnm. It is possible that Perles confused Gesenius with Fürst, since both were the authors of Hebrew and Aramaic lexicons.

⁸Fishbane (Fishbane 1976: 3-4), but see the dissenting opinion by Emanuel Tov (Tov 1994: 64-65).

⁹The lengthened nûn on these verbs may be explained by pause (GKC, §§66f; Joüon and Muraoka 1991, §72b). The theory that these forms represent examples of a yaqattal tense in Biblical Hebrew has not won widespread acceptance (Waltke and O'Connor 1990, §29.4d, n. 67).

¹⁰Examples of )l'mf referring to a hê are Num 14:19 (htf)#&fnF); Ps 8:4 (ht@fn:nFwOk@); and Ps 30:8 (ht@fd:ma(Vhe).

¹¹Examples of )l'mf referring to an alep are 2 Sam 12:1 (#$)rF; Prov 24:7 (twOm)rF); and Neh 7:71 ()wOb@rI).

¹²I have only been able to find one case in the Masorah of the Leningrad Codex where a vowel was actually named alongside the note )l'mf. It occurs on the word sroyq' in Neh 7:47. The note reads dwyb lOm lO, meaning that this is the only occurrence of this form and it is written with a plene yôd. Since there are two vowels in the word sroyq' (a serê and a holem) which could take plene vowels, the word yôd may have been written to clarify to the reader which vowel was the subject of the discussion.