Translating the Psalms
O sing to the Lord a new song. . . .
—Psalm 96:1 (98:1,149:1)
Some of the fundamental texts of Chinese antiquity are so ambiguous, a scholar once told me, that not only what they mean is in dispute, but even what they are about. When I heard that, I felt sorry for the Chinese and the Sinologists. It’s all very well to say that if ambiguity makes you uneasy you’re an authoritarian personality, but surely such texts are too much of a good thing.
Of course, neither is ambiguity completely absent from the more Occidental languages and literatures. When I was young, one of my Latin teachers used to amuse himself by giving his students a sentence of five short words to translate: Mea mater sus est mala. (Punctuation was unknown to the ancients.) It can mean: “My mother, O pig, is wicked.” It can also mean: “Priestess, my sow is ugly”; or, “Come on, goddess, a boar is eating the apples” or peaches, or pomegranates, or bad things. And so on. But that sentence is exceptional. If a Latin text is hard, that’s seldom because it’s more than normally ambiguous.
Similarly with Hebrew. Spinoza didn’t doubt he knew what the Bible was saying, since he didn’t doubt he knew Hebrew. And that wasn’t only because our printed Bibles are “vocalized” or “pointed,” i.e., have vowel signs. Arthur Koestler said he could never learn to read unpointed Hebrew, and complained about the imprecision and ambiguity of words presenting themselves to a reader in the bare bones of their consonants. But we don’t normally read isolated words, we read sentences, or words in a context. Without a sentence or a context, vowels don’t always help: taken by itself, what exactly does sound mean? On the other hand, if you’re in a car heading from New York to New Jersey and you pass a marker with the legend G WSHNGTN BRDG, you have no right to grumble about ambiguity. Spinoza understood not only the vocalized Bible, but also the unvocalized books of the Rabbis and Maimonides.
If he was confident he knew what his Hebrew (and Latin) authors meant, we should be more confident. A century after him, Rousseau was still able to define etymology as a science in which vowels count for nothing and consonants for very little. Then Indo-European and Semitic historical linguistics hardly existed, now they are firmly established. We have just finished this year’s reading of Genesis in the synagogue. With the help of the late E. A. Speiser’s translation and notes, in the Anchor Bible, we could understand Genesis better than the men of an earlier day. Above all, Speiser was a linguist.
Reading the first fifty psalms with the help of Mitchell Dahood’s translation and notes, also in the Anchor Bible,1 what do I understand? I’m not sure. The only thing I’m sure of is that Father Dahood’s volume is different in kind from Speiser’s. Someone who has no Hebrew can be given the Anchor Genesis with the assurance that it will bring him close to the original, and to the world of things and thoughts in which that came into being. You would be doing such a person a disservice if you gave him Dahood’s Psalms 1. Even if he has Hebrew, the book may irritate him beyond bearing. I am myself an equable sort, but reading Dahood, though it pleased me sometimes, exasperated me often. That is less his fault than the editors’. His work simply doesn’t belong in the Anchor Bible, with its implicit promise to the common reader. It is a kind of demonstration project.
Dahood, a Jesuit, teaches Ugaritic literature at the Pontifical Biblical Institute. (Ugaritic is a language first uncovered about forty years ago in Syria. Most scholars think it is a form of Canaanite, and therefore closely related to Hebrew. The Ugaritic texts are several hundred years older than the oldest parts of the Bible.) Together, Dahood’s introduction, translation, and notes make up a sort of manifesto. He is saying, in effect: “The day of the Bible scholar not expert in Ugaritic is past, particularly if he specializes in such books as Psalms and Job. Only Ugaritic can help with those books, because even the Septuagint. the most ancient version, no longer understands their Canaanite archaisms—of vocabulary, grammar, style, imagery, and the rest. Of course, the Masoretes don’t understand either, since they’re later still. Hence we may ignore what the Masoretes added to the text or imposed on it: vowels, punctuation, verse division, sometimes word division. The psalms were always in liturgical use, so that the consonantal text the Masoretes received and transmitted is trustworthy.” In Dahood’s words: “[In] the first fifty psalms . . . resort to emendation can be justified in fewer than a half a dozen instances.”
Most of this is seriously meant, but it can’t all be. For instance, when it suits Dahood’s purpose he is ready enough to follow the Septuagint’s example, and even—because he’s a Catholic priest?—the Vulgate’s. (Most unscientific, that. The Vulgate is late and has little independent value as a witness.) Between the lines Dahood seems to be saying: “Ugaritic gives me a bag of tricks. With Houdini’s tricks, he could have himself shackled and immured, and yet escape. With mine, I can hold myself to the transmitted consonants and yet make of them pretty well what I choose, by adducing Ugaritic precedents. I don’t really believe all the things I say here. I say them only to make you see the virtuoso performances that Ugaritic makes possible, and the stodginess you’re condemned to without it.”
At one point Dahood becomes aware that this is what he sounds like, and he hurries to deny it’s what he means. He has told us things like these: that the suffix -i, first person in Hebrew, can be understood as third person, if convenient, in the light of “Phoenician and probably Ugaritic”; that similarly a troublesome l-, “to,” can be taken as an Ugaritic vocative or emphatic particle (“O” or “indeed, surely”); and that a difficult l?’; (l’;), ‘not,” can be read as *l?’ (since it, too, is l’;), “Victor, Victorious One.” He assures us that while “this . . . may sound much like ‘you name it, we have it,’” it isn’t. Really? I find it hard to believe that so many prepositions have to mean “from” quite so often—not only the one we always knew meant “from,” but also those we thought meant “to” and “in” and “on.” When there isn’t much perceptible difference between in” and “from,” why insist that thunder bashamayim, “in the sky,” should be “from the sky”?
An example of the lengths Dahood will go to in order to avoid facing the need for emendation—or, alternatively, of the joy he has in showing off his bag of tricks: If only because in the bad old days Ps. 2: 11-12 were a favorite proof-text for Christian theologians trying to convince Jews that the Hebrew Bible foretold Jesus the Christ, those two verses are well-known for their difficulty. In the Masoretic Text the end of the first and the beginning of the second are wegilu bire’adah. nash-sheqü-bhar. King James: “. . . and rejoice with trembling. Kiss the Son . . .”; Jewish Publication Society: “. . . and rejoice with trembling. Do homage in purity . . .”; Revised Standard Version: “. . . with trembling kiss his feet . . .” (this last annotated as “conjectural. The Hebrew of 11b and 12a is uncertain”). RSV’s conjecture follows the recommendation of the critical apparatus in the third edition of Biblia Hebraica. The consonants that have caused all that trouble are wgylw br’dh nšqw br. With a minimum of rearrangement and the loss of only one w, Biblia Hebraica tells us to add the first word to the fourth, reading br’dh nšqw brglyw, i.e., bire’adah nashsheqü beraglaw—“with trembling kiss His [sc, the LORD’S] feet”: neat, economical, and satisfactory.
Dahood goes all around Robin Hood’s barn: “. . . and live in trembling, O mortal men. . . .” He keeps the consonants, but divides the last two words in his own way—nš qbr. (He says this is “no consonantal change,” but note that he has dropped the w after q. If challenged, he would probably say that the w was a mistaken semi-vowel introduced into the text by a scribe trying to make the spelling a little fuller. Canaanite texts are characterized by scriptio defectiva, skimpy spelling.) Either because he is constrained by the new meanings he has erected, or just for the fun of it, Dahood gives “live” for gilu instead of “rejoice.” He has to twist and turn for that, and I don’t think it comes off; but no matter. The real innovation is “O mortal men”: nš qbr, vocalized neshe qebher, “men of the grave.” Neshe, “men of”? It means “women of”; “men of” is anshe. Quite all right—“. . . nashim, ‘men’ [and therefore neshe, ‘man of’], is well-documented in Ugaritic. . . .” (He gives only one example.) Shouldn’t the consonants of neshe be nšy, not nš;? Blame the skimpy spelling. The Bible offers us compounds, Dahood says, in which the first element is ish, “man of,” or bene, “sons of,” and the second element is mawet or temutah, “death.” “Men of the grave” is of a piece with such constructions.
The fact remains that neither Ugaritic nor Hebrew has a straightforward instance of neshe qebher—or ish/anshe qebher, or ben / bene qebher. If RSV calls “kiss his feet” conjectural, what shall we call Dahood’s “O mortal men”? He doesn’t give us “O mortal men” in a discursive note, as a possible alternative to a standard translation. It is his translation. And then, demolishing his own elaborate structure, he notes: “The recent proposal of Henri Cazelles . . . in Oriens Antiquus 3 (1964), 43-45, who translates ‘Saluez le Brillant’; [Pay your respects to the Shining One], on the basis of Ugar. brr, ‘to shine,’ is possible though it requires further confirmation.” Yet, determined to immortalize his mortal men, Dahood won’t change his translation. One doesn’t usually expect such farce in work of this kind. It is almost as delightful as his saying, in a note to Ps. 36, that another scholar’s emendation “is too clever by half.”
It’s unsettling that Dahood can’t even be relied upon not to be directly helpful. Every now and then one of his suggestions is so elegant that it compels admiring assent. Close by, a like suggestion wobbles and misses, because there it is complicated and ungainly.
Take denominative verbs. Dahood is fond of them. Ugaritic, he says, liked to form verbs from nouns, especially parts of the body, and from numerals; and such verbs, until now mostly unrecognized, abound in the text of Psalms. In 22:26 (25) he is fine. King James, JPS, and RSV all translate, more or less, “From Thee comes my praise [i.e., the praise of me] in the great congregation,” while Dahood says, “One hundred times will I repeat to you my song of praise [i.e., the song of praise by me] in the great congregation.” Let us dispose of secondary matters first: tehillati is equally “my praise” and “my song of praise.” (The Hebrew for “Psalms” is tehillim.) The problem is in the first part of the verse. KJV, JPS, and RSV straightforwardly render me’ittekha (m’tk), “from thee (is),” as “from Thee comes.” For Dahood those consonants clothe a verb based on me’ah (m’h), “hundred.” He reads m’tk as m?’/?/?/?/??2=m?’/?/?/?/, a pi’el verb, + -??, here dative: “I have repeated (repeat, shall repeat) to you a hundred times.” Why is it spelled m’tk and not, as we should expect, m’ytyk (or m’tyk)? Scriptio defectiva again.
Here the denominative verb is a success, in 20: 8 (7) a failure. KJV: “Some trust in chariots, and some in horses; but we will remember the name of the LORD our God.” JPS: “Some trust in chariots, and some in horses; but we will make mention of the name of the LORD our God.” RSV: “Some boast of chariots, and some of horses; but we boast of the name of the LORD our God.” (I prefer “Some invoke chariots, and others horses; but we the name, etc.”) There is only one verb in the Hebrew text—in the second clause, where “we” is its subject. In the first clause it is implied, with “some” and “others” as subjects. The verb is nazkir, from the root zkr, “remember”—in this form, “cause to be remembered”; and all the translations cited so far understand it in that sense. Not so Dahood, who renders: “Some through chariots, and others through horses, but we through the Name of our God are strong.” Nazkir, he thinks, is a denominative verb from the zakhar (zkr) that means “male.” Just as elsewhere nagbir is from gebher or gibbor (gbr), and means “we shall be powerful,” so here nazkir means “we shall be (are) male, i.e., masculine, i.e., strong.”
Dahood is wrong. Zakhar means only “male.” It refers to what is common to rams and infant boys, distinguishing them from ewes and infant girls. It is biological only—one might say, anatomical; whereas gbr is social. Gbr and its derivatives, in the sense of “man, manliness, power,” are frequent in Psalms, but zakhar, “male,” is not to be found there at all. In the whole Bible no form of hizkir suggests the meaning “to be male,” let alone “to be strong”; and every locution similar to our beshem . . . nazkir (see especially Exodus 23: 13, Joshua 23: 7, and Isaiah 48: 1) is related to “cause to be remembered, mention.”
What led Dahood astray? The critical apparatus of Biblia Hebraica, noting the Septuagint’s “we shall be made great,” records that some would therefore emend nazkir to nagbir. Dahood wants to accomplish the same thing without emending, so he shows off—in a rather unfortunate way, because we have to ask ourselves whether he is so lacking in linguistic finesse as to confuse the different connotations of gbr and zkr. But it is as a linguist that he addresses us.
Generally, the verb in biblical poetry, always difficult, with Dahood enters the domain of total arbitrariness. Whatever he prefers, he can justify: perfect imperfects, precative perfects, -ah as Canaanite-influenced 3rd-person perfect masculine singular, -u as Canaanite-influenced 2nd-person imperfect masculine singular. That is true also of his infixed -t-‘s, enclitic -m ‘s, intensive b-‘s, and inserted ki’s. Not that such things may not exist, but that he makes entirely too free with them. My most common marginal note is the exclamation point, meaning “Everything can be anything!” On a higher turn of the spiral, the linguist Dahood seems to have put us back in the pre-linguistic conditions Rousseau laughed at.
A final comment on these matters. Of the Forty-first Psalm Dahood says: “The language of this lament is very archaic (probably Davidic era) and difficult, so that the translation of several clauses is doubtful.” The Davidic era—that is, about 1000-950 BCE. Most of the psalms, therefore, are in his view later; but not nearly so late as was thought two generations ago, when the critics spoke confidently of Maccabean psalms. From this and other things Dahood says, we may infer that he thinks most of the psalms to be no more than 350 to 400 years younger than this archaic one. That is to say, he perceives a marked difference in language arising in the course of four centuries. But he also says that the Ugaritic texts were composed or first written down about 1800 BCE, or earlier. They are at least eight centuries older than the oldest psalms. How, then, does he allow himself to understand the language of the psalms as if it were contemporary with his Ugaritic? Ugaritic has something to tell us, but what, precisely? Surely something a little less unequivocal than Dahood appears to believe.
An analogy. That German Knabe and Knecht are cognate with English knave and knight is certain, but we would be making a mistake if we thought we could explain one set by the other. Again, magistrate is from magis, “more,” and minister from minus, “less,” but it doesn’t follow that a police-court magistrate is greater than a minister of state. It is taking much for granted to assume that what a word probably meant in Ugaritic in 1800 BCE it certainly meant in Hebrew in 800 BCE. Because ‘pr is “mud” in Ugaritic—so Dahood tells us—that is how he translates Hebrew ‘aphar. But the biblical evidence is for “dry earth, dust,” as we have always supposed. It is good for us to know what the meaning was a thousand years earlier. It is always good for us to know things. Only, then we have the job of determining how to understand our knowledge, and what to do with it. Rarely will we be well-advised to perform tricks with it.
What dahood thinks of the large questions about the psalms—types, functions, origins, theology—he doesn’t tell us here, but promises us in his concluding, second volume. Or rather, here he neither considers them at the length they require nor refrains from considering them at all, but touches on them lightly, which is to say, unsatisfactorily. Some psalms he characterizes as an individual’s laments or prayers for deliverance. Then he translates certain words, novelly, as “eternal life” and “resurrection.” Why should a lament or a prayer for deliverance include pleas for such boons? What is being prayed for, right now, is surcease here below. And it won’t do for Dahood to throw out a remark about Canaanite belief in an afterlife, like the Greeks’ Hades and Elysian Fields, and deduce that the Israelites had such a belief, too. The subject needs more than a hit-and-run treatment. Anyway, if that really is an individual’s prayer for deliverance, then the sickness or evil that has him in thrall can only be paranoia. All those foes, enemies, and evildoers who compass him about, digging pits and laying snares for him! The “I” of those prayers must be a special kind of individual—an exemplary one, probably a royal one, who in his person and life enacts conflicts of a national or even cosmic bearing.
I have said that Dahood is conservative about the (consonantal) text. But textual conservatism doesn’t necessarily imply conservatism about the meaning of the text, or the theology of its authors. The Scandinavian myth-and-ritual school, for instance, is reluctant to emend, but ready to find paganism everywhere. A representative work, by a disciple of the masters of that school, is G. W. Ahlström’s Psalm 89: Eine Liturgie aus dem Ritual des leidenden Königs, Lund 1959. Conservative about the text, Ahlstr?m is nevertheless sure he is dealing with a ritual of the suffering king, in the cult of a YHWH who is a dying and rising god like Tammuz-Adonis (adon, “Lord”). Ahlstrom is even suspicious of the name David. It means “beloved.” Is it really a name? Perhaps it is the appellation of an officiant? Tammuz is called “beloved.”
Dahood doesn’t go so far—probably. Of the Twenty-ninth Psalm he says that it is a “Yahwistic adaptation of an older Canaanite hymn to the storm-god, Baal,” reminding us that more than thirty years ago H. L. Ginsberg called it “a Phoenician hymn in the Psalter.” That the Israelites took it over from the Canaanites seems hard to dispute. But if so, why does Dahood translate the first verse as “. . . Give Yahweh, O gods, give Yahweh glory and praise . . .”? RSV: “. . . Ascribe to the LORD, O heavenly beings, ascribe to the LORD glory and strength. . . .” (Note to “heavenly beings”: “Heb sons of gods. ”) JPS: “Ascribe unto the LORD, O ye sons of might, ascribe unto the LORD glory and strength.” (Note in the Soncino Psalms: “sons of might. Hebrew ‘sons of the mighty’ or ‘sons of God,’ a double plural of benel. The same phrase occurs in 89: 7, where the parallelism shows that the angelic hosts are intended. . . .”) This isn’t quite the purest or most austere monotheism, but surely the Israelites wouldn’t have left the psalm out-and-out polytheistic? In a note Dahood says: “In Canaanite mythology the bn ilm, ‘the sons of El,’ . . . are the minor gods . . . of the pantheon. . . . In the Old Testament the term was demythologized and came to refer to the angels or spiritual beings who are members of Yahweh’s court and do his bidding. . . .” Then why not translate, with RSV, “heavenly beings”? Why “gods”?
Above all, why “gods” in the Twenty-ninth Psalm? Father Dahood may not know, but that continues to hold a special place in the Jewish liturgy. It is a Sabbath psalm par excellence, sung Friday night and again Saturday morning, when the Torah scroll is returned to the Ark. (Compare the letter by Rabbi Simeon J. Maslin here,3 commenting on Novinsky and Paulo’s “The Last Marranos,” May 1967, which showed that to this day the Portuguese Marranos recite the same psalm—without knowing what the Hebrew means, of course, and therefore corrupting the text in their oral transmission of it.) In all these many centuries, our pious and learned have been invoking angels—celestial beings—not gods. It must have been always so, from the time when Jews (or Israelites) first began the liturgical use of this psalm.
Another point about liturgy is suggested by the last verse of the Twentieth Psalm, frequent in Jewish prayer: YHWH hoshi’ah hammelekh ya’anenu bheyom qor’enu. JPS follows the punctuation of the Masoretes: “Save, LORD; let the King answer us in the day that we call.” RSV, like the ancient versions, divides the verse differently, and probably better: “Give victory to the king, O LORD; answer us when we call.” (Again Dahood is ostentatiously fancy.) A Jew who tries to learn from the Bible scholars may agree that RSV is right, but when he leads the congregation in prayer it will not occur to him to shift the pause from hoshi’ah to hammelekh.
In short, what is the text that is being translated? Is it one or many? Fixed or labile? If you translate what the text was “originally,” aren’t you ignoring what it is “really”? Haven’t we the right to say to Dahood that a hymn with an invocation to gods wasn’t yet an Israelite psalm; and that since he’s translating Israelite-Jewish psalms rather than Canaanite hymns, only “heavenly beings” will do? Going to the very origin, Dahood may mistranslate. Yet the ideal of translation, generally, is to render a text as nearly as possible as it was understood by the writer (or “writer”) and his first hearers. Which is only another reason why the enterprise of translation, so necessary and useful, is doomed from the start. E. H. Gombrich says, in Norm and Form; “. . . a work of art carries with it the barnacles of its voyage through the centuries . . . anything we say or write about a painting may change it in a subtle way.” The text won’t sit still for you.
One moral of all this is Utopian, but inescapable. It’s a Jew’s business to know Hebrew. Reading a psalm, or intoning one, he should have simultaneously present to his mind as many different meanings as possible. He can’t depend on a translation, or translations.
What is close to you, or should be close to you, you should know. If you know what’s remote and not what’s close, you’re Levantine—like the upper class in the Egypt of the pashas and effendis, whose beaux-esprits, while knowing little of Arabic or the Arabs, might write delicate little essays about Proust. (So I have seen them described.) Greek is “objectively” a more important language to know than Latin, for literature, philosophy, history of science, or even rabbinics or New Testament; but Latin is the language of the West, and when only one of the two is taught in the West, it will rightly be Latin. If we may trust T. S. Eliot and Edmund Wilson, Dante is greater than Shakespeare; yet an Englishman or American has to know Shakespeare, and only ought to know Dante. A German without Goethe is uncultured, an Englishman or American with Goethe is a curiosity. And so with Jews and Hebrew.
Maybe this is the meaning for us of the old Jewish sigh that it’s hard—or, as someone said about something else, it’s a complex fate—to be a Jew. You are that lady’s son the nuclear geophysicist, and you have to be able to weigh one translation of Psalms against another.
1 Volume 16 of the Anchor Bible, Doubleday, 1966, xliii + 329 pp., $6.00.
2 His M/?/'ēṯ/?/ḵā must be a typographical error.
3 Letters from Readers, October 1967.