TODAY, WE’RE AIMING HIGH, attempting to answer the question of how the Hebrew title “Adonai” was translated into Greek. It’s a bit tricky and nuanced, but I believe doable, now with the help of my conversation with Albert Pietersma. As a micro-recap, we have seen recently that, generally speaking:
- Yahweh has hugely-greater frequency than Adonai, which is particularly rare in the Pentateuch.
- Adonai leans a lot on Yahweh when it occurs, the two usually appearing together in a “combo” format.
- There are Jewish traditions, manuscript retranscription hesitations and errors that point to semantic proximity and even overlap of the two terms, Adonai and Yahweh.
For further reference please see the bottom of today's post for where to look.
So as I reach into the theological fridge for my Greek yoghurt to share the news about how the Adonai translation went, I wanted to also let you know that my email to Albert Pietersma received a short but meaningful response. I basically wrote to him an edited form of my 2016 post: Why This Research Matters, in which I highlight two important differences between Yahweh and Adonai translations into Greek within the book of Psalms: Yahweh still receives special treatment grammatically with fewer articles than expected, in line with the tradition of the Pentateuch. Furthermore, on one sample lexical unit, ἐπὶ τὸν κύριον, I seemed to have a clear point, with only Adonai receiving the middle word in the Greek there, “τὸν”, or “the Lord”. At the time of writing the email I didn’t know it yet, but a third differentiating factor arises for Psalms. But surely, even with these two differentiating factors alone, Pietersma was going to have to recognise that this Mr. Nobody here had stumbled over a wrong-headed assumption and perhaps now reconsider how his NETS translation handled “the Lord”? I'd still like them to do that by the way. Here's what Pietersma said and how combined with the results I am going to share with you today, there is good cause for me to rethink my overall picture of Adonai translation:
AP: I wonder if our respective foci aren’t at odds. My point here is simply that the Hebrew divine names Yahweh, Yah, and Adonai are all rendered by Greek kyrios. Differently put, though the translator might have opted for transcriptions like Yao, Ya, or Adonai, by opting for kyrios he foregoes differentiation at the word level. Period! Full stop! Needless to say, differentiation at the grammatical remains an option, but that is a different issue. Yes?
First of all: fantastic to be able to get direct insights from specialists like Pietersma! Second, yes the Greek translators could quite easily have avoided any differentiation issues by transcribing the words differently. If they'd really wanted to incorporate a Kyrios translation for one of them, they could have left the other untouched by transliterating it into the new alphabet. Actually, I should probably mention that transliteration does very occasionally happen for both:
- · For Yahweh, toward the end of Psalms, you may notice this standalone line often repeated at the end of these smaller Psalms: Praise the LORD! In the Hebrew you have two words for this, הַֽלְלוּ־יָֽהּ - “Hallal” and “Yah” (as in Yahweh), which in Greek combine to form the single: αλληλουια, literally, Allelujah!
- · For Adonai, on one of the Old Greek manuscripts passed down to us in Judges 16:28, it reads Αδωναιε κύριε for the NASB of: “Then Samson called to the LORD and said, “O Lord GOD...”. Here this translator in this manuscript does something I have seen hardly anywhere else[i], transliterating the Hebrew Adonai into Greek Αδωναιε.
However, exceptions prove the rule sometimes, and that is the case here - thank you Dr. Pietersma for reminding me of this glaring option.
So, the translators chose, collectively and more or less unanimously, to use the same Greek word κύριε to translate both of these different Hebrew names for the Israelite Deity, Adonai and Yahweh.
Yet Pietersma seems to acknowledge the possibility of grammatical differentiation. Yet, how significant should the grammatical issue that I raise be and how on Earth would you represent that in an English translation like NETS?
Well, when producing a massive translation project such as any Bible you might have at home, consult in a seminary or look up online, you are benefitting from the fruit of ridiculous numbers of hours of collaboration to maintain consistency of practice across the whole thing. So if striving to produce a single-volume translation where grammatical differences occur consistently throughout the texts, then you might want to consider a strict translation scrapping the English article for LORD (Yahweh) and maintaining it for the Lord (Adonai). To most readers, I think this could seem confusing or like a curious typo. But hang on a second, is there a consistency of practice across the whole thing? Isn’t one of the point of NETS to draw out the particularities of each book? To answer these questions we need to see what actually happens to the articles in the Greek translation of Adonai in nominative and genitive cases across the Septuagint:
Adonai into nominative and genitive Kyrios
LXX nominative + genitive arthrous (HUMAN LORD FIGURE)
|
LXX nominative + genitive anarthrous (HUMAN LORD FIGURE)
|
Adown - human lord - "arthricity"
|
LXX nominative + genitive arthrous (GOD)
|
LXX nominative + genitive anarthrous (GOD)
|
Adonai - GOD - "arthricity"
|
|
In total
|
107
|
6
|
94.7%
|
15
|
327
|
4.4%
|
Genesis
|
25
|
1
|
96.2%
|
0
|
0
|
n/a
|
Exodus
|
1
|
0
|
100.0%
|
1
|
2
|
33.3%
|
Numbers
|
2
|
0
|
100.0%
|
0
|
0
|
n/a
|
Deuteronomy
|
1
|
0
|
100.0%
|
0
|
1
|
0.0%
|
Joshua
|
0
|
0
|
n/a
|
0
|
2
|
0.0%
|
Judges
|
2
|
0
|
100.0%
|
0
|
0
|
n/a
|
Ruth - Job
|
72
|
4
|
94.7%
|
0
|
6
|
0.0%
|
Psalms & Proverbs
|
1
|
1
|
50.0%
|
11
|
12
|
47.8%
|
Isaiah
|
2
|
0
|
100.0%
|
1
|
42
|
2.3%
|
Jeremiah
|
0
|
0
|
n/a
|
0
|
6
|
0.0%
|
Lamentations
|
0
|
0
|
n/a
|
1
|
11
|
8.3%
|
Ezekiel
|
0
|
0
|
n/a
|
1
|
212
|
0.5%
|
Daniel-Malachi
|
1
|
0
|
100.0%
|
0
|
33
|
0.0%
|
Result: Like we expect of Yahweh, Adonai (and Adown) applied to God receives the same treatment as Yahweh, not many articles at all, with the exception of Psalms. When a human referent is presented, the presence of articles is radically higher. Rather than representing a Bible-wide tendency of article-adding to Adonai translation, Psalms is actually the outlier by far! Amazing how quickly our expectations can skew our analysis.
It might seem puzzling to see books missing in the table above or even fewer counts than mentioned in my recent posts discussing relatively low occurrence rates of Adonai. The reason for this is to do with how Greek articles can be seen to function even with proper names in certain cases. It has been clear to me for almost two years now since I started thinking about the Greek “anarthrous” translation of Yahweh that this anarthrous rule is only really particularly discernible in the nominative and genitive Greek cases. Of the remaining three, accusative, dative and vocative, “Lord” in the accusative (κύριον) and dative (κυρίῳ) needs lots of articles, while in the vocative (κύριε) it virtually never requires an article. So I hope that helps explain why we are looking at even fewer instances of Adonai and why in the table above many books are not represented or have been grouped together - they may literally have zero occurrences of a translation of Adonai into the nominative (κύριος) or genitive (κυρίου) cases.
But where we do have data, apart from the Psalms, we can deduce that Adonai seems to have been “swept along” with Yahweh's anarthrous rule, with very low rates of what I am calling “arthricity”, i.e. articles relative to what we might expect on average for a normal title.
The Pentateuch Legacy
Before we move to examples of genitive or nominative translation beyond the Pentateuch, let's illustrate this similarity of Adonai and Yahweh translation by going back to some of the vocative texts. These would have been among the very first to be translated into Greek in the third-century BCE Septuagint project targeting the first five books, including significant passages from Exodus and Deuteronomy:
Exodus 15:17 (NASB):
“You will bring them and plant them in the mountain of Your inheritance,
The place, O LORD [Yahweh => κύριε], which You have made for Your dwelling,
The sanctuary, O Lord [Adonai => κύριε], which Your hands have established.
Deuteronomy 3:24 (NASB)
‘O Lord GOD [Adonai Yahweh => κύριε κύριε], You have begun to show Your servant Your greatness and Your strong hand; for what god is there in heaven or on earth who can do such works and mighty acts as Yours?
- In Exodus 15:17 we have simple parallelism with undifferentiated κύριε.
- In both Deuteronomy 3:24 and 9:26 we have an Adonai-Yahweh “combo“ reflected in a simple repetition of “κύριε κύριε”.
We know that Yahweh is a super-common term in the Pentateuch (see Adonai vs Yahweh - Two Charts for the One Lord and BlueLetterBible search here), especially once Moses receives the revelation of Yahweh’s name in early Exodus. We can also assume that any term selected had to be meaningful but also practical in order to help Jews avoid recreating issues of tiptoeing around the Name all over again in the new host language. So Exodus and Deuteronomy simply substitute in this solution and aren't fussed by the repetition. Let's look now though at two combos from Genesis, which do seem to reflect an early concern about repetition:
Genesis 15:2 (NASB)
Abram said, “O Lord GOD [Hebrew: Adonai Yahweh => Greek: δέσποτα => NETS: “O Master”], what will You give me, since I am childless, and the heir of my house is Eliezer of Damascus?”
Genesis 15:8 (NASB)
“He said, “O Lord GOD [Hebrew: Adonai Yahweh => Greek: δέσποτα κύριε => NETS: “O Master, Lord”], how may I know that I will possess it?”
What is the nature of this apparent concern? As I said, we know that there is nothing super-sacred about saying “Lord” or “Master” in and of itself, as they are used throughout the Greek Bible and as these are terms also used for human rulers. But what is going on with Gen 15:2 throwing in this new word “δέσποτα” to translate both Yahweh and Adonai simultaneously and why is it different in 15:8? Actually, the Genesis translator in both instances achieves something that the translator of Deuteronomy was apparently not so concerned about: avoiding redundant repetition. What doesn't matter - if you are the Genesis translator - is quite how you go about that avoidance work. Heck, Adonai occurs so rarely (at this stage), it's not like you need to scratch your head for ages to come up with some complex rule, so it just doesn't matter if there is inconsistency between 15:2 and 15:8.
A final point from the Pentateuch: because Adonai can be used to say “my Lord”, conveyed in Greek “the Lord of me”, the article must be maintained here (see Exodus 34:9 “ὁ κύριός μου”). The translator team are simply not fussed that the non-possessable Yahweh will become indistinguishable from the possessable Adonai in their fused translation.
Thus we can summarise that the first translator team left a legacy for subsequent translators with the following approximate set of implicit instructions:
- Treat Adonai as you would Yahweh
- Try to avoid repeating κύριε when you can, but there are none of the theological stakes of the Hebrew originals
- Adonai possessives will need to follow usual Greek rules with the article
Nearly all the subsequent translators “got the memo”.
Subsequent Hebrew History strengthens the emerging Adonai-Yahweh synonymy
Initially, Joshua translator seems to be aware of the Genesis translation alternative of despota (Jos 5:14) and affirms the simple parallelism in 3:11 and 3:13 with genitive application: “the ark of the Lord [Adonai]” and “the ark of the LORD [Yahweh], the Lord [Adonai] of the whole Earth”. Previously, only two anarthrous genitives had preceded this, both of which were combos (Ex. 23:17 and 34:23). Like Exodus, he continues to avoid this redundancy, via condensing into a single κυρίου or κύριε (3:13 and 7:7).
Judges translator only has 4 Adonai to work with and continues the work of redundancy avoidance in 16:28 with the exceptional Αδωναιε κύριε (see above) and seems to experiment in 6:22 with a possessive to separate the two κύριέ: κύριέ μου κύριε, which will be picked up in later historical books. So Adonai is still rare at this stage and all four occur in the typical vocative – O Lord – sense.
Ruth: no Adonai to work with.
Samuel has a super-concentrated passage in 2 Sam 7:18-29 (which will certainly have interested source critics), providing all of a sudden six Adonai-Yahweh combos. This translator picks up on Judges translator’s κύριέ μου κύριε and slams it down for all six occurrences of the Adonai-Yahweh combos in 2 Sam 7. Perhaps we could speculate that Judges and Samuel translators’ (assuming they are not one and the same translator) double-emphasis of Yahweh's lordship is maintained and the inelegance of the strict repetition of κύριέ κύριε is avoided.
Kings appears to be from a different translating mind again although with the same legacy behind him. This translator doesn't mind the inelegance of repetition (1 Ki 8:53), and reasserts that Yahweh's stuff can be stated in exactly the same terms as Adonai's stuff, anarthrously. Our two examples here are “the ark of the covenant of Adonai” (1 Ki 3:15) and the anarthrous lexical unit “ἐνώπιον κυρίου” here applied to Adonai as with Yahweh: “in the sight of Adonai”. 1 Ki 22:6 and 2 Ki 7:6 introduce our first conspicuous nominative anarthrous for Adonai: κύριος (no “ὁ”).
Ezra 10:3 and its translation are not relevant to my survey as the Greek translator clearly understands a human referent and deviates from the Hebrew in any case.
Nehemiah is on redundancy alert in 10:29, appearing to accept some sacrifice of the initial proper name character of Yahweh (“observe all the commandments of Yahweh, our Lord” => “observe all the commandments of our Lord”).
Job's a bit weird - but it's interesting in the divine aspect of Adonai. In 28:28 we read the rather famous line: “Behold, the fear of the Lord (Adonai), that is wisdom”. In Greek this becomes something more like “God-worship, that is wisdom”.
Isaiah
Moving ahead into the prophets, the redundancy issue of what I am calling “Adonai-Yahweh combos” intensifies for the translators, and will reach its climax in Ezekiel. The Isaiah translator in 1:24 and 3:1 keeps to an already Greek-transliterated divine name “Sabaoth” (e.g. 1 Samuel 1:3, NETS: “Lord God Sabaoth”) after avoiding a repetition via δεσπότης instead of the usual condensed κύριος to translate Adonai alongside Yahweh. We have already seen this, of course, precedented back in Genesis. NETS shifts path slightly here, going for the rather peculiar “the Sovereign, the Lord Sabaoth, says”.
3:15 simply avoids translating Adonai-Yahweh completely - this avoidance tactic is evidenced elsewhere too. Generally, repetition avoidance is systematic in Isaiah. 3:17 applies an explicit divine swap similar to the Job instance while 3:18 introduces us to the first of many (35) anarthrous simple nominative translations for Adonai into κύριος, with only one articulated instance “ὁ κύριος” in 49:14. As opposed to my initial expectations (which were partly influenced by scholars like Wever's comments), Isaiah, if anything, is stronger on the anarthrous rule even than the Pentateuch itself! Genitive translations of Adonai are also consistently anarthrous, even when a Yahweh combo is not at stake (see 28:2, ὁ θυμὸς κυρίου, the wrath of [the] Lord).
Jeremiah and Lamentations
Jeremiah translator appears to be different to Isaiah and doesn't transliterate “Sabaoth” and isn't quite as systematic in combo redundancy avoidance (44:26 reads the double-anarthrous of “κύριος κύριος”), but is utterly faithful to article avoidance throughout the six nominative forms of κύριος (there are no genitives).
Lamentations is worth a separate mention in that its Adonai occurrences are all super-relevant to this survey, occurring separately from Yahweh (no combos) and nearly exclusively in the cases we are integrating (nominative and genitive). Result? Only 1 occurrence out of 12 relevant occurrences includes the article (1:15a): 8%. I particularly appreciate the existence of the Lamentations translation witness because it allows us to see Adonai treatment more independently from direct Yahweh translation influences of the many combos in the other Hebrew books in the Hebrew canon surrounding it, and this figure of 8.3% happens to remain very consistent with what we have learned to expect from Pentateuch-Yahweh-translation rates.
Ezekiel: thus says Adonai-Yahweh aka κύριος
Ezekiel is the big mama of any study on Adonai because of a single saying, repeated so often it's like a mantra. The full version in Hebrew it reads unwaveringly: אָמַר אֲדֹנָי יְהֹוִֽה, “thus says [the] Lord”. The Greek varies slightly between λέγει κύριος and λέγει κύριος κύριος. I suspect my count not to be perfect, but the following statistical overview should give a good idea:
- Of the 224 Adonai in Ezekiel, 213 occurrences are nominative or genitive and 212 are anarthrous
- There are 215 Adonai-Yahweh combos and the translator avoids repetition 166 times (77%) via condensation into a single κύριος
- λέγει κύριος also translates “says Yahweh” when no Adonai is present (e.g. 4:13)
Daniel and the Minor Prophets
The translations of these prophetic texts are entirely anarthrous when it comes to translating Adonai into the nominative (29 times) and the genitive (twice).
Daniel's few contributions to Adonai are mainly vocative “O Lord”, but there is one nominative κύριος in 1:2, rendered anarthrously. Hosea has a single nominative κύριος in 12:14, also rendered anarthrously. Next, Amos is a major Adonai user, if you remember the concentrations I reported, featuring the highest Adonai concentration in the entire Hebrew Bible. Its 24 occurrences mean that one word in every 126 is “Adonai”. Amos translator is pretty varied in his approach to the 21 Adonai-Yahweh combos: usually he'll tend to avoid the possibility of repeating “κύριος κύριος”, but not always (5:3, 7:2, 7:5, 9:5), and the avoidance tactics are twofold: either condensing into a single κύριος or replacing one of the κύριος with “ὁ θεός” (God, e.g. 8:9).
Obadiah in 1:1, Habakkuk in 3:19 and Zephaniah in 1:7 all solve their respective Adonai-Yahweh combos in line with Amos' “ὁ θεός” substitution method and Micah uses the condensing method for 1:2. Micah is also anarthrous when he translates a combo and a simple Adonai occurrence, all within his second verse!
Rounding off the Hebrew Bible, Malachi remains in perfect keeping, it would seem, with the sharing of the anarthrous rule to Adonai translations, with an anarthrous genitive κυρίου in 1:12 and an anarthrous nominative κύριος in 3:1.
Ok - erm, John, wait a minute, haven't you forgotten about the Psalms?!
Psalms: the Outlier
Don’t worry, I definitely have not forgotten the Psalms! As stated, Psalms appears in a different state in the current critical Greek text to the rest of the Septuagint with respect to its translation of Adonai now on at least three independent criteria:
1 1. Greek articles: Adonai to nominative Kyrios and genitive Kyriou translation “arthricity” rate of 50% relative to Yahweh rate of 18%.
2. ἐπὶ τὸν κύριον: it turns out that "ἐπὶ" before κύριον when translating "to Yahweh" systematically removes the accusative article τὸν: Ps 4:5, Ps 21:7, Ps 22:8, Ps 31:24, Ps 32:10 and 11, Ps 37:3, Ps 40:3, Ps 55:22. Not so in Psalm 130:6 translating Adonai.
3. Adonai-Yahweh redundancy indifference: Psalms translator makes by far the least effort of any of the translators to remove the redundancy of Adonai-Yahweh combos (only one time out of thirteen, in 30:8).
2. ἐπὶ τὸν κύριον: it turns out that "ἐπὶ" before κύριον when translating "to Yahweh" systematically removes the accusative article τὸν: Ps 4:5, Ps 21:7, Ps 22:8, Ps 31:24, Ps 32:10 and 11, Ps 37:3, Ps 40:3, Ps 55:22. Not so in Psalm 130:6 translating Adonai.
3. Adonai-Yahweh redundancy indifference: Psalms translator makes by far the least effort of any of the translators to remove the redundancy of Adonai-Yahweh combos (only one time out of thirteen, in 30:8).
What should NETS or other modern translations do with this information? First I need to think about what am I going to do with this information? As I already confessed, I used to think, erroneously, that Psalms was the example for the rest of the LXX. Now, on the above criteria, it is the clear outlier. But Pietersma handed me another important criterion: synonymy. Here, Psalms does much better, even falls into step with the rest of the LXX and should not be neglected.
Psalm 38:15: For I hope in You, O LORD;
You will answer, O Lord my God.
(ὅτι ἐπὶ σοί κύριε ἤλπισα σὺ εἰσακούσῃ κύριε ὁ θεός μου)
Psalm 97:5: The mountains melted like wax at the presence of the LORD,
At the presence of the Lord of the whole earth.
(ἀπὸ προσώπου κυρίου
ἀπὸ προσώπου κυρίου!)
ἀπὸ προσώπου κυρίου!)
Maybe in the months and years ahead a more detailed analysis will reveal more reasons for the Psalms grammatical differentiation, e.g. in relation to the more modest increase in Greek articles for Yahweh translations, the specific genre and poetic goals of the translators within their understanding of the genre and so on. But these examples should still serve as a witness that even Psalms translator sees Adonai and Yahweh as synonymous at critical points.
Another hypothesis well worth testing out in my view: that the Psalms translation took place earlier (or later) than the bulk of the other translations and represents an earlier and less complete assimilation of Adonai to Yahweh.
Some options for translations based on this research
Because of this synonymy criterion, NETS and other translations are probably safe to continue treating Adonai and Yahweh similarly, perhaps even scrapping the small caps “Lord” system or applying it to non-combo Adonai also. If you want to borrow the idea of translating the Hebrew into “Lord” from the Greek translators, why not do it as they did it? Sometimes, however, we have seen translation committees opt for transliteration, both in ancient and more recent times.
- 1. Transliterate both Adonai and Yahweh. If we want to be most faithful to the most original form of the Old Testament text, then we should consider the possibility of simply stating “Adonai” for ינָדֹאֲ and “Yahweh” for הוִ֗היְ֝ . That, however, would lose a sense of lordship that would have been there for the Hebrews on Adonai.
- 2. Transliterate Yahweh but translate Adonai to “Lord”. A problem here is that once you introduce a title for either term, then the issue of the article comes up in an artificial way since we have seen it affects both terms. However, given the scarcity of Adonai references and the space for removing articles that “Lord Yahweh” still provides, that might not be a major problem.
- 3. Translate Yahweh and Adonai to one of the following
a. “the Lord”
b. “the Lord”
c. “Lord”
d. “Lord”
This option also helps make sense of reading the New Testament’s many citations of the Greek Old Testament.
For a translation like NETS that seems to want to highlight specific book translations and structure, my money is on a representative version of 3, where the grammatical distinctions of both Psalms and human Lords could be represented via the presence or absence of the English definite article, “the”.
For a more harmonising translation, the rendering needn’t be complicated by Psalms’ divergent grammatical practice and could translate Yahweh and Adonai consistently throughout as “Lord” or “Lord”.
OK, that pretty much wraps it up for today, I hope you have found it interesting or informative and my handling of Adonai careful. My next steps are to proceed with the Yahweh arthricity rates, which is a much significant data set than Adonai (readers may recall that I have myself only done Psalms and Ezekiel for the time being).
Related posts:
[i] Judges 13:8 also mentions Αδωναιε alongside κύριε, albeit in the reverse order and constitutes a simple addition (i.e. does not serve to remove an Adonai-Yahweh redundancy) and Jeremiah in 34:5 uses αδων to translate a theatricalised scene of subjects calling out in lament for their human master.
No comments:
Post a Comment
Thanks very much for your feedback, really appreciate the interaction.