My concern here is not with what Jesus actually said or what language he said it in. I teach Ancient Hebrew as a living language, and I am on the lookout for comprehensible Hebrew texts in biblical or Second Temple idiom that I can use to help develop my own and my students’ fluency. To be sure, there are many excellent ancient Jewish prayers in Hebrew, but because it is known and used, the Lord’s Prayer is a good choice for my Christian students. In the absence of a Hebrew textus receptus the problem is deciding which translation to use.
My original post compared Delitzsch (1877), Salkinson-Ginsberg (1886) and the 1977 Bible Society of Israel. I recently stumbled upon two additional versions published by Christian Czerskier (1830 / rev. 1836) and Christian Hermann Friedrich Bialloblotzky (1833) while searching for Hebrew translations of the Anglican Book of Common Prayer. As both versions have the advantage of being farther removed from the changes that took place when Hebrew was adopted as the primary spoken language of Jews in the land of Israel, I thought I would update my translation comparison by adding in these two older and less well known versions.
I also intended to comment on the lack of attention to Hebrew versions of the Lord’s prayer. … But then I checked ATLA and discovered Jean Carmignac’s 1978 analysis of an astonishing 68 different Hebrew translations of the Lord’s prayer ranging from the 9th century to 1976.[3]
Carmignac explains that both the Czerskier and Bialloblotzky translations are derived from the 1814 translation by the London Society for Promoting Christianity amongst the Jews. Czerskier adopted the London Society’s 1833 revision (77 n. 116), while Bialloblotzky made one change to William Greenfield’s 1831 revision of the London Society’s 1814 translation. As I had settled on Bialloblotzky as the more-biblical sounding translation (to my ear), I was pleased to see the London Society’s translation included alongside Delitzsch in Carmignac’s shortlist of eight “that stand out above all the rest.”
Carmignac’s essay also defends his own translation on the basis of parallels in the Dead Sea Scrolls.
In the rest of this post, I compare Delitzsch (1877) [D] with Greenfield-Bialloblotzky (1831 / 1833) [B] and Carmignac (1969 / 1978) [C], adding notes mostly drawn from Carmignac (1978):
| (D) Delitzsch (1877) | (B) Bialloblotzky (1833) | (C) Carmignac (1969) |
|---|---|---|
| אָבִינוּ שֶׁבַּשָׁמַיִם יִתְקַדַּשׁ שְׁמֶךָ תָּבֹּא מַלְכוּתֶךָ יֵעָשֶׂה רְצוֹנְךָ | אָבִינוּ אֲשֶׁר בַּשָּׁמַיִם יִתְקַדֵּשׁ שְׁמֶךָ תָּבֹא מַלְכוּתֶךָ יְהִי רְצוֹנְךָ | אָבִינוּ אֲשֶׁר בַּשָּׁמַיִם יִקָּדֵשׁ שִׁמְכָה תָּבוֹא מַלְכוּתְכָה יֵעָשֶׂה רְצוֹנְכָה |
| כַּאֲשֶׁר בַּשָׁמַיִם גַּם בָּאָרֶץ | כַּאֲשֶׁר בַּשָּׁמַיִם כֵּן גַּם בָּאָרֶץ | כַּאֲשֶׁר בְּשָׁמַיִם וְעַל אָרֶץ |
| אֶת־לֶחֶם חוּקֵּנוּ תֵּן־לָנוּ הַיּוֹם | לֶחֶם חֻקֵּנוּ תֵּן לָנוּ הַיּוֹם | לַחְמֵנוּ לְמָחָר תֵּן לָנוּ יוֹם יוֹם |
| וּמְחַל־לָנוּ עַל־חֹבוֹתֵינוּ | וּסְלַח לָנוּ אֶת־חֹבֹתֵינוּ | וְשָׂא לָנוּ נְשֵׂינוּ |
| כַּאֲשֶׁר מָחַלְנוּ גַּם־אֲנַחְנוּ לְחַיָּבֵינוּ | כַּאֲשֶׁר אֲנַחְנוּ סֹלְחִים לְבַעֲלֵי חוֹבֵינוּ | כַּאֲשֶׁר גַּם אֲנוּ נָשָׂאנוּ לְמַשֵּׂינוּ |
| וְאַל־תְּבִיאֵנוּ לִידֵי נִסָּיוֹן | וְאַל תְּבִיאֵנוּ לְמַסָּה | וְאַל תְּבִיאֵנוּ בְמַסָּה |
| כִּי אִם־תְּחַלְּצֵנוּ מִן־הָרָע | כִּי אִם הַצִּילֵנוּ מִן הָרָע | כִּי (אִם) הַצִּילֵנוּ מִן הָרָשָׁע |
| כִּי לְךָ הַמַּמְלָכָה וְהַגְּבוּרָה וְהַתִּפְאֶרֶת לְעוֹלְמֵי עוֹלָמִים אָמֵן | כִּי לְךָ הַמַּמְלָכָה וְהַגְּבוּרָה וְהַתִּפְאֶרֶת לְעוֹלָמִים אָמֵן |
Our Father which art in heaven
Πάτερ ἡμῶν ὁ ἐν τοῖς οὐρανοῖς
D: אָבִינוּ שֶׁבַּשָׁמַיִם
B / C: אָבִינוּ אֲשֶׁר בַּשָּׁמַיִם
Although the shortened relative pronoun, שׁ, becomes common in Late Biblical Hebrew, Carmignac (p. 60) notes that the DSS prefer the more common Biblical Hebrew form אשׁר, and that אשׁר בשׁמים appears in the Ten Commandments (Exod 20:4; Deut 5:8) and at Qumran while שׁבשׁמים is first attested in Rabbinic literature.
Hallowed be thy name
ἁγιασθήτω τὸ ὄνομά σου
D: יִתְקַדַּשׁ שְׁמֶךָ
B: יִתְקַדֵּשׁ שְׁמֶךָ
C: יִקָּדֵשׁ שִׁמְכָה
After reviewing other translations, Carmignac felt “a strong inclination” to change his 1969 translation of ἁγιασθήτω from the Niphal (יִקָּדֵשׁ) (also in Greenfield’s revision) to the Hithpael (יִתְקַדֵּשׁ) form adopted by the majority of Hebrew translations, noting the use of the Hithpael form in the Jewish Qaddish prayer and in Qumran. Since the meaning of Niphal and Hithpael forms can overlap, the Hithpael (as in D and B) is to be preferred. (Carmignac’s שִׁמְכָה instead of שְׁמֶךָ simply follows Qumran orthography.)
Thy kingdom come
ἐλθέτω ἡ βασιλεία σου
D: תָּבֹּא מַלְכוּתֶךָ
B: תָּבֹא מַלְכוּתֶךָ
C: תָּבוֹא מַלְכוּתְכָה
Carmignac simply prefers the fuller Qumran orthography here and throughout the prayer. I assume the dagesh in Delitzsch’s תָּבֹּא is a mistake. Carmignac explains that the Late Biblical Hebrew מַלְכוּת refers to “the act of reigning,” while מַמְלָכָה means “realm.”
Thy will be done
γενηθήτω τὸ θέλημά σου
D: יֵעָשֶׂה רְצוֹנְךָ
B: יְהִי רְצוֹנְךָ
C: יֵעָשֶׂה רְצוֹנְכָה
The difference between the Niphal jussive יֵעָשֶׂה preferred by Delitzsch and Carmignac and the Qal jussive יְהִי adopted by Bialloblotzky is, according to Carmignac, between “to be accomplished” and “to become.”
In the Bible רָצוֹן is only combined with the verb היה when it is preceded by the preposition לְ (לְרָצוֹן) (Lev 22:20-21; Ps 19:15), while עשׂה is quite often collocated with רָצוֹן on its own in the sense of doing someone’s will (e.g., Ezra 10:11; Ps 40:9; 103:21; 143:10; multiple passages in the DSS; Avot. 2.4).
However, the DSS do combine היה and רָצוֹן in reference to God’s will (e.g., 1QS xi 18; 1QH ix 22, xviii 4; 4Q393, 4Q418). And in the Mishnah, יְהִי רָצוֹן does occur in connection with prayers for God’s will to be done (Ber. 9.3 [rendered by OAM as “May it be God’s will”]; Taan. 4.8; Avot. 4.8).
This leads to a bigger and more exegetically significant question: Does the prayer (primarily) invite human participation, in effect offering oneself up to the divine will and asking to be conformed in action to God’s will so that God’s will is done in the pray-er’s doing? Or does the prayer (primarily) invite God to do God’s eschatological will (as in Taan. 4.8; Avot. 4.8)? Here, I’m guessing יֵעָשֶׂה רְצוֹנְךָ is a better translation—in other words, the prayer’s emphasis is on human involvement in doing God’s will—but I still sort of prefer the 1814 London Society’s יְהִי רְצוֹנְךָ.
In earth, as it is in heaven
ὡς ἐν οὐρανῷ καὶ ἐπὶ γῆς
D: כַּאֲשֶׁר בַּשָׁמַיִם גַּם בָּאָרֶץ
B: כַּאֲשֶׁר בַּשָּׁמַיִם כֵּן גַּם בָּאָרֶץ
C: כַּאֲשֶׁר בְּשָׁמַיִם וְעַל אָרֶץ
Carmignac’s version attempts to preserve the difference between Greek prepositions. I prefer Greenfield-Bialloblotzky’s emphatic rendering of καί (see Deut 12:30 for the expression). Carmignac counts 25 different ways the phrase has been translated into Hebrew. Take your pick.
Give us this day our daily bread
τὸν ἄρτον ἡμῶν τὸν ἐπιούσιον δὸς ἡμῖν σήμερον
D: אֶת־לֶחֶם חוּקֵּנוּ תֵּן־לָנוּ הַיּוֹם
B: לֶחֶם חֻקֵּנוּ תֵּן לָנוּ הַיּוֹם
C: לַחְמֵנוּ לְמָחָר תֵּן לָנוּ יוֹם יוֹם
The Delitzsch and Greenfield-Bialloblotzky renderings, which differ only in the spelling of חֹק and the presence or absence of the direct object marker, draw on Prov 30:8 where לֶחֶם חֻקִּי refers to the speaker’s daily food allotment (NJPS / NIV: “my daily bread”). Carmignac’s לְמָחָר (“for tomorrow”) proposal seems closer to the meaning of ἐπιούσιον, but יוֹם יוֹם renders Luke’s καθ‘ ἡμέραν. For Matthew’s σήμερον, הַיּוֹם is to be preferred.
And forgive us our debts
καὶ ἄφες ἡμῖν τὰ ὀφειλήματα ἡμῶν
D: וּמְחַל־לָנוּ עַל־חֹבוֹתֵינוּ
B: וּסְלַח לָנוּ אֶת־חֹבֹתֵינוּ
C: וְשָׂא לָנוּ נְשֵׂינוּ
Delitzsch’s מחל means “forgive” in Rabbinic Hebrew, but not in the Hebrew Bible or in the DSS. (According to the DSS database in Accordance, 4Q267 is the only possible, but very fragmentary, exception.) Carmignac prefers the combination of נשׂא, which can mean forgive, and the cognate, נְשִׂי, which appears once in 2 Kings 4:7 in the sense of “debt.” I prefer Bialloblotzky’s use of סלח, a common verb with the primary meaning “to forgive.” A singular form of the noun חוֹב does occur once in Ezek 18:7 with the sense of guilt or debt; the feminine form חוֹבָה shows up in the Mishnah. I’m siding in favour of clear meaning over word-play. Again, take your pick.
As we forgive our debtors
ὡς καὶ ἡμεῖς ἀφήκαμεν [maj. text ἀφίεμεν] τοῖς ὀφειλέταις ἡμῶν
D: כַּאֲשֶׁר מָחַלְנוּ גַּם־אֲנַחְנוּ לְחַיָּבֵינוּ
B: כַּאֲשֶׁר אֲנַחְנוּ סֹלְחִים לְבַעֲלֵי חוֹבֵינוּ
C: כַּאֲשֶׁר גַּם אֲנוּ נָשָׂאנוּ לְמַשֵּׂינוּ
“Here we reach the maximum disagreement with fifty-eight divergent opinions” (p. 66). Delitzsch’s לְחַיָּבֵינוּ adopts the adjective חַיָּב, which appears once in Sirach 8:5 (according to DCH) as well as in the Mishnah.
Carmignac continues the word play on the root נשׂא by constructing a Hiphil participle with an otherwise unattested meaning of “debtor.” The London Society’s 1814 translation renders τοῖς ὀφειλέταις ἡμῶν as לחובינו, apparently reading a Qal participle from the verb חוב, which does not occur in the Hebrew Bible but does at Qumran and the Mishnah.
The Greenfield-Bialloblotzky revision seems mistakenly to conflate the London Society’s חוֹבֵינוּ (“our debtors”) with לבעלי חובותינו (“the masters of our debts”), a phrase borrowed from a Hebrew translation of Matthew completed ca. 1520-1530, which, in turn, draws on בַּעַל מַשֵּׁה in Deut 15:2, substituting the more common חובה for Deuteronomy’s משׁא (a hapax). According to Carmignac, “the mere length of this phrase makes [the] suggestion hardly possible” (see pp. 66, 30-31 for discussion).
In this case, I prefer the London Society’s reading: כַּאֲשֶׁר אֲנַחְנוּ סֹלְחִים לְחוֹבֵינוּ.
And lead us not into temptation
καὶ μὴ εἰσενέγκῃς ἡμᾶς εἰς πειρασμόν
D: וְאַל־תְּבִיאֵנוּ לִידֵי נִסָּיוֹן
B: וְאַל תְּבִיאֵנוּ לְמַסָּה
C: וְאַל תְּבִיאֵנוּ בְמַסָּה
As Carmignac notes, מַסָּה is better attested than נִסָּיוֹן. The difference between the prepositions בְּ and לְ in this context seem minimal, though the Hebrew Bible usually has בְּ in connection with the place.
But deliver us from evil
ἀλλὰ ῥῦσαι ἡμᾶς ἀπὸ τοῦ πονηροῦ
D: כִּי אִם־תְּחַלְּצֵנוּ מִן־הָרָע
B: כִּי אִם הַצִּילֵנוּ מִן הָרָע
C: כִּי (אִם) הַצִּילֵנוּ מִן הָרָשָׁע
Carmignac prefers רָשָׁע over רָע because at Qumran רָשָׁע has the sense of a wicked entity, but in the LXX πονηρός almost always translates רע not רשׁע.
For thine is the kingdom, and the power, and the glory, for ever. Amen.
Maj. text: Ὅτι σοῦ ἐστιν ἡ βασιλεία καὶ ἡ δύναμις καὶ ἡ δόξα εἰς τοὺς αἰῶνας. Ἀμήν.
D: כִּי לְךָ הַמַּמְלָכָה וְהַגְּבוּרָה וְהַתִּפְאֶרֶת לְעוֹלְמֵי עוֹלָמִים אָמֵן
B: כִּי לְךָ הַמַּמְלָכָה וְהַגְּבוּרָה וְהַתִּפְאֶרֶת לְעוֹלָמִים אָמֵן
C: No translation provided
This results in an eclectic text:
אָבִינוּ אֲשֶׁר בַּשָּׁמַיִם
יִתְקַדֵּשׁ שְׁמֶךָ
תָּבֹא מַלְכוּתֶךָ
יֵעָשֶׂה רְצוֹנְךָ or יְהִי רְצוֹנְךָ
כַּאֲשֶׁר בַּשָּׁמַיִם כֵּן גַּם בָּאָרֶץ
לַחְמֵנוּ לְמָחָר תֵּן לָנוּ הַיּוֹם
וּסְלַח לָנוּ אֶת־חֹבֹתֵינוּ
כַּאֲשֶׁר אֲנַחְנוּ סֹלְחִים לְחוֹבֵינוּ
וְאַל תְּבִיאֵנוּ בְמַסָּה
כִּי אִם הַצִּילֵנוּ מִן הָרָע
One final note: The Hebrew version of the Lord's Prayer in the Pater Noster Chapel in Jerusalem pictured in the Wikimedia Commons photograph at the beginning of this post, reproduces a 1593 translation by Elias Hutter with a few minor changes (Carmignac 38).
- I say “supposed” because, as I have explained elsewhere, I do not assume Jesus taught—or prayed!—only in Aramaic. I note in passing that the Aramaic word for “father” is not the smoking gun it may appear to be. To be sure, the Aramaic word, “Abba,” used by Jesus (and other Jews) in addressing God as father, was adopted by the Greek-speaking early church (see Rom 8:15; Gal 4:6). But the use of “Abba” for father need not mean Jesus taught and prayed the “Lord’s Prayer” in Aramaic. After all, Hebrew-speaking Israelis today use “Abba” as a familial address. The word may originate as an Aramaic loanword, but it has been adopted into modern Hebrew. Let’s imagine for a moment that Second Temple Period Hebrew did the same.↩︎
- The next most popular post, on Martin Luther’s Preface to the Revelation of St. John (1522), rings in at 14,500 views. “Reconstructed Koine Greek Pronunciation” (8,510 views) and “Ancient Greek Picture Flashcards in Anki” (5,710 views) come in 3rd and 4th.↩︎
- Jean Carmignac, “Hebrew Translations of the Lord’s Prayer: An Historical Survey,” in Biblical and Near Eastern Studies: Essays in Honor of William Sanford LaSor, ed. Gary A. Tuttle (Grand Rapids: Eerdmans, 1978), 18–79. More recent attempts to reconstruct a Hebrew Lord’s prayer include Charles R. Day, “The Lord’s Prayer: A Hebrew Reconstruction Based on Hebrew Prayers Found in the Synagogue,” Conspectus 7 (2009): 22–31 and Alexander Ponomariov, “The Lord’s Prayer in a Wider Setting: A New Hebrew Reconstruction,” Journal of Northwest Semitic Languages 41.1 (2015): 71–100. Neither Day nor Ponomariov appear to have consulted Carmignac’s review of earlier translations. This is unfortunate, for, as Carmignac noted, “The value to any research project of having access to collections of previous efforts is precisely the help they afford in avoiding what is specious and ephemeral.” More helpful is David N. Bivin and Joshua N. Tilton’s massive 2011 essay, which can be found here: https://www.jerusalemperspective.com/769/. Eitan Berkowitz, “The Hebrew Translation of the Carolingian Lord’s Prayer: A Case Study in Using Linguistics to Understand History,” AJS Rev. 44.1 (2020): 23–47, offers a fresh analysis of the 9th century Hebrew version that survives in Latin transliteration. ↩︎
