Zephaniah 3:1: Modern vs. Ancient Translations?

By God’s grace, English translations of the Bible are generally of very high quality, far higher than the translations of any other ancient text.  More effort is put into securing just the right sense and nuance when translating the Bible than anything else, because so much is at stake.  Due to my over-education, I have the rare privilege of reading English translations not only alongside the original Hebrew, but alongside other translations used by the earliest Christians in Greek (the Septuagint), Syriac (the Peshitta), and Latin (the Vulgate).  And when I came to Zephaniah 3:1, I noticed something strange: unusually, all the English translations I looked at disagreed with all the early versions.  What’s going on here?The Masoretic Hebrew text of Zeph. 3:1 would be read something like hôy mōrʾāh wenigʾālāh hāʿîr hayyônāh.  I will refer to those words repeatedly.

Almost all modern English translations are heavily indebted to the King James Version (KJV), which translates this verse: “Woe to her that is filthy and polluted, to the oppressing city!”  We can break down the translation word-by-word: hôy means “Woe,” mōrʾāh is interpreted as a feminine singular active participle of the Qal (“simple”) stem from the root m-r-ʾ (whose meaning we will discuss shortly, but interpreted here as “filthy”), we means “and,” nigʾālāh is a feminine singular participle of the Nif`al (“passive”) stem from the root g-ʾ-l (here, “polluted”), hāʿîr is the definite article (“the”) attached to the word “city,” and finally hayyônāh is the definite article attached to the feminine active participle of the Qal stem of the verb “to oppress” (stem y-n-h).  A straightforward translation of a straightforward verse.

In comparing other English versions translated from the Hebrew (not the Douay-Rheims and the Brenton Septuagint translations which are translated from the Latin Vulgate and the Greek Septuagint versions, for which see below), there is not much disagreement on this translation.  Some modern versions have adopted a wide range of ways to translate “Woe” into a more contemporary idiom (how many times in a day do you say “woe”?), but otherwise the only disagreement among the translations seems to be that the word “filthy” is replaced by “rebellious” among many modern English translations (e.g. NIV, ESV, NASB, NKJV, CSB, ASV, and NRSV).  We can understand this alternate translation by noticing that the root m-r-ʾ does not occur anywhere else in the Bible, so perhaps it is simply a variation of the common root m-r-h (which means “to rebel”).  This sort of interchange is not uncommon among verbs whose final root letter is an h, while the meaning “filthy” may be nothing more than a contextual inference from the following word “defiled.”  Any of these English translations seem solid.

And yet the ancient versions seem to tell a different story.  The earliest known written translation of the scriptures is the Greek Septuagint, which reads (in Rahlfs’ ed.): Ὦ ἡ ἐπιφανὴς καὶ ἀπολελυτρωμένη, ἡ πόλις ἡ περιστερά (which I would translate “O notable and redeemed one, the city, the dove”).  The only words this shares with the English translations are “the city” and “and.”  The Syriac Peshitta is similar, with a different ending: ܐܘܝ ܡܕܝܢܬܐ ܝܕܝܥܬܐ ܘܦܪܝܩܬܐ ܡܕܝܢܬܗ ܕܝܘܢܢ (“Woe, known and redeemed city, the city of Jonah”).  The Latin Vulgate, which always prefers to follow the Hebrew rather than the Greek, nevertheless agrees with it here except for one point of contact with the modern English versions more familiar to us: vae provocatrix et redempta civitas columba (“Woe, provoking and redeemed city, the dove”).  So while there are individual places where each of these ancient translations departs from the others (Septuagint’s softening of “woe,” Peshitta’s “Jonah,” and the Vulgate’s “provoking”), in general they agree with one another against the modern English versions.  How did we get here?

The ancient interpretations are certainly tenable based on the Hebrew.  We can take them word-by-word.  The Hebrew interjection hôy can mean “woe,” but it is not as specific as, for example, wāy (modern “vey”), and it can be a more generic interjection (as in the Septuagint).  The Septuagint and the Peshitta (but not the Vulgate) interpret mōrʾāh as a feminine Pu`al (intensive passive) or Hof`al (causative passive) participle of the root r-ʾ-h, whose base meaning is “to see,” and thus “one widely seen” or “one caused to be seen,” hence the translations “notable” or “known.”  All three versions translate nigʾālāh as a feminine Nif`al (“passive”) participle from the root g-ʾ-l (here meaning “redeemed”; both meanings are securely attested).  And finally, the Septuagint and the Vulgate interpret hayyônāh as the definite article with the noun “dove” (while the Peshitta must omit this definite article and the preceding one, reading ʿîr yônāh to get the meaning “city of Jonah” instead of hāʿîr hayyônāh, since the name “Jonah” is also the Hebrew word for “dove”).  In other words, although almost all modern English translations pick the negative meanings for these adjectives, the ancient versions mostly pick positive meanings to describe the city.  (The Aramaic Targum Jonathan seems to be entirely different, sharing only the word “city” with any other translation, ancient or modern!  So I have omitted it in my analysis here.)

Who is right?  On strict linguistic grounds it is hard to adjudicate these differences.  Since the city is Jerusalem, and so far as we know Jonah never went there (he was a northern prophet), it would be very surprising if the Syriac Peshitta reading “city of Jonah” was correct.  The translation of the first adjective as “filthy” (KJV) depends upon giving a meaning to this root which is otherwise unknown, but then the root itself is otherwise unknown.  (Often if a Hebrew root is otherwise unknown, scholars wonder if it really is a root, or if we should read the word differently.)  But the meanings “conspicuous” or “rebellious” for the first adjective, “defiled” or “redeemed” for the second adjective, and “oppressing” or “dove” for the final word, all seem roughly equally well justified.  In consequence, which reading is preferred will largely come down to other factors, whether positive or negative adjectives make more sense in this context, or whether one is inclined or disinclined to trust the early translations.

But what if the fact that this verse can be read both ways is not a frustrating failure of clarity, but rather an important point about the message of this verse?  Zephaniah’s prophecy, like most Old Testament prophecy, is also poetry, and poets delight in double meanings.  In this case, the duality may not be idly playing with words, but very serious: the Lord is simultaneously expressing his love for Jerusalem, and indicting them for their defilement with oppression.  On the one hand he uses affectionate pet names to call to her and remind her of her salvation, “Oh!  The city, the dove, is famous and redeemed,” but at the same time our righteous God declares it time to call a spade a spade by saying, “Woe!  The oppressive city is rebellious and defiled!”  The greater God’s favor toward his people, the greater our betrayal of God’s love by our acts of rebellion and oppression, so the two readings are two wings of the same bird, bringing the Lord’s message to his people, a message of God’s gracious relationship with us, and at the same time a message of judgment for our iniquities.  And in Hebrew, those can be the same words, but in English, no one would accept a translation that read, “(Oh/woe)!  The (dove/oppressive) city is (famous/rebellious) and (redeemed/defiled).”  Perhaps we need to allow this dual message of grace and judgment to alight in our hearts and prick us to repentance, turning away from oppression and returning into the arms of our God who loves us.


  1. Interesting. I didn’t realize there were verses that could vary so radically in translation. It’s always hard for me to get an intuitive feel for wordplay in other languages. However, I’m left with one question: what is the rabbinical interpretation of the Masoretic text? (Or is there a standard one?) For a prophecy that’s not obviously Messianic, I’d be inclined to give that a fair amount of weight in figuring it out.

    1. I should also note that this degree of ambiguity is highly unusual, perhaps unique! Biblical translations are generally very good, and you can often get a sense of the ambiguity by reading different English versions. What struck me about this case was the combination of English translations’ near unanimity disagreeing with ancient versions’ near unanimity; I’ve never seen anything like it.

  2. Good question! A priori, I would not prioritize rabbinic interpretations of the Masoretic text, especially as the Masoretic text is in a number of places inferior to the other ancient witnesses, and almost no rabbinic commentators predate 1000 CE (and the rabbinic commenters we have were also not native Hebrew speakers). Asking for a “standard” rabbinic interpretation is also not likely to lead anywhere, but one can check various rabbinic interpretations and see if there are commonalities.

    The earliest thing that might qualify as a rabbinic commentary is in fact Targum Jonathan (which I mentioned in passing in the original post), and which reads at this verse, “Woe of the one hastening and redeemed, the city which increases that I might be angry” (or something like that; my grasp of eastern Targumic Aramaic is weak). So apart from “Woe,” “redeemed,” and “city,” it charts a unique course unrelated to either the other ancient versions or the modern English translations.

    I would need to know the Talmud a whole lot better to know how to find out whether the Talmud ever cites this verse, but otherwise the next rabbinic commentary I know of is that of Rashi (d. 1105), whose importance as a rabbinic biblical commenter is unparalleled. I found his commentary translated at chabad.org:
    “Woe to her who is filthy and polluted”: Until now, Zephaniah was speaking of Ninveh, but now he refers to Jerusalem: Woe to her who is destined to be plundered and sullied like dung. It will be polluted with the filth of its iniquity.
    “filthy”: as in (Lev. 1:16) “its crop” ; and (Nahum 3:6) “And I will make you like dung.”
    “the dove like city”: Like a silly dove, without a heart.
    So Rashi’s reading actually agrees with KJV in the first half, but disagrees in the second half. (And Rashi provides a different etymological justification for the translation “filthy,” although I often find Rashi’s etymologies clever and implausible.) Nevertheless, Rashi was read not only by medieval Jewish Bible scholars, but also by medieval Latin Christian Bible scholars, so KJV might in fact be drawing on a tradition directly indebted to Rashi.

    After Rashi, I think the next rabbinic commentary on Zephaniah would be found in the work of Radak (David Kimhi), who died in 1235. But I cannot find an online version to check how he handles the verse. Nor can I find an online version of the commentary of Abrabanel (d. 1508) to check, and that concludes all the rabbinic commentaries on Zephaniah written before 1500 of which I have been able to learn in response to your question.

    So it looks like Rashi might be the missing link between the ancient versions and the KJV (which was heavily influenced by the Vulgate, but was trying to be a new translation from the Hebrew Masoretic text), and especially provides an explanation for the otherwise unaccountable translation “filthy.” But Targum Jonathan is just unique. I hope this tour of rabbinic interpretations is helpful!

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s