The Book of 5 Ezra

(Online version of a lecture by J. R. Davila on 16 February, 2007)

(The sources for the content of these lectures vary considerably.  Some, like last week’s, come mostly from my own research.  Others have a mixture of my research and that of others.  This one is based almost entirely on the research of Theodore Bergen, who has published much important work on 5 Ezra, for which, see the bibliography.  It is also with his permission that I have posted his translation of 5 Ezra, which is the best available. Professor Bergren will also be providing the “5 Ezra” contribution for the More Old Testament Pseudepigrapha Project.)

THE NOMENCLATURE OF 2 ESDRAS AND ITS COMPONENTS:  The naming of the various books attributed to Ezra in the ancient manuscripts and early editions of the Bible is quite inconsistent and somewhat confusing.  I will not survey this problem in any detail here, but it is worthwhile to set down the standard usages among scholars when dealing with the Latin book that contains prophecies attributed to Ezra, including 5 Ezra.

  • 2 Esdras refers to the Latin version of the book, containing 16 chapters total.
  • 4 Ezra refers to chapters 3-14 of 2 Esdras.  4 Ezra is an apocalypse translated from Greek (now lost) into Latin (and also Syriac).
  • 5 Ezra refers to chapters 1-2 of 2 Esdras, which survive only in Latin and which seem to constitute a separate work that perhaps was familiar with 4 Ezra.  It is the subject of today’s lecture.
  • 6 Ezra refers to chapters 14-15 of 2 Esdras, which also was probably originally a separate work.  It too survives complete only in Latin, although one Greek fragment is preserved as well.  We will cover this book later in the semester.

THE NOMENCLATURE OF THE LATIN BIBLE TRANSLATIONS:  It may be helpful as well to review briefly some information about the ancient Latin translations of the Bible.

  • The Old Latin (Vetus Latina) is an early translation (or cluster of translations) of the Greek scriptures (the Septuagint or LXX and the New Testament) into Latin.  The origin of this translation is unclear, but it appears in our sources first in Africa in the second century C.E., it was revised in Europe, and it survives in various forms in widely variant manuscripts and quotations.
  • The Latin Vulgate is a translation of the Hebrew Bible and the Greek New Testament from the original languages into Latin.  It was produced by the church father Jerome in Bethlehem in the late fourth century C.E.

5 EZRA:  TEXT AND MANUSCRIPTS:  There are two recensions of 5 Ezra, generally known as the French and Spanish recensions.  The earliest French manuscript (the St. Germain manuscript) consists of a complete Bible; is dated to 821/22 CE; and includes 5 Ezra4 Ezra, and 6 Ezra, in that order.  One leaf of 4 Ezra was cut out, due to its containing a theologically objectionable rejection of prayers for the dead.  There are sixty-one other manuscripts of the French recension, but sixty of them descend from the St. Germain manuscript (they too are missing the material on the page cut out of that manuscript) and so only one is of independent value.  There are six Spanish manuscripts, all of which give the order 4 Ezra6 Ezra5 Ezra, with some indicating a major or minor division between the first two and 5 Ezra.  The logical implication is that 4 Ezra and 6 Ezra became a unit first, with 5 Ezra added afterward either at the beginning or end.  There is also good evidence that the end of 4 Ezra was mutilated during this process, as was possibly also the beginning of 6 Ezra.

TWO RECENSIONS:  There are good reasons for thinking that the Spanish recension (S) is closer to the original text.  Unfortunately, the French recension (F) is the better of the two for the text of 4 Ezra, so all of the major translations of 2 Esdras (including the RSV) are of the French recension.  Bergren’s translation is an eclectic reconstructed text that is based mostly on the Spanish recension.  Time does not permit a full survey of the variants (which can be found in Bergren’s monograph), but I will review several that illustrate the point.

  • 5 Ezra 1:1 S refers to “Ezra son of Chusi in the days of Nebuchadnezzar.”  There is no mention of a Chusi in the biblical book of Ezra, and Ezra did not live in the time of Nebuchadnezzar.  But F 1:1-4 corrects the text to conform to the book of Ezra with a biblical genealogy that sets Ezra in the time of Artaxerxes.
  • In 1:11 S the transition between God’s address to Ezra (vv. 5-10) shifts without transition to an address to Israel.  But F softens the transition and makes it explicit in vv. 11-12.
  • In 1:39-40, S refers to Elijah and Enoch and adds ten of the minor prophets in a random order, plus a Mattathias (= Malachi?).  F once again assimilates the passage to the Bible, first omitting Elijah and Enoch in accordance with Luke 13:28, then tidying up the list of minor prophets so that all twelve appear in the correct order.
  • In 2:5-7 S has unmarked transitions between the speech of the “mother,” which ends in 2:5a; the speech of Ezra in 2:5b-6; and the speech of God, which begins in 2:7.  The content of all three make the speaker clear.  The mother is “desolate.”  Ezra invokes God as father to punish the people who brought their mother to destruction.  And God curses those people because they have spurned his covenant.  But F tries to make the transitions clearer and doesn’t get it quite right.  Verse 5b is taken to begin the speech of God, so that God addresses Ezra as “father” and God tells Ezra to bring the people and their mother to ruin!  It is very difficult to make any sense of F and it is very likely a botched attempt to improve on S.

In general, F seems more polished and less repetitive than S.  They seem to be connected to an earlier, more original version, but one can debate their exact relationship to it and to each other.  F looks to be fairly heavily revised in comparison to S.  Unfortunately, it is impossible to tell the relationship of S to the original composition.  It may be a very faithful rendition of it or it too may have been heavily revised, like F.  F illustrates that such heavy revision happened and we can never assume that it didn’t even when we lack positive evidence that it did.  This example underlines the point made last week about the problems with relying on the surviving texts of the Old Testament Pseudepigrapha.

ORIGINAL LANGUAGE:  5 Ezra is preserved only in Latin, so this must be our starting point, and the question becomes whether there are good reasons to move backwards from the Latin version that lies behind our two recensions to a hypothetical Vorlage in another language (presumably Greek, possibly translated from Hebrew or Aramaic).  We have good evidence that both 4 Ezra and 6 Ezra were translated from Greek, but no comparable case can be made for 5 Ezra.  No Greek fragments of 5 Ezra survive and there are no compelling examples where corrupt or senseless Latin in 5 Ezrasuddenly makes sense if we posit a misunderstanding of an original Greek version (although one or two such cases are possible).  Nor does the vocabulary contain enough transliteration of unusual Greek words to point toward translation from Greek.  There are numerous Semitisms and a few Graecianisms in the Latin, but the text shows dependence on the Old Latin translation of the LXX, and both can be explained by the writer’s familiarity with biblical translation Latin.  The Latin of 5 Ezra is fairly unliterary and unsophisticated, which makes it unlikely that its translation from Greek would have been so idiomatic as to be scarcely detectable.  In short, the best working hypothesis is that 5 Ezra was composed in Latin.

DATE OF COMPOSITION:  The best evidence for narrowing down the date of 5 Ezracomes from quotations:  quotations of 5 Ezra in other works (for a terminus pro qua – a date by which it must have existed) and quotations of other works in 5 Ezra (for a terminus a quo – a date before which it cannot have existed).

First, let us consider quotations of other works in 5 Ezra.  The work De altercatione ecclesiae et synagogue (an anti-Jewish composition presenting a hypothetical debate between the Church and the Synagogue) contains free quotations of 5 Ezra 1:24 and 2:2-3, which are attributed to “Esdras.”  This work is widely but not entirely universally taken to have been composed in the mid-fifth century C.E.  In addition, the Acta Sylvestri (a fifth-century work that describes the adventures of Sylvester, the bishop or pope of Rome in the early fourth century) quotes 5 Ezra 1:32abc, a passage secondarily interpolated into some manuscripts of 5 Ezra.  The quotation is attributed to “Hesdras.”  The implication is that the book and the interpolation must have existed by the time of the quotation.  There are other quotations of the book, but these are the earliest that survive.

The book of 5 Ezra also quotes from or alludes reasonably unambiguously to a number of earlier works.  5 Ezra 2:2-5a alludes to Baruch 4:8-23.  (The book of Baruch seems to have been written by 70 C.E.)  Of greater interest are the allusions to the New Testament.  5 Ezra 1:30 seems to be dependent on Matthew 27:37//Luke 13:34 and is closest to Matthew’s version, although we cannot rule out direct dependence on Q (on which both Gospel passages arguably rely).  This implies that 5 Ezra existed by sometime in the second half of the first century C.E.  Likewise, there is possibly a connection between 5 Ezra 1:32 and Matthew 23:34-35//Luke 11:49-50 (or on the Q passage behind them), although this contributes nothing more to the question the dating of the work.  Also, 5 Ezra 2:42 has some connection with Revelation 7:9.  Likewise, 5 Ezra 2:42-43 seems to be dependent on Shepherd of Hermas 83:1 (9.6.1), which involves a crowd with a tall man who is later (89:7-8) identified with the “Son of God.)

In general, the biblical quotations (especially those in the Spanish recension) are closer to the Old Latin than the Vulgate.  This seems to indicate that the author used the Old Latin translation.  One possible reason would be that 5 Ezra was written before Jerome translated the Vulgate (again, in the late fourth century), but it is also possible that the author wrote later than this and was either unfamiliar with the Vulgate or else knew it and preferred the Old Latin for some reason.

To sum up, 5 Ezra seems to have existed by sometime in the fifth century C.E. at the latest.  It is difficult to be sure how much earlier it existed, but it seems familiar with compositions from the first half of the second century C.E.

FORM AND GENRE:  The book of 5 Ezra is largely a collection of prophetic oracles.  I have not yet found form critical analyses of these oracles in the secondary literature, but we can say in general that they are oracles of woe against the people of Israel.  The end of the book (2:42-48) has been argued to be a small apocalypse (a revelation mediated to a human recipient by a divine being).  In it, Ezra reports that he had a vision of a crowd of people being crowned by an exceedingly tall young man.  Ezra questions an angel and learns from him that these are the immortal saints being crowned by the Son of God.

ORIGIN AND BACKGROUND:  Who wrote 5 Ezra and why?  As discussed last week, on methodological grounds the best place to start with such questions is with the earliest manuscripts and quotations of a work and to work backwards from there only as necessary.  Taking that approach, there are good reasons to consider the most reasonable working hypothesis to be that 5 Ezra was composed by a Christian in Latin between the second and fifth centuries C.E.  I have already discussed the questions of the original language and the date above.  As for the authorship, first, the earliest surviving manuscripts were preserved and copied by Christians and, more importantly (since the earliest manuscripts are comparatively late) the work was known and quoted in Christian contexts as early as the fifth century.  Second, the work itself seems to drawn on Christian holy books in a way that is most likely for a Christian author (and unlikely for a Jewish one).  The writer seems to be aware of the Synoptic tradition, the book of Revelation, and the Shepherd of Hermas.  Robert Kraft has raised the objection that many of the ideas and phrases in the apparent allusions are generic, and this is a reasonable point.  But I believe that the cumulative force of the allusions outweighs that objection.  Third, the content is most naturally read as Christian.  The rejection of Jacob, Judah, and Israel (1:2) in favour of “another nation” (1:24), a people coming from a distance who have not seen prophets but who accept their antiquity (1:35-36) makes more sense as referring to Gentiles than as an inner-Jewish conflict.  Likewise, 5 Ezra seems to reject the validity of the Jewish ritual cult — including the festival days, the new moon, and the sabbath — in 1:31, an attitude consistent with gentile Christianity.

Fifth Ezra is best understood as a late antique compendium of Christian oracles which promotes the validity of Christianity over Judaism by putting thinly disguised supersessionist ex eventu (after-the-fact) predictions of the triumph of Christianity in the mouth of the ancient Jewish authority-figure Ezra.

(c) 2007
Reproduction beyond fair use only on permission of the author.

Contact details

St Mary’s College
The School of Divinity
University of St Andrews
South Street
St Andrews
Fife KY16 9JU
Scotland, United Kingdom

Tel: +44 (0)1334 462850 
Fax: +44 (0)1334 462852