The Twenty-Seven-Book New Testament Before Athanasius

Abbreviations

CNT = Bruce Metzger, The Canon Of The New Testament (New York: Oxford University Press, 1997)

GSN = B.F. Westcott, A General Survey Of The History Of The New Testament Canon (New York: The MacMillan Co., 1896)

HOJ = Barbara Bruce, trans., and Cynthia White, ed., Origen: Homilies On Joshua (Washington, D.C.: The Catholic University Of America Press, 2002)

HOL = Joseph Lienhard, trans., Origen: Homilies On Luke, Fragments On Luke (Washington, D.C.: The Catholic University Of America Press, 1996)

INT = D.A. Carson and Douglas Moo, An Introduction To The New Testament (Grand Rapids, Michigan: Zondervan, 2005)

TCD = Lee McDonald and James Sanders, edd., The Canon Debate (Peabody, Massachusetts: Hendrickson Publishers, 2002)

WHO = John McGuckin, The Westminster Handbook To Origen (Louisville, Kentucky: Westminster John Knox Press, 2004)

It's common for Athanasius to be cited as the earliest extant source to advocate the twenty-seven-book New Testament canon of Evangelicalism. Sometimes the claim will be made with some sort of qualification that allows for a collection of those twenty-seven books in some earlier source. We'll be told, for example, that Athanasius is "usually cited" as the earliest source (TCD, 476) or that he "listed" the books first (TCD, 486). If Athanasius is "usually cited", then what's the position of those who don't cite him? And if Athanasius was the first to "list" the books, did somebody advocate the twenty-seven-book collection earlier without presenting the books in the form of a list? Often, such qualifications aren't explained. Sometimes an author will make seemingly inconsistent, or at least unclear, comments on the subject (compare Bruce Metzger's comments about Athanasius on 7, 212, and 292 of CNT with his comments on 139 of the same work).

As Everett Ferguson explains (TCD, 319), Athanasius' canon probably reflects an earlier tradition. Athanasius appeals to "the church" (Festal Letter 39:2), what was "handed down" (39:3), and what was "appointed by the fathers" (39:7). In section 4 of the letter, he refers to the number of Old Testament books as something that was "handed down", so he seems to be including the numbering of the books and their formation in a canon, not just their existence. And a wide diversity of sources writing shortly after Athanasius advocate the same canon (Jerome, Augustine, etc.). Athanasius' appeals to the traditions of the fathers and the acceptance of the same canon by other sources writing shortly afterward suggest that the canon predates Athanasius. It's doubtful that so many sources from so many backgrounds, in so many locations, etc. would accept the same canon in that timeframe if nobody held to that canon before Athanasius.

The issue here isn't whether Athanasius' canon was universally accepted. It wasn't. The letter of Athanasius itself refers to the potential or reality of other canons, and we know that other sources advocated canons different from the one Athanasius advocates. But it does seem likely, from the considerations mentioned above, that Athanasius wasn't the first to hold to the twenty-seven-book canon.

Everett Ferguson writes:

"However, it may be noted that according to Augustine, the Donatist schismatics, who arose out of conflicts resulting from different responses to the command to hand over scriptures [in the Diocletianic persecution in the early fourth century], had no different canon from the Catholic churches of North Africa....Cresc. 1.37; cf. the definition of these in Unit. eccles. 19.51: 'The canonical scriptures of the Law and the Prophets to which are added the Gospels, the Apostolic Epistles, the Acts of the Apostles, and the Apocalypse of John.'" (TCD, 317 and n. 97 on 317)

Augustine accepted our twenty-seven-book canon (On Christian Doctrine, 2:8:13), as did other North African sources contemporary with him (TCD, 320).

The earliest extant source I'm aware of who advocates that twenty-seven-book canon is Origen, more than a century before Athanasius. (For a helpful treatment of Origen's view of the canon, which clears up some misconceptions and cites some significant passages, see William Oliver's article here.) Near the end of his life, Origen commented:

"our Lord, whose advent was typified by the son of Nun [Joshua], when He came, sent His Apostles as priests bearing well-wrought trumpets. Matthew first sounded the priestly trumpet in his Gospel. Mark also, Luke and John, each gave forth a strain on their priestly trumpets. Peter moreover sounds loudly on the twofold trumpet of his Epistles; and also James and Jude. Still the number is incomplete, and John gives forth the trumpet-sound in his Epistles and Apocalypse; and Luke while describing the Acts of the Apostles. Lastly however came he who said: 'I think that God has set forth us Apostles last of all,' and thundering on the fourteen trumpets of his Epistles, threw down even to the ground the walls of Jericho, that is to say all the instruments of idolatry and the doctrines of philosophers." (Homilies On Joshua, 7:1, as cited here)

This passage is controversial. The scholarly reactions to it are diverse. Bruce Metzger argues that the passage is genuine and seems to refer to our twenty-seven-book canon (CNT, 139-140), which is the position I hold. Everett Kalin argues that the passage "owes less to Origen than to Rufinus, who translated it into Latin (the Greek original is lost)" (TCD, 389). Barbara Bruce allows for both possibilities (HOJ, n. 5 on 75).

I want to outline five potential objections to my view of this passage, accompanied by a response to each.

1. The passage may have been interpolated or altered by Rufinus.

There's no direct evidence that the passage or any significant portion of it comes from Rufinus rather than Origen. As far as I know, the passage is in every copy of Origen's Homilies On Joshua.

John McGuckin refers to Rufinus as "generally a reliable translator" (WHO, 31).

Barbara Bruce, in her introduction to a recent translation of the homilies, comments that the "general reliability" of Rufinus' Latin translations of Origen have been vindicated, despite the doubts of earlier scholarship and some scholars in our day (HOJ, 11). She continues:

"Other studies have confirmed the paraphrastic nature of his [Rufinus'] work, but have judged the changes to make for clarity and the thought to remain faithful to the Greek....After explaining how he had expended much labor on changing the hortatory manner of the homilies on Leviticus into the form of an exposition and supplying what was wanting in the homilies on Genesis and Exodus, he said he translated the homilies on Joshua and a few others 'just as we found them, literally and without great effort.' Annie Jaubert, in her French translation of the Homilies, supported Rufinus's statement. She noted constructions that were more dependent on Greek than on Latin syntax and a curtness of speech and density of expression that gave the feel of unpolished notes he may have been working from." (16-18)

Notice that Rufinus explained his different intentions in different contexts. Sometimes he intended to produce a paraphrase. Other times he intended a more literal rendering. And he specified that his work on Origen's Homilies On Joshua was a case of the latter.

It should also be noted that Rufinus did the translation of these homilies upon request (HOJ, 23-24). He could have interpolated a passage or altered the text in the process of doing a translation for somebody else, but the fact that he was doing the work at somebody else's request suggests that he wasn't so interested in altering the text as to initiate the process himself. He was doing a translation that somebody else asked him to do.

And how much interest in canonical issues do we find in the writings of Rufinus? I'm aware of only one passage in his writings in which he lists the New Testament canon, the commonly cited passage in his Commentary On The Apostles' Creed (37). He doesn't seem to have been so concerned about the New Testament canon as to be motivated to change the text of Origen. And as we'll see below, some of the arguments of those who reject this passage in Origen suggest that Rufinus would have been involved in changing other passages in Origen as well. The more passages there are that have to be dismissed as forgeries, the less likely the forgery theory becomes.

Bruce Metzger notes that the passage in question represents "classic Alexandrian oratory" (CNT, 139). (Origen lived in Alexandria for much of his life.) He also notes that the order of the canonical books in Origen is different from the order of the books in Rufinus (CNT, 140). The only other place where we find a comparable order of the books is in the Catalogue Claromontanus, an Eastern source. Origen lived in the East, but Rufinus lived in the West. B.F. Westcott noted that this passage in Origen is "characteristic of Origen's style" (GSN, 368). Barbara Bruce notes the possible influence of Clement of Alexandria on a portion of the passage (HOJ, n. 4 on 74). Clement is more likely to have influenced Origen than to have influenced Rufinus.

And why would Rufinus forge a canon list with ambiguities like the ones discussed below? Why would he refer to the letters of John without specifying their number? By the time of Rufinus, it was widely known that the number of Johannine epistles was in dispute. Somebody interpolating a canon into Origen for the purpose of adding credibility to that canon probably isn't going to leave the number of Johannine epistles unspecified. That ambiguity makes more sense coming from Origen than it does coming from Rufinus. Similarly, if it's to be argued that Origen's list doesn't include Revelation, why would Rufinus leave it out? Compare the passage in Origen to Rufinus' list of the New Testament canon in his Commentary On The Apostles' Creed (37). He specifies three epistles of John, and he includes Revelation. Why, then, would he forge a passage in Origen that doesn't number the Johannine letters and leaves out Revelation?

Furthermore, Origen's Homilies On Joshua repeatedly suggest the canonicity of Old Testament books that Rufinus rejected (HOJ, 116, 194, 206). Did Rufinus alter the text to support his New Testament canon, but retain contradictions of his Old Testament?

Both the internal and the external evidence are against the notion that the canon in this passage of Origen came from Rufinus. It seems to be Origen's canon.

2. Origen may not be referring to a canon of scripture.

It's possible that Origen was doing something other than referring to the canon. He may have been referring to ecclesiastical books, without any intention of suggesting that all of them are scripture. But how likely is it that his reference to something other than a canon of scripture would happen to align with our twenty-seven-book canon? And it's more natural to see the passage as addressing books of scripture rather than a mixture of scripture and non-scripture.

Origen says that the books he's referred to "cast down the walls of Jericho and all the devices of idolatry and dogmas of philosophers, all the way to the foundations" (HOJ, 75). He's addressing books that are authoritative and sufficient to accomplish "all" and go to "the foundations". He opens this section of the homily by saying that these books are apostolic and instigated by Jesus Christ. That sounds like a canon of scripture.

Though Origen doesn't directly say that he's addressing the canon, the nearby context gives us a further indication that he was doing so. In section 2 of the same homily (the apparent canon is in section 1), Origen associates the trumpets of section 1 with books of scripture:

"Make for yourself trumpets hammered thin from the Holy Scriptures. From there bring forth perceptions; from there, discourses; for, on that account, they are called trumpets hammered thin." (HOJ, 76-77)

3. The canon Origen refers to in this passage is inconsistent with his comments on the canon elsewhere.

Origen did sometimes comment that books like 2 Peter and 2 and 3 John were doubted by some people. We have to distinguish, though, between what Origen doubted and his references to the doubts of others.

And he sometimes changed his beliefs, as other ancient Christians did, including on canonical issues. B.F. Westcott cites the example of Origen's inconsistency on who wrote Hebrews (GSN, n. 3 on 370).

As noted above, Origen's Homilies On Joshua were composed near the end of his life (HOJ, 19 and n. 5 on 24). Everett Ferguson explains that it's important to take note of the timing of the writings of Origen that are cited in canonical discussions (TCD, 319 and n. 104 on 319). The works of Origen that were composed in the last years of his life treat the disputed books as scripture.

Aside from 2 and 3 John, which I'll discuss below, all of the disputed books of the New Testament are cited as scripture outside of this passage in Origen that I'm addressing. He cites Hebrews as scripture (HOJ, 100). He refers to James as if it's scripture (HOJ, 112). He does the same with 2 Peter (HOJ, n. 5 on 75). And Jude (Commentary On Matthew, 10:17). In one passage, he refers to the authors of the New Testament, and his inclusion of James and Jude implies his acceptance of the two books that bear their names as canonical:

"Isaac, therefore, digs also new wells, nay rather Isaac’s servants dig them. Isaac’s servants are Matthew, Mark, Luke, John; his servants are Peter, James, Jude; the apostle Paul is his servant. These all dig the wells of the New Testament." (source)

Origen also refers to Revelation as scripture, as I'll discuss below. Thus, even if we set aside this passage in Homilies On Joshua for the moment, we have references in Origen to every book of the New Testament as scripture, except 2 and 3 John. Would those who speculate that Rufinus altered the passage in Homilies On Joshua also speculate that he altered these other passages in Origen? As I said above, the more passages there are that have to be speculatively dismissed as forgeries, the less likely the forgery theory becomes.

4. How do we know that the epistles of John are to be numbered at three? Origen doesn't give a number.

As I've explained above, these objections tend to pull in different directions. Yes, Origen doesn't number the Johannine epistles. But we would expect Rufinus to number them, if he had altered the text so as to make Origen seem to support his canon. It would make more sense for those who want to raise these objections I'm addressing to be selective in which ones they raise rather than using all of them.

Though Origen doesn't number the letters of John, he does use the plural. He's referring to at least two. I'm not aware of any early source who refers to four or more epistles of John, and we know that Origen refers elsewhere to three that are attributed to the apostle (Eusebius, Church History, 6:25). Thus, Origen is referring to either two or three in the passage under consideration.

Eusebius, writing about fifty years after Origen, includes 2 and 3 John among the epistles that are used in most churches, but are disputed by some (Church History, 2:23). It seems that 2 and 3 John were accepted by a majority, even though a minority disputed them. The apparent majority acceptance of the two letters adds weight to the notion that Origen accepted them.

Jerome writes (Lives Of Illustrious Men, 18) that "many" attribute 2 and 3 John to another early church leader named John, not the son of Zebedee. Notice that even these people who are disputing authorship by the apostle John don't deny that the two letters have the same author. It's doubtful that these people would have accepted one epistle while rejecting the other.

It's also unlikely on internal grounds that one of the two letters would be accepted while the other is rejected. The three Johannine letters have many similarities in vocabulary and theme (INT, 674), and 2 and 3 John both identify their author as "the elder" (2 John 1, 3 John 1). The internal evidence suggests common authorship, which adds further weight to the conclusion that Origen accepted both 2 and 3 John, not just one of them.

Of all the sources who commented on the canonicity of 2 and 3 John, the one who had the closest relationship with Origen probably is Dionysius of Alexandria. He was a contemporary of Origen, was bishop in a city Origen had lived in for much of his life, and was a supporter of Origen in times of controversy and persecution (WHO, 22 and n. 124 on 22). Dionysius refers to 2 and 3 John as genuine works of the apostle (Eusebius, Church History, 7:25). Origen may not have agreed with Dionysius on this issue. But Dionysius' view of 2 and 3 John, in light of the close relationship between the two men, does add more weight to the conclusion that Origen held the same view.

I agree with Jeffrey Pettis' conclusion:

"Origen attributes the authorship of the Johannine New Testament writings (the Fourth Gospel, the letters of John, and the book of Revelation) to the apostle John, the son of Zebedee (ComJn 1.1, 6; 5.3; HomGn 7.4)." (WHO, 134)

5. As Bruce Metzger observes (CNT, 139), most manuscripts of Origen's Homilies On Joshua don't include a reference to Revelation in this passage. Why should we think that Origen included Revelation in his canon?

It should be noted that even if we excluded the reference to Revelation, it would still be highly significant that a twenty-six-book canon, so close to our canon, was advocated by Origen.

The minority of manuscripts might be correct. But even if they aren't, I think it's probable that Revelation was left out unintentionally. These homilies were delivered orally by Origen and written by others. Origen may have unintentionally left Revelation out. Or the scribes who wrote the homily may have made a mistake.

Something similar occurred with a list of the Jewish Old Testament canon in Origen. Eusebius cites a list of the Jewish Old Testament in the writings of Origen, but leaves out the Minor Prophets (Church History, 6:25). We know, from other sources, that both the ancient Jews and Origen accepted the Minor Prophets as part of the Old Testament. They must have been left out of this particular listing by mistake. The same sort of thing may have occurred with Revelation in Origen's Homilies On Joshua 7:1.

As I've discussed elsewhere, Revelation was widely accepted in the early centuries of church history. Origen repeatedly refers to it as a genuine work of the apostle John (WHO, 134) and repeatedly treats it as scripture, including in works written around the same time as his Homilies On Joshua (HOL, 54, 100-101; Against Celsus, 6:23). Most significantly, he treats Revelation as scripture elsewhere within the Homilies On Joshua (HOJ, n. 71 on 36, n. 38 on 58, n. 44 on 66, n. 27 on 129, 176). He most likely intended to include Revelation in his canon in Homilies On Joshua 7:1.

I conclude, then, that Origen is the earliest extant source to advocate the twenty-seven-book New Testament canon. But even if it would be argued that he held a twenty-five-book or twenty-six-book canon instead, the similarity to our canon would be highly significant. If the twenty-seven-book canon isn't found in Origen's extant writings, it at least seems to have been held by some other sources predating Athanasius' Festal Letter 39, for reasons explained above.