The Textual Reliability of the New Testament

I was interviewed by Nick Peters on his Deeper Waters podcast today. Two-hour, live program. He’s going to interview Craig Blomberg next week about his new book. Nick is doing a terrific ministry. I encourage you to give a listen:

http://www.blogtalkradio.com/grok558/2014/04/19/the-textual-reliability-of-the-new-testament

Review of Trobisch’s User’s Guide to the Nestle-Aland 28

users_guide

Review of David Trobisch, A User’s Guide to the Nestle-Aland 28 Greek New Testament,
SBLTC 9 (Atlanta: Society of Biblical Literature, 2013). Pp. viii + 69; $25.95.

The much-anticipated publication of the Nestle-Aland Novum Testamentum Graece, 28th edition, in December 2012, instantly created a need for a user’s guide similar to what Kurt Aland and Barbara Aland’s The Text of the New Testament, 2nd edition (Eerdmans, 1989), pp. 232–60 and passim, did for the Nestle-Aland 26th edition. David Trobisch answered the call with his User’s Guide to the 28th edition, which appeared in November of 2013.

This User’s Guide however, is significantly different from the material in Aland-Aland’s Text. Whereas the latter is a scholarly introduction to (and unashamedly a promotion of) the NA26, Trobisch’s User’s Guide is significantly simpler and has only 14 pages devoted to the scholarly use of this handbook edition of the NT. The User’s Guide has three chapters (1–54) and three sections of supporting material at the end (55–69). The chapters progress in intended readership from those who have had little or no Greek (chapter 1: “Structure and Intention of the Edition,” 1–25), to graduate students who have learned Greek and have some comprehension of biblical studies (chapter 2: “Exercises and Learning Aids,” 27–39), to a brief chapter intended for use by “researchers and teachers who interpret the New Testament professionally” (viii), presumably including professors and advanced students (chapter 3: “NA28 as an Edition for Scholars,” 41–54).

Although the second chapter is useful for students, the rationale for the first chapter is puzzling. Why would someone without knowledge of Greek want to use a Greek text at all, especially one as concise (due to the myriad abbreviations, sigla, etc.) and scholarly as the Nestle-Aland? And this being the longest of the three chapters, complete with the Greek alphabet, diphthongs, and other elementary material needed to pronounce ancient Greek, it seems to be a waste of space to some degree. Even in this introductory chapter, Trobisch got some facts wrong. For example, he says that γχ is pronounced ‘nch’ as in ‘anchovies’ (9); the text of the NA28 was produced by “an international editorial committee” (2 [italics added]; see also 49), when the title page indicates only that the Münster institute produced this particular edition; and the canon of the shorter reading or lectio brevior “only applies to two readings that are superficially combined” (24), when the consensus among textual critics is that this rule applies to those variants that have more words than the alternative, whether they are a combination of older readings or not (cf. the variants in John 3:13 and Rom 8:1, for example). Nevertheless, some of the material in the first chapter is helpful for students of Greek. I would recommend eliminating this chapter and combining the best features with what is already in chapter 2.

The second chapter includes helpful information about some of the changes between NA27 and NA28, including the dropping of consistently cited witnesses of the second order, how to use the distinct apparatus for the Catholic Epistles, and a discussion on the Eusebian Canons for the Gospels. On this last item, it should be noted that the Nestle-Aland tradition continues to list the numbers in the Canons as Arabic and Roman numbers. Although this is useful as a tool for the modern student in comparing the Gospels, it is unhelpful for those who spend time on the actual manuscripts, since the Eusebian Canons are found in manuscripts entirely by Greek letters (see https://danielbwallace.com/2014/04/13/conversion-table-for-the-eusebian-canons to download the PDF of a conversion table). This chapter takes the student through the NA28 Introduction, Apparatus, marginalia, and various other features of the book, with exercises sprinkled throughout.

Chapter 3 is a useful introduction to a behind-the-scenes look at the decisions made in Münster concerning the format, text-critical decisions and approach, and differences from the previous edition of the Nestle-Aland text. But Trobisch overstates things when he calls this new edition a “thoroughly revised edition” (vii). To be sure, the apparatus has been thoroughly updated, but the only textual differences are in the Catholic Epistles. Trobisch makes both commendations and criticisms of the 28th edition. In the first section which systematically goes through differences between this and NA27, some of the negative features of the 28th come to light—even though Trobisch explicitly addresses limitations of this new edition in the second section, “Limitations of the NA28.”

Gone are any explicit conjectural emendations, whereas the NA27 listed over 100 of them (one of which was followed [Acts 16:12], though both Bruce Metzger and Kurt Aland disagreed with the rest of the committee), and NA28 adds one more to the text (2 Peter 3:10). (At the same time, neither of the variants in these two passages is a true conjecture since there are versions that have these readings. Bruce Metzger and Bart Ehrman, The Text of the New Testament, 4th ed. [Oxford: OUP, 2005] 230, implicitly define a conjecture as having no support in Greek manuscripts, versions, or fathers: the need for conjectural emendation for the New Testament is “reduced to the smallest dimensions” because “the amount of evidence for the text of the New Testament, whether derived from manuscripts, early versions, or patristic quotations, is so much greater than that available for any ancient classical author…”)

NA28 also eliminated the useful subscriptions for the NT books found in previous editions, a most unfortunate decision. They have however retained the inscriptions, though Trobisch says that these, too, got the ax (43).

The number of witnesses cited in the apparatus is significantly reduced, and any comparison with previous editions of the Greek NT by Tischendorf, Westcott-Hort, von Soden, and others is eliminated.

The lack of such valuable features means that students and scholars will need to continue using their NA27 in conjunction with NA28. Trobisch notes that 33 textual changes occurred in the Catholic Epistles (44), though there are actually 34 (see NA28, 50*–51* for the list). A brief discussion of the sea-change in Münster from the “local-genealogical method” (which Barbara Aland once told me was not within the bounds of reasoned eclecticism) to the “Coherence-Based Genealogical Method” or CBGM concludes the chapter.

A final criticism of this booklet is that although the author provides links to several sites which host images of NT manuscripts, he overlooks the website for the Center for the Study of New Testament Manuscripts (www.csntm.org), which has one of the largest collections of high-resolution digital images of Greek NT manuscripts on the Internet, most of which have been photographed by CSNTM in the last twelve years. Included on this site are images of the Chester Beatty papyri, which CSNTM digitized in the summer of 2013, working with the papyri at the Chester Beatty Library in Dublin.

In sum, I anticipated that this work would be useful for students learning the ropes of NT textual criticism, but the gaps, errata, and proportion leave me somewhat disappointed. Even though there are many helpful features, the work is overall quite uneven. I hope that a second edition which corrects these deficiencies will soon be forthcoming (some of these deficiencies have been corrected in the second German edition of this book), since such a volume is needed for anyone using the Nestle-Aland 28th edition of Novum Testamentum Graece.

Daniel B. Wallace

Conversion Table for the Eusebian Canons

For several years now, the staff at the Center for the Study of New Testament Manuscripts (www.csntm.org) have utilized the Eusebian Canons to quickly find their place in Gospel manuscripts. These Canons are found in the inner margin of the Nestle-Aland, Novum Testamentum Graece, 28th edition (as well as earlier editions). They are written in Arabic numbers over Roman numerals. As helpful as this is for those working in the printed text of the Gospels, it is difficult to use when examining manuscripts precisely because one has to convert on the fly numbers to letters if he or she needs to locate where they are in the text efficiently. Of course, determining what passage one is reading is usually fairly easy by simply keying in a few Greek words in sequence and checking what the manuscript says against a printed text in a Bible software program. But at times this can be tricky. For example, if the text is difficult to read or has variant spellings, finding one’s place may require several attempts on the computer. Codex 0322, a two-leaf palimpsest that CSNTM discovered in 2004, yields only a few letters on each page to the eye. But the Eusebian Canons are still completely intact and guided us to recognize the text as Mark 3.17–4.1; and 6.10–22. With that to guide us we were able to discern two or three ‘Western’ readings in this majuscule.

For others who are interested in the Eusebian Canons, I am attaching the document (Eusebian Canons conversion table) that we use when examining Gospel manuscripts. There are three columns for each canon: arabic number, Greek letter, and scriptural reference. I am sure a few errata have made their way into this conversion table, and would invite corrections so that we can improve on it.

Can We Still Believe the Bible?

Can-We-Still-Believe-the-Bible-200x300

Craig Blomberg, Distinguished Professor of New Testament at Denver Seminary, has written another outstanding volume. Blomberg is a committed evangelical, but not one with a closed mind. As he says in his preface about the environment of Denver Seminary (quoting Vernon Grounds, former president of the school), “Here is no unanchored liberalism—freedom to think without commitment. Here is no encrusted dogmatism—commitment without freedom to think. Here is a vibrant evangelicalism—commitment with freedom to think within the limits laid down by Scripture.” Blomberg’s writings have always emulated this philosophy. His research in the secondary literature is consistently of superb quality, and his discussions of problem passages and issues, especially in the Gospels, is always well informed. Rather than clutter the narrative with documentation, Blomberg has wisely used endnotes instead of footnotes (though I personally prefer footnotes, I understand that most readers see them as a distraction). This book has nearly 50 pages of endnotes, almost one fifth of the whole book. Blomberg knows his stuff.

I received a prepublication draft of the book, Can We Still Believe the Bible?, and was asked to blog about it. More specifically, I was asked to blog about the first chapter, “Aren’t the Copies of the Bible Hopelessly Corrupt?”

This first chapter addresses the number one apologetic issue of our time—Did the scribes get it right when they copied the scriptures? No longer is the main attack on the Christian faith framed in the question, Is the Bible true? It is now the preliminary question, How do you even know that the Bible you have in your hands accurately represents the original documents? History, as many ancients conceived of it, is circular rather than linear. In this case, that’s true: “Hath God said?” is the original attack on God’s word, way back in the Garden. We’ve come full circle once again.

In this chapter, Blomberg rightfully shows the misrepresentations of the situation by Bart Ehrman, in his book, Misquoting Jesus. For example, of the approximately 400,000 textual variants among New Testament manuscripts, many who read Misquoting Jesus get the impression that this one datum is enough to destroy the Christian faith. But the reality is that less than one percent of all variants are both meaningful and viable. And even Ehrman himself has admitted that no cardinal doctrine is jeopardized by these variants.

Blomberg lays out a compelling argument, with much nuance, about the reliability of the NT and OT manuscripts. His chapter on the text of the Bible is organized as follows:

  • Misleading the Masses
  • The Truth about Variants (New Testament, Old Testament)
  • Did Originals Originally Exist?
  • Comparative Data
  • Avoiding the Opposite Extreme
  • Conclusion

In the opening section, the author takes on Bart Ehrman’s wildly popular book, Misquoting Jesus. In characteristic fashion, Blomberg critiques both what Ehrman does and doesn’t say, doing all with wisdom and wit. He points out, for example, that virtually nothing in Misquoting Jesus is new to biblical scholars—both liberal and evangelical, and all stripes in between. Non-scholars, especially atheists and Muslim apologists, latched onto the book and made preposterous claims that lay Christians were unprepared for. Ignorance, in this case, is not bliss. Earlier in the chapter when Blomberg mentioned that there are as many as 400,000 textual variants among the manuscripts, he bemoans: “It is depressing to see how many people, believers and unbelievers alike, discover a statistic like this number of variants and ask no further questions. The skeptics sit back with smug satisfaction, while believers are aghast and wonder if they should give up their faith. Is the level of education and analytic thinking in our world today genuinely this low?” (13).

He then discusses the two major textual problems that Ehrman zeroes in on: Mark 16.9–20 and John 7.53–8.11. He makes the insightful comment that the probable inauthenticity of these passages is news to laypeople because they tend not to read the marginal notes in their Bibles and because “more and more people are reading the Bible in electronic form, and many electronic versions of the Bible don’t even include such notes” (15).

In passing, I’d like to make three comments about the ending of Mark’s Gospel:

  1. Blomberg says that there is no passage elsewhere in Mark that has nearly as many variants as 16.9–20 (p. 19). This may be true, but he doesn’t document the point. It has often been said about the pericope adulterae, but I’m not sure about the ending of Mark.
  2. Blomberg cites Travis Williams, “Bringing Method to the Madness: Examining the Style of the Longer Ending of Mark,” Bulletin of Biblical Review 20 (2010), to the effect that “the style of writing in the Greek significantly differs from the rest of Mark’s Gospel” (19). This article was first read at the southwest regional Evangelical Theological Society meeting shortly after Travis was an intern of mine. He did an outstanding job on the paper; hence, its publication in BBR. Since this publication another student of mine, Greg Sapaugh, wrote his doctoral dissertation at Dallas Seminary on “An Appraisal of the Intrinsic Probability of the Longer Endings of the Gospel of Mark” (2012). Both scholars came to the same conclusion: the language of Mark 16.9–20 is anomalous and almost surely was not written by the person who wrote Mark 1.1–16.8.
  3. When discussing whether the real ending of Mark’s Gospel was lost, Blomberg says, “The open end of a scroll was the most vulnerable part of a manuscript for damage; perhaps Mark literally got ‘ripped off’!” (20). He goes on to argue against this, seeing that Mark’s intention was to conclude his Gospel at v. 8. Although Blomberg is right to note that Mark was almost certainly written on a roll instead of a codex, he doesn’t mention the great difficulty that this poses for those who think that the real ending was lost. Ancient rolls were almost always rolled up for the next reader. Assuming that to be the case for Mark, the ending of the Gospel would be the most protected part.

Blomberg also highlights many of other major passages that Ehrman wrestles with, such as Mark 1.41, Heb 2.9, and Luke 22.43–44. In the process, he notes that of the two standard Greek New Testaments in use today—the Nestle-Aland text and the United Bible Societies’ text—the latter includes only the most important textual problems (1438 of them) and a perusal of these textual problems reveals that “the only disputed passages involving more than two verses in length” are Mark 16.9–20 and John 7.53–8.11 (18).

The author takes pains to introduce the discipline of textual criticism to lay readers. He discusses some of the major textual problems (or, rather, those with much emotional baggage because of their long history in the printed Bible) in the NT (including Matt 5.22; 6.13; Acts 8.37; and 1 John 5.7–8), patiently going through the evidence, showing that the wording in the KJV is spurious because it is poorly attested in the manuscript evidence and/or has strong internal evidence against it.

The question is then raised, Why are these passages (including the two 12-verse texts mentioned earlier, along with Luke 22.43–44) sometimes printed in our modern translations? Blomberg gives a nuanced answer, but the bottom line (in my view) is this: Translations follow a tradition of timidity. My own examination of over 75 translations in a dozen different languages reveals the same monotonous story: Translators keep these passages in the text of their Bibles because to do otherwise might upset some uninformed Christians. But Ehrman has let the cat out of the bag. Just as Edward Gibbon’s Decline and Fall of the Roman Empire pointedly athetized the Trinitarian formula in 1 John 5.7–8 over two centuries ago, so Ehrman has done the same for Mark 16 and John 8. When Gibbon wrote this note in his six-volume work, it scandalized the British public. A hundred years later, the Comma Johanneum did not even show up as a marginal note in the Revised Version of 1881. It is time for us to relegate these likely inauthentic texts to the footnotes. Otherwise, we will continue to placate uninformed believers, setting them up for a Chicken Little experience when they read books like Misquoting Jesus. Sadly, tens of thousands of college students, raised in a Christian home, have abandoned the faith because of fear of embarrassment over these issues, especially due to Misquoting Jesus. In recent years, it has been estimated that over 60% of kids coming from Christian homes abandon the faith by the time they get done with college. It is time for pastors and other Christian leaders to educate the masses about the reality of the transmission of the Bible. If we don’t, the fallout will only get worse.

Blomberg also discusses more routine textual variants (what he calls “ordinary and uninteresting,” the latter description of which I would disagree with :-)), giving a glimpse to the discipline of NT exegesis to outsiders. (At least he does correct this a bit later: “The vast majority of textual variants are wholly uninteresting except to specialists [italics mine].”) Almost anyone who has spent time with the textual apparatus is amazed at how little the vast majority of variants affect the meaning of the text.

In his treatment of the gap that exists between the originals and the early copies, he argues that “One may fantasize about all kinds of wild changes being introduced between the first, complete written form of a given book and the oldest copy we actually have, but it will be just that—fantasy…” (35). I’d like to offer some supplemental reasoning for why this is almost certainly true: Against the supposition that the older the manuscripts that are discovered, the more likely it is that we will find new, authentic readings, we can simply look at the last 130+ years. That’s when all but one of the NT papyri (our oldest manuscripts) have been discovered. How many earth-shaking, new readings have commended themselves to scholars as autographic among these 128 NT papyri? None, zero, zilch. Not a single new reading since the discovery of the NT papyri has been viewed by textual scholars as authentic. Does this mean that the papyri are worthless? Not at all. Rather, they usually confirm readings that scholars already thought were authentic. Now, with even earlier evidence found in the papyri, the arguments are stronger. This shows that the methods of textual scholars since the work of Westcott and Hort (1881–1882) are, in broad strokes and in many particulars, on target. But, with regard to Blomberg’s point, it also shows that if history is any indication, it would be foolish to think that any not-yet discovered readings will some day grace the text of our critical Greek New Testaments instead of finding a place in the apparatus of also-rans.

In comparing the copies of the NT with other ancient Greco-Roman literature, Blomberg argues well that Christians need not feel embarrassed about the relatively small gaps between the originals and the earliest copies (most NT books have copies within a century of the completion of the NT), since the gaps for other literature are far greater (hundreds of years). Further, the differences between the copies for, say, the apocryphal literature is remarkably greater than for the NT copies. He mentions as an illustration the Coptic text of the Gospel of Thomas and the three Greek fragments (though he incorrectly dates them to the second century [36]), citing Tim Ricchuiti’s excellent study (in Revisiting the Corruption of the New Testament).

In his last section before the conclusion, “Avoiding the Opposite Extreme” (37–40), Blomberg offers some excellent insights about the ludicrousness of a perpetual miracle of exact copying of the text (akin to the argument that Muslims use about the Qur’an and some KJV advocates come close to arguing about the TR): “But think of just what kind of miracle this would need to be for it really to have occurred. Not only would God have superintended the process of a select group of biblical authors penning their documents so that their words reflected precisely what God wanted to have written; God would also have needed to intervene in the lives of all the tens of thousands of copyists over the centuries to ensure that not one of them ever introduced a single change to the texts they were reproducing” (39). He goes on to expound on this topic, with remarkable clarity and logic. Definitely a good read.

Errata

There are a few errors of fact and misleading statements in Blomberg’s new release.

  1. Page 15: The author says that Ehrman’s Orthodox Corruption of Scripture, on which Misquoting Jesus was based, was Ehrman’s doctoral dissertation. Actually, Ehrman wrote his dissertation on the text of Didymus the Blind. Orthodox Corruption is Ehrman’s most influential scholarly work, but it was not his dissertation.
  2. Page 16: It would be “extraordinarily unlikely that we shall ever again find variants that are not already known.” Actually, it is very likely that we will find variants in almost every new MS discovered. They are almost always so trivial that they would not warrant mention in an apparatus, however. What is unlikely in the extreme is that any of these MSS will have new readings that convince scholars of their authenticity.
  3. Page 24: The textual problem in Rom 5.1 is discussed; Blomberg notes that the difference between ‘we have faith’ and ‘let us have faith’ is one letter in Greek: it is either an omicron or an omega. He says that the forms would have been similar, but gives the capital letters (Ο, Ω) instead of the majuscule letters (ο, ω), which is what the oldest MSS are written in.
  4. Page 27: The author suggests that every single second- and third-century papyrus of the NT was “written with the very careful handwriting of an experienced scribe…” This, however, is not true. The penman of P75, for example, was probably not a professional scribe (according to E. C. Colwell), although he produced a very careful text, painstakingly writing out one to two letters at a time. Further, even later scribes were definitely not professional. For example, P10, P93, and P99 were either done for private use or were perhaps schoolboy exercises. I pointed out in one of my debates with Ehrman (SMU, 2011; DVD available here) that a comparison of P66 and P75 reveals that the more professional scribe (P66) produced the less careful text. Zachary Cole, who is currently working on his doctorate in NT textual criticism at Edinburgh University, wrote his master’s thesis at Dallas Seminary (2012) on “Scribal Hands of Early New Testament Manuscripts.” This thesis was in response to Ehrman’s claims that the earliest scribes were not professional and therefore their text was not carefully produced. Several of the second- and third-century papyri were judged to be less than professionally done, including especially P9, P18, P24, P78, and P98, but also including as many as 27 other papyri. And Cole concluded that all this is irrelevant, since the training of the scribe is no necessary indicator of the quality of his text.
  5. Page 27: “no orthodox doctrine or ethical practice of Christianity depends solely on any disputed wording.” I would word this a bit differently. We can definitely say that no cardinal doctrine depends on any disputed wording, but I think there are some places in which less central teachings—both in terms of orthodoxy and orthopraxy—are based on texts that are disputed. For example, whether exorcists casting out particularly pesky demons need to pray and fast depends on a variant in Mark 9.29, and the particulars of the role of women in the church may depend, in part, on 1 Cor 14.34–35 (a passage that, although found in all MSS, is disputed by some scholars).
  6. Page 34: “the original copy [sic] of a biblical book would most likely have been used to make countless new copies over a period of several centuries…” Blomberg cites the important study by George Houston on the longevity of papyrus documents, which Craig Evans exploits to the effect that the original documents would have perhaps lasted several centuries. I think that Evans may be arguing his case a bit too strongly, especially in light of patristic evidence to the contrary. We do have two or three ancient patristic statements to the effect that the autographs still existed into the second or third centuries, but they have generally been regarded as ahistorical comments without substance behind them. Nevertheless, an important point to consider is that these ancient writers demonstrate, from a very early period, a desire on the part of the ancient church to seek out the oldest MSS to establish the wording of the original. And Blomberg is quite right that the ancient scribes surely would have copied the autographs multiple times, thus disseminating direct copies spanning a period of more than one or two generations.
  7. Page 37: Gutenberg’s printing press is dated c. 1440; it should be dated c. 1454.
  8. Page 38: Fifteenth-century Catholic reformer, Erasmus: sixteenth century is meant.
  9. Pages 16–17 has what looks to be the most egregious error: “Although Ehrman doesn’t total all the numbers, Wallace does, and the result is that those 400,000 variants, if there are that many, are spread across more than 25,000 manuscripts in Greek or other ancient languages.” In the next paragraph he asserts: “This is an average of only 16 variants per manuscript… Nor are the variants spread evenly across a given text; instead, they tend to cluster in places where some kind of ambiguity has stimulated them. Paul Wegner estimates that only 6 percent of the New Testament and 10 percent of the Old Testament contain the vast majority of these clusters.”I think Blomberg means that there is an average of 16 unique variants per MS. That would be essentially true, though we really should restrict the count to Greek MSS since the translations have too many problems to be able to discern at this stage whether the wording is a true variant from the Greek or simply a looser translation. On his use of Wegner: I’m out of the country right now and can’t look at my copy of Wegner. But it is simply not true that only 6% of the NT contains “the vast majority of these clusters.” I’m not sure what Blomberg is trying to say here. Perhaps he meant that the major textual problems of the NT are found in only 6% of the text. That may well be the case, but in this case the number seems too high.

These are, for the most part, rather niggling criticisms. Overall, this chapter is an excellent corrective to the extreme skepticism of Bart Ehrman and those who have followed in his train. It is well researched, clearly written, and deserves to have a wide reception among believers today, as does the book of which it is a part. One can hope that pastors and church leaders will wake up to the fact that we are losing the intellectual battle for the millennials, and we have only ourselves to blame. Bringing spiritual grace and academic rigor to the table is needed, and Blomberg is one of the evangelical gatekeepers leading the way.

What Does “We are God’s fellow-workers” in 1 Corinthians 3.9 Really Mean?

Translations and Commentaries

The King James Version in 1 Cor 3.9 reads, “we are labourers together with God…” This unambiguously suggests that Paul and Apollos were considered in some sense on the same level with God. Of course, ‘in some sense’ covers a multitude of possibilities, but there nevertheless seems to be an underlying tone of synergism and mutual credit. (A similar translation is in the French Nouvelle Version2: “nous sommes ouvriers avec Dieu,” and in La Sacra Bibbia: “Noi siamo infatti collaboratori di Dio.”)

Most modern translations take a more neutral stance, translating 1 Cor 3.9a as “we are God’s fellow workers” (ASV [‘fellow-workers’], RSV, NASB and NASB 1995, NKJV, ESV, NIV), “we are God’s coworkers (HCSB, TNIV, and NAB2 [‘co-workers’; 2010]), “we do share in God’s work” (NJB), “wir sind Gottes Mitarbeiter” (Luther 1985), or “nosotros somos colaboradores de Dios” (Reina Valera2).

But some translations take a different interpretation. The NET Bible has, “we are coworkers belonging to God”; the REB says, “we are fellow-workers in God’s service”; the NIV 2011 reads, “we are coworkers in God’s service”; the TEV has, “we are partners together working for God”; the NRSV reads, “we are God’s servants, working together”; and “we are both God’s workers” is in the NLT2.

Thus, we see in the translations three different views: (1) Paul and Apollos are co-workers with God; (2) the statement is ambiguous, though tending toward the first view; and (3) Paul and Apollos are co-workers with each other in service to God.

Commentaries overwhelmingly fall into the third group, with some supporting the first interpretation. In the first group belong Theodore of Mopsuestia, Calvin, and Robertson and Plummer (ICC). In the third are Findlay (Expositors), Moffatt (Moffatt NTC), Barrett (BNT), Héring, Fee (NICNT), Kistemaker, Furnish (JBL 80 [1961]), Wolff (THKNT), Horsley (Abingdon), Collins (SP), Thiselton (NIGTC), and Keener (NCBC).

Significantly, if we were to chart out the translations and commentaries chronologically, we would see a tendency toward the neutral view (translations) and especially toward the third view (commentaries).

Table of Interpretations and Translations of 1 Cor 3.9

co-workers

with God

God’s co-workers

co-workers

for God

Theodore of Mopsuestia (d. 428)

Calvin (1546)

KJV (1611, 1769)

Findlay (1900)

ASV (1901)

Robertson and Plummer (19142)

Moffatt (1938)

RSV (1952, 1971; slight variation)

Reina Valera (19602)

NASB (1960–1977)

Furnish (1961)

Héring (1962)

TEV (1968–1992)

Barrett (19712)

Nouvelle Version2 (1978)

NKJV (1982)

NIV (1984)

Luther2 (1985)

Fee (1987)

NRSV (1989)

REB (1989)

NJB (1990)

Kistemaker (1993)

La Sacra Bibbia (1994)

NASB 1995

Wolff (1996)

Horsley (1998)

HCSB (1999)

Collins (1999)

Thiselton (2000)

ESV (2001)

TNIV (2001)

NLT2 (2004)

NET (2005)

Keener (2005)

NAB2 (2010)

Totals

6

13

19

Thus, of the 38 works perused, half see Paul and Apollos as co-servants only with each other, both working for God. The latest authority that sees them as co-workers with God was in 1978, a French translation of the Bible. The latest source for this view in English was the ICC commentary by Robertson and Plummer in 1914. From 1938 on, every commentary consulted regarded Paul and Apollos to be in the service of God. To be sure, this is hardly a representative sampling; it represents only the works I have at hand in my personal library. I understand that Weiss (1910), Davies (1972), and H. D. Betz (1986) all adopted the first view (according to John G. Lewis [DPhil, Oxford, 2003] who also takes this position), but I have not verified it yet. See also Victor Paul Furnish, “Fellow Workers in God’s Service,” JBL 80 (1961) 364 (whole article, 364–70), who adds Lightfoot and Wendland as supporters of the first interpretation, and Heinrici and Parry as supporters of the third.

Issues in Greek Grammar

Fee, Thiselton, and Furnish give some of the strongest arguments for taking the genitive θεοῦ as possessive or purpose (‘for God’) here, which center on the flow of argument in the context. But not one of the works examined gives any grammatical parallels to this understanding of the text. The normal Greek rule is that a genitive attached to a συν-prefixed noun/substantive will be a genitive of association, and thus translated ‘with.’ In order for the third view to gain some traction, at least some clear examples need to be produced of a genitive dependent on a συν-prefixed noun which is other than associative. Otherwise, it is just wishful thinking. So, are there any instances of such a genitive?

(Excursus: The Role of Syntax in Exegesis)

The role that Greek syntax plays in exegesis is often neglected in exegetical literature. I would say, in fact, that for most exegetes, any meaningful discussion of syntax seems to be wrenched out of them, if discussed at all. Too frequently, commentators will appeal to the “context,” almost as though just uttering that word magically settles all issues. It’s as if the first one to utter this incantation wins the argument! But context is not a given; it must be construed. And if no syntactical parallels can be found to support an interpretation which may seem probable from the context, then the interpretation cannot be certain. Exegetes have long seen this when it comes to lexical studies. They know they can’t simply invent a meaning for a word that it never has elsewhere just because the context seems to favor it. They know they are on much more solid footing if they can find some parallels lexically, especially if they are in Hellenistic Greek. It would seem that syntax should play the same role, but curiously it almost never does. In this brief paper, my purpose is to illustrate, with parallels as close as I can determine, of συν-prefixed substantives with non-associative genitive modifiers. Only if such parallels can be produced can one then bring in the contextual arguments for 1 Cor 3.9.

Genitive of Association in Hellenistic Greek

New Testament

(All parallel references are in bold for ease in locating them.) In Rom 11.17, Paul speaks of the Gentiles as συγκοινωνὸς τῆς ῥίζης (“fellow-partakers of the root”). The Gentiles are seen as fellow-partaker with their Jewish counterparts of the root. The genitive is objective (‘partake of the root’). In 1 Cor 1.20 we read ποῦ συζητητὴς τοῦ αἰῶνος τούτου; (“where is the debater of this age?”). The word ‘debater’ means, loosely, dialogue partner. The genitive is temporal. In 1 Cor 9.23 we read ἵνα συγκοινωνὸς αὐτοῦ γένωμαι (“so that I may become a participant in [the gospel]”). Again, an objective genitive. In 2 Cor 1.24 again the associative genitive is implied (συνεργοί ἐσμεν τῆς χαρᾶς ὑμῶν: “we are fellow-workers [with you] for your joy”). In Eph 3.6 the head noun takes an objective genitive once again: εἶναι τὰ ἔθνη … συμμέτοχα τῆς ἐπαγγελίας (“in order that the Gentiles might become … fellow partakers [with the Jewish believers] of the gospel”). In Acts 21.30 we read that the whole town was in an uproar and ἐγένετο συνδρομὴ τοῦ λαοῦ (lit., “there became a rushing together of the people”). The idea here is most likely subjective (“the people rushed together”). In 1 Peter 3.7 the men are instructed to treat their wives with all due respect because they are “fellow-heirs [with their wives] of the grace of life” (συγκληρονόμοις χάριτος ζωῆς).

In 1 Thess 3.2 we have a very interesting illustration. There Paul declares that Timothy is τὸν ἀδελφὸν ἡμῶν καὶ συνεργὸν τοῦ θεοῦ. This passage, like 1 Cor 3.9, has been understood in two different ways. Some take τοῦ θεοῦ as a genitive of association (“fellow-worker with God”), while others see it as possessive/purpose (“fellow-worker [with us], belonging to God/for God”). Many important witnesses read διάκονον τοῦ θεοῦ (“servant of God”) instead of συνεργὸν τοῦ θεοῦ; others have διάκονον τοῦ θεοῦ καὶ συνεργὸν ἡμῶν (“servant of God and our fellow-worker”); both of these would eliminate the possibility of θεοῦ being an associative genitive. Assuming that the Nestle-Aland28’s text is autographic (in spite of it being poorly attested), the variants may have come about because of a desire to clarify the meaning while simultaneously eliminating one possible interpretation. In spite of the argument of some exegetes, it seems to be saying too much to argue that the variants arose because scribes were offended at the idea that Paul put himself and his colleagues on the same plane with God. An equally plausible interpretation is that here the scribes wanted to clarify that this was not Paul’s meaning, since the context was more ambiguous than 1 Cor 3.9’s context, and the scribes there saw clearly that Paul did not see a synergism between God and men. Since there are no variants listed there in the NA28 apparatus, scribes may well have seen Paul’s meaning here, as well, to be clearly against a divine-associative view.

Now, except for Acts 21.30 and 1 Thess 3.2 (both of which could possibly be construed otherwise), none of these is an illustration of a συν-prefixed noun with a personal dependent genitive. Yet this is precisely what we see in 1 Cor 3.9, which begs the question as to whether θεοῦ γάρ ἐσμεν συνεργοί really can mean “we are fellow-workers [with each other] for God.” But at least what the parallels have shown is that the genitive of association may, at times, be implied rather than stated, even when another genitive is related to the συν-prefixed noun. And this would be in line with seeing θεοῦ in 1 Cor 3.9 as non-associative. But are there any unambiguous illustrations of a συν-prefixed noun with a non-associative personal dependent genitive? Most of the illustrations used in the previous paragraph are all found in my book, Greek Grammar Beyond the Basics: An Exegetical Syntax of the New Testament (130). But more are needed, especially those that are closer parallels.

Consider the following:

Josephus:

Antiquities of the Jews 8.133 (8.5.2): σύνοδον ἀνθρώπων (“a gathering of people”). σύνοδος is translated ‘fellow-traveler’ or ‘traveling companion’ in Ignatius, Ephesians 9.2, but such a meaning will not work here. The genitive ἀνθρώπων probably is a genitive of apposition/content (“a gathering made up of people”).

Antiquities of the Jews 16.351 (16.10.9): συνθῆκαι τῶν δανείων (“covenant [between the two men] concerning the debt”). Here the implied genitive is personal, while the genitive stated is a genitive of reference.

Antiquities of the Jews 17.51 (17.3.1): εἰς τὴν σύνοδον τῆς Ἀντιπάτρου μητρός (“for the meeting [arranged] by Antipater’s mother”). One might normally assume that after σύνοδος the genitive would indicate association. In this instance, however, Antipater’s mother arranged the meeting between others. Similarly, Josephus, Vita 311 (60), and Josephus, Vita 368 (66).

Antiquities of the Jews 19.107 (19.1.14): τοῖς συνωμόταις κινδύνων (“to [his] co-conspirators in the risk”). The implied genitive is personal while the stated genitive is a genitive of reference.

Jewish War 4.148 (4.3.6): συνεργοὺς τῶν ἀσεβημάτων (“fellow-workers for their ungodly acts”). The implied genitive is personal while the stated genitive is a genitive of purpose.

Jewish War 4.240 (4.4.3): τὴν σύνταξιν ὑμῶν (“your army” in the sense of “the army in service to you”).

Philo:

Philo, Somniis 1.193: πρὸς τὸ τῶν φίλων ἔλθῃ συνέδριον (“when he comes into the assembly of friends”).

Philo, Specialibus 1.29: συνεργοὺς τῆς ἀπάτης (“co-workers [with each other] in deception”). Same word as is found in 1 Cor 3.9. The implied genitive is personal and associative, while the stated genitive is reference.

Philo, Contemplativa 40: τὰς κοινὰς συνόδους αὐτῶν (“their common assemblies”). The common assemblies which consists of them; thus, a genitive of apposition or content.

Greek Pseudepigrapha:

Rechabites 11.7: συνηθείᾳ τοῦ γάμου (“companionship [between them] of marriage”). Not personal, but the personal is implied.

Psalms of Solomon 4.1: συνεδρίῳ ὁσίων (“council of holy men”—that is, council comprising holy men, not a council in association with holy men).

Justin Martyr:

1.63 (27): τοῖς Μωσέως συντάγμασι (“the collected writings of Moses”—i.e., Moses wrote these collected writings).

Conclusion

It seems to be sufficiently established that a genitive dependent on a συν-prefixed substantive can indeed be other than a genitive of association in Hellenistic Greek. And this is even found in some instances in which (a) either a genitive of association is not stated or in which (b) the genitive that is present is personal. In the first category are the following texts: Rom 11.17; 1 Cor 1.20 (possibly); 1 Cor 9.23; 2 Cor 1.24; Eph 3.6; 1 Peter 3.7; Josephus, Ant. 16.351; Josephus, Ant. 17.51; Josephus, Ant. 19.107; Josephus, Jewish War 4.148; Philo, Specialibus 1.29; and Rechabites 11.7. (Both 1 Cor 3.9 and 1 Thess 3.2 may belong here, too, but since the former is our target passage and the latter is equally disputed, they should not be counted.) In the second category are Acts 21.30; Josephus, Ant. 8.133; Josephus, Ant. 17.51; Josephus, Jewish War 4.240; Philo, Somniis 1.193; Philo, Contemplativa 40; Psalms of Solomon 4.1; and Justin Martyr 1.63. One text in particular is doubly parallel: Josephus, Ant. 17.51. All of this paves the way for both 1 Cor 3.9 and 1 Thess 3.2 to indicate association between men in the service of God.

Having established the syntactical parallels, we now need to proceed to look at the context of 1 Cor 3.9. Fee succinctly states the argument: “In the Greek text, the emphasis is altogether on God: ‘God’s we are, being fellow workers; God’s field, God’s building, you are.’ Some have suggested that Paul here intends, as the KJV has it, ‘we are laborers together with God.’ But everything in the context speaks against it: the emphatic position of the genitive (‘God’s’) suggests possession, as do the following, equally emphatic, genitives, which are unambiguously possessive; the argument of the whole paragraph emphasizes their unity in fellow labor under God, an argument that would be undercut considerably if he were now emphasizing that they worked with God in Corinth; and finally, these new ‘slogans’ serve as the climax of the whole paragraph, in which the emphasis is decidedly on God’s ownership, not on Paul’s and Apollos’s working with him in Corinth” (G. D. Fee, The First Epistle to the Corinthians, NICNT [Grand Rapids: Eerdmans, 1987] 134).

To sum up, the translation of 1 Cor 3.9 as “we are co-workers belonging to God” (NET) or the like is strongly justified in light of the most probable construal of the context, and is legitimate in light of the syntactical parallels.