This provocative and stimulating monograph, originally a doctoral dissertation done under John Kloppenborg’s supervision, makes a strong, sustained argument that some places in the NT manuscripts (MSS) are so defective that a scholarly conjecture is needed to restore the wording of the original text. This view is not new, though its heyday was centuries ago when the database for the NT text was but a fraction of what it is today.

The book has two main sections: Theory (the need and rationale for conjecture), with three chapters (pp. 3–78), and Practice (pp. 79–184), which focuses on five conjectural emendations in James covered in four chapters (conjectures in 3:1, 4:2, 4:5; and two that the author rejects, in 1:1 and 2:1).

Wettlaufer opens his monograph by defining conjectural emendation as “the act of restoring a given text at points where all extant manuscript evidence appears to be corrupt” (p. 3). He gives no further definition. This is a surprising omission, since the definition of conjecture is key to his entire project. But one person’s conjecture is another person’s minimally attested reading. Wettlaufer never thinks of patristic writings, let alone ancient versions, as part of the material evidence to be included when assessing whether a reading is really a conjecture. Yet later he quotes Bruce Metzger and Bart Ehrman on what conjecture means: “the amount of evidence for the text of the New Testament, whether derived from MSS, early versions, or patristic quotations, is so much greater than that available for any ancient classical author . . .” (p. 16, quoting The Text of the New Testament [4th ed; Oxford: OUP], 230, emphasis added by Wettlaufer).

In the first chapter, the author argues that conjectural emendation is a necessary part of classical textual criticism. This is true, but the dearth of data, especially early data, is the primary reason for it. The average classical Greek author has fewer than 20 copies of his or her writings still in existence. Further, the earliest extant copies for such an author come well over half a millennium after the original was written. But the NT is not like this at all.

The NT boasts of tens of thousands of handwritten copies (in Greek, Latin, Syriac, Coptic, and several other languages) as well as over a million quotations by the church fathers. Nothing in the Greco-Roman world even comes close. Other ancient texts present scholars with a dearth of evidence; NT scholars enjoy an embarrassment of riches. And yet, even with the discovery of the papyri in the last 150 years, not a single newly discovered reading has commended itself to the majority of NT scholars as authentic. The papyri tend to support old readings, but in no place (except perhaps one) have they supplanted the readings that were already known. The fact that the papyri—collectively covering about half the NT—offer virtually no new authentic readings is telling against the possibility that somehow the autographic wording has been lost. If it had been, would we not expect to find more than one place where the papyri and only the papyri got it right?

In his second chapter, the longest in this section, Wettlaufer gives three arguments against conjecture: “some overestimate the witness of the extant manuscript base, some misestimate the influence of faith in textual criticism, while some underestimate the necessity of even trying to restore the text of the New Testament” (p. 14). I agree strongly with his second and third points, but his first—the only positive argument for conjecture—is the weakest.

Here the author says that the primary reason scholars reject conjectural emendation is because of the abundance of MSS (p. 16). But that’s only partially true. He cites Westcott and Hort, yet ignores much of their argument: “in the New Testament the abundance, variety and comparative excellence of the documents confines this task of pure ‘emendation’ within so narrow limits that we may leave it out of sight for the present . . . ” (p. 17, italics added by Wettlaufer). Two sentences later he declares, “Popularity is no necessary indicator of truth.” True—yet that is not what these scholars, or more recent scholars, are saying. It is the abundance, age, quality, and variety of witnesses that argue against the need for conjecture—not just the abundance, and not just Greek MSS.

In this chapter, the author’s best argument, prima facie, is this: “It is surprising, in fact, how often the accepted text actually depends on only a handful of extant manuscripts” (p. 29). Citing a study by Maurice Robinson, Wettlaufer notes that there are 30 places in the NA27 in which “the reading accepted as authentic in that edition was supported by but a single manuscript” (p. 30; Wettlaufer lists only 29). This is a far smaller number than I expected. If there are only 29 or 30 places where the authentic text is found in but one MS, it would seem that the chances of no witnesses to the original reading would have to be significantly less than this. This statistic, then, actually hurts his case. Further, almost all of these 29 instances have more witnesses than just one.

On one or two occasions the author has misread f1 or f13 to mean one witness (the ‘f’ means family and includes several Greek MSS). Once or twice he assumed that sine acc. in parentheses meant that the reading was not exactly the same as what the allegedly lone witness had. But sine acc. means “without the accent,” and is thus not a genuine textual variant. Wettlaufer also did not recognize the pc as meaning pauci—i.e., a few other Greek MSS. In total, almost two thirds (19 of 29) of alleged single-witness authentic readings have more than one witness.

Not only this, but when one examines other apparatuses, the remaining ten instances are reduced to four. And of those four, only one affects the meaning of the text in any significant way. If there are but four places in which the standard critical text is based on only one witness, the statistical probability that conjectural emendation is needed for several passages in the NT is greatly diminished. In short, cumulative arguments are decidedly not on Wettlaufer’s side.

In the final section of chapter two, the author mentions the recent trend among some notable textual critics (chiefly, Ehrman, Parker, and Epp) to get away from even attempting to recover the autographic text. He adds, “When textual criticism had the straightforward goal of recovering a single original text, conjectural emendation was simply one way to attain that goal” (p. 58). This is a significant point. To the extent that modern textual critics have thrown in the towel after trying to reconstruct the autographic text, this very despair serves conjectural emendation well. If the data are really so minimal that we simply cannot reconstruct the original from the existing witnesses, conjectural emendation may provide a way forward. But the question of whether the situation is really that bleak is a different matter. Many textual critics continue to believe that the original wording is to be found somewhere among the surviving witnesses.

His third chapter offers some safeguards on abusing conjecture. Here the author gives some decent controls over making a conjecture: it must fit in the immediate and larger context, must not be counter to the author’s linguistic usage, and must not resolve a problem that exists only for modern students.

Wettlaufer begins his section on practice with rationale for using James as a case study: “Of all the texts in the New Testament it is probably the epistle of James that bears the most potential for sound conjectural emendation” (p. 81). He speaks without further discussion of an “initial period of rejection,” when James was copied and preserved less than other NT texts (p. 81). We will address the reception history of James at the end of this review.

Rather than discuss each of the three conjectures that Wettlaufer accepts for James, we will focus on just one as representative of his approach. In his treatment of Jas 3:1 he notes that although the “traditional text” explains the rise of the others, it itself is inadequate. Why? Because reading the verse as “Not many of you should become teachers . . . since you know that we will receive a greater judgment” means that teachers will be judged. The author argues that the normal meaning of κρίμα (“judgment”) cannot be softened to “scrutiny.” Thus, Jas 3:1 is out of sync with the following verses which instruct the readers to curb their tongues, but not to be silent, and that only teachers, in this reading, would be judged.

What reading does he offer instead? “Do not become garrulous teachers . . . .” The authentic reading was πολύλαλοι instead of πολλοί, in this case, and it fell out very early on via haplography: ΠΟΛΥΛΑΛΟΙ→ΠΟΛ(ΥΛΑ)ΛΟΙ→ΠΟΛΛΟΙ (p. 97). Wettlaufer admits, “It is not the neatest example of haplography, but it is precisely the type of messy mistake to which the earliest untrained scribes were prone” (p. 97). Yet, in early majuscule script, the alpha looks less like the lambda than Wettlaufer’s capital letter illustrations reveals. For example, in P46 (c. 200 CE), here is a representative line with three alphas and three lambdas:

© The Trustees of the Chester Beatty Library Dublin. Photographed by CSNTM.

Although it is possible that an early scribe overlooked the ΥΛΑ and wrote πολλοί instead of πολύλαλοι, it is less likely in the earliest period, precisely the time that Wettlaufer thinks the error must have occurred.

There are other problems with this conjecture. First, it is not really a conjectural emendation—at least, according to the definition of Metzger and Ehrman, viz., a reading not found in Greek MSS, versions, or church fathers—since it is found both in fifth-century and ninth-century patristic citations.

Second, Wettlaufer’s proposed emendation (not new with him, as he notes) actually introduces problems that, in my view, are bigger than all those that he solves. This is seen in v. 1b: “since you know that we will incur a greater judgment.” There are two issues here—the first person plural verb and the comparative adjective. If Wettlaufer’s conjectural emendation is correct, two other emendations would be needed to fix the problems still remaining in the text. First, the μεῖζον would need to be seen as an early scribal corruption, but Wettlaufer never comments on this; thus, the problem of the “greater” judgment implying a lesser judgment for all still remains. Second, λημψόμεθα is still a problem if the conjectural emendation is adopted: The “we” is referring back to διδάσκαλοι and as such it does not alleviate the problem that teachers, including James, would incur greater judgment. Paraphrased it could be rendered: “Do not become garrulous teachers, for even we who are teachers, even if not garrulous, will incur a greater judgment.” The first person verb will not allow for the interpretation that Wettlaufer claims for it, since all teachers then would still incur judgment, just as in the traditional text. For Wettlaufer’s conjecture to work, he would also have to emend the verb to either second person or third person: “you will incur [λήμψεσθε] greater judgment” or “they will incur [λήμψονται] greater judgment.” Wettlaufer seems to recognize this problem when he mistranslates the verse as “do not become garrulous teachers or receive a greater judgment!” (p. 99). In this paraphrase, he illegitimately treats λημψόμεθα as though it were λήμψεσθε. In the end, although Wettlaufer’s conjectural emendation is ingenious on many fronts, it ultimately leaves the text with more exegetical problems than it solves, and requires two more emendations in order to satisfy all the problems that Wettlaufer mentioned.

In the last chapter of this section, the author finally offers his reconstruction of the transmissional history of James—but as an assumption which he needed to demonstrate. He assumes that only one archetype was produced from which all copies of James were ultimately derived (pp. 183, 184).

This is the Achilles’ heel of Wettlaufer’s method. When he speaks of classical texts for which conjecture is a necessity, he does not note that for many, if not most, classical texts a direct line of descent can be constructed from the available evidence. Sometimes there are two archetypes that all extant witnesses go back to. But there are rarely multiple archetypes. But is it really likely that only one archetypal MS of the autograph of James was made? Wettlaufer’s mention of Isho’Dad’s and the Old Latin Speculum’s reading for Jas 3:1, and Severian (late fourth/early fifth century) as following a text that is not found in extant MSS for Jas 4:5, illustrates the very thing that Wettlaufer seems opposed to, viz., that there is more than one archetype of James from which all MSS were ultimately derived. Further, he never discusses the reception history of this letter except to say that it was “rejected” in the second century. To be sure, that seems to relate to James’s canonical status in many quarters, but this does not mean that it was not read or copied. The very prospect of it being rejected by some presupposes that there were several copies circulating in the second century. Do all these go back to a single defective archetype?

One of the arguments that the author repeatedly mentions is that since our earliest MSS often deviate from each other in significant ways, there must be many gaps in the MS record in the earliest period. True enough, but this does not help his view of a single, defective archetype for James from which all witnesses derived. The variety of early witnesses shows that more than one archetype was circulating in the early period.

Those who accept conjectural emendation for this letter need to address the reception history of the epistle and give some plausible hypothesis for the single archetype view. Or else they need to show that even if there was more than one archetype, the corruption of the text in certain places is one that is likely to have been repeated in more than one early MS. (Wettlaufer’s argument for corruption of the autographic text of Jas 4:5 is his strongest case, but his explanation for how the corruption came about is his weakest, as even he admits.) Unless this is done, the chances that all of the first-generation copies of James made the same mistake must be judged exceedingly remote. And it is this factor that has, often unstated, caused NT scholars to reject conjecture, even if the reading has a certain attractiveness to it.

The proliferation of relatively good copies (though none exactly corresponding to the originals) from an early period offers the most likely scenario. And to the extent that this is true, along with the likelihood that later scribes would at times consult more than one exemplar, the necessity for conjectural emendation decreases commensurately. Wettlaufer has made out as good a case as can be made for conjectural emendation in the NT, but in the end his view creates more problems than it solves and does not stand up to the most likely reception history of James.