Preparing my new volume on the historicity of Jesus for next year, I’ve found that one of the works published since my first volume that warrants attention in my new one is Early Classical Authors on Jesus (T&T Clark, 2022) by Margaret H. Williams (hereafter ECAJ). Here I will review this book, and comment on some of its arguments.

Overview

Apart from Josephus (whom Williams does not cover; see below), this is probably the most comprehensive treatment of the earliest references to Jesus outside Christian literature that has been published since 2014. There is an earlier multivolume set from 2019, The Reception of Jesus in the First Three Centuries, that I also need to look at (a patron has generously funded my procurement of that so it is on the way), and ECAJ is an expansion of Williams’ chapters there. But ECAJ I could afford to buy myself. It’s been on my desk a while, as I have been slowly going through it.

As I mentioned, it does not cover Josephus: Williams says that Josephus doesn’t count as a “classical” author (p. 2 n. 4). I don’t understand why not; he is a competent Greek author and a Roman citizen. Not treating him as a Classical author is kind of like treating an American Jew as not really an American. So I suspect the real reason Williams skips him is closer to what she says on p. 54, that the material in Josephus “has suffered so badly through subsequent Christian ‘editing’ that Josephus’s original words (assuming that there is a genuine Josephan core to this evidence) can no longer be identified with confidence.” And that’s all she has to say on the matter (she instead just refers readers to a four-page treatment by J.W. van Henten in the 2019 set I’m now awaiting as complete enough to serve). This means she doesn’t even think the “James” passage is worth any further comment (see Josephus on Jesus? Why You Can’t Cite Opinions Before 2014).

Instead, Williams covers the standard gallery: Pliny the Younger, Tacitus, Suetonius, Lucian, and Celsus. Full stop. She knows enough to not even mention Thallus or Phlegon (the evidence is fairly conclusive that neither ever mentioned Jesus; see my chapter on this in Hitler Homer Bible Christ). And her reasons for dismissing Mara bar Serapion as unusable (literally “unwise” and not “prudent” to include, p. 4) are worth notice: “no consensus” on its date, ranging “from the first to the sixth century”; no consensus on its author even not being a Christian; and “if these unresolved issues were not enough, recently it has been proposed that the passage” in question “is a Christian interpolation” into “an earlier pagan letter,” citing an old article by Kathleen McVey I hadn’t noticed before; this led me to a more recent analysis concurring by Catherine Chin. Now having read both McVey and Chin, I must conclude they are right: that letter (in the form we have) looks more like a 4th century Christian school exercise. So we’ve all been right to dismiss it as unusable. The Talmud, meanwhile, is in Hebrew and not early, so it genuinely wouldn’t qualify for inclusion in Williams’ study (see On the Historicity of Jesus, Ch. 8.1).

Overall, ECAJ is a useful work. Williams provides background for the authors and texts she examines, and throughout is thorough enough in citation and discussion that anyone who wishes to treat of these authors on Jesus definitely must read her book. You don’t have to agree with her (I often don’t), but you do need to make use of her research and, if necessary, address her arguments. Regardless of her faith position (which I do not know nor presume), I do sense a slight apologetic leaning (she seems intent on trying to get evidence for Jesus when she can, exhibiting a bias you wouldn’t find from a more skeptical expert; she even gullibly trusts Eusebius/Hegesippus when no one should, e.g., p. 20), but she is an expert Classicist who allows herself to be constrained by objective methodologies enough to make her efforts at that even look difficult (you can often tell when she is speculating, for example, and doesn’t actually have evidence to cite). So even when she is wrong, she has (usually) been careful enough in presenting her argumentation that you can tell. Which is a relief compared to amateur, crank, and more openly apologetical works that more often handwave or obfuscate (intentionally or not).

I also notice that she is aware of mythicism as a competing position to historicism, and though certainly a historicist, she seems diplomatically neutral on the subject (finding both sides excessive in their treatment of the material). She is also, like me, a properly trained “classicist and ancient historian.” This would leave her open to the usual ad hominem (as we hear from the likes of Bart Ehrman) that she is out of her element, when the rest of us know it’s closer to the other way around: she is actually more qualified to assess historical sources than most Jesus historians (whose degrees and training tend instead to be in “theology” or other ancillary fields that lack the contextualizing breadth of training received by classicists and historians). She even outlines some reasons why in her Introduction.

But alas, Williams does not appear to have really explored the historicity debate. For example, she credits to me the argument that Jesus should “have been attested in pagan sources” earlier than we have, when in fact I extensively argue that this doesn’t follow (OHJ, Ch 8, esp. 8.13, and my conclusion to 8.4: “I will regard minimal historicity to be as capable of predicting this state of affairs as minimal mythicism, and therefore this silence does not affect the probability of either”). Hence I say exactly the same thing she does (ECAJ, pp. 5–11). Even Lataster (Questioning, p. 438) zeroes this out as evidence in his final calculation). That plus the fact that she does not cite any page numbers for where I am supposed to have said the opposite signals to me that she merely threw in a reference to my book on presumption, and not on a basis of having actually read it. Williams also does not know (ECAJ, p. 17) that I don’t count Tacitus as interpolated in my argument against historicity, which is why I published that argument separately. She cites my article in VC (which you’ll find now in HHBC) and I’ll say more on that argument below; but in OHJ, Ch 8.10, you’ll find I reference this as a possibility, but come to exactly the same conclusion she does, that even if authentic it is not independent testimony.

This is a continuing problem with historicists. They don’t actually read the peer reviewed studies criticizing historicity. But her book isn’t about a comprehensive examination of the historicity debate, so I will only evaluate it on its subject: the five authors she examines. Indeed, even on those, their value as evidence for historicity is not her stated focus. And my focus will mainly be on errors or faulty arguments, which should not be mistaken for a critique of the entire book, which is mostly sound.

Pliny the Younger

Williams first tries to build an argument that Pliny merely “pretends” to know nothing about Christianity, an argument that though well-contextualized (pp. 19–34), is completely implausible on its own terms. Her case also relies in part on later Christian fabrications about the popularity of Christianity from Eusebius/Hegesippus, known fabricators of just that sort of thing (as I noted above; more below). But her theory is also internally incoherent. A savvy executor of Roman law would not shy away from stating and adhering to his experience prosecuting Christians or observing or reading dispatches on such prosecutions; there was no risk at all of this being mistaken for a dangerous sympathy, but entirely the opposite. Williams’ hypothesized motive for Pliny should therefore have had exactly the opposite effect, of motivating him to dutifully read out the party line and asking for advice only on what he genuinely needed advice upon.

The real issue in Pliny’s dispatch to Trajan, acknowledged by Trajan’s reply, was the use of anonymous accusations as a basis for killing people, not the mere prosecution of criminals (or even Christians specifically); and only secondarily what crimes Christians are supposed to be guilty of (and thus how they should be punished for them), which Pliny evidently resolved on his own was “illegal assembly” (with which Trajan implicitly concurs). Williams (p. 47 n. 52) erroneously believes there was a special law against Christians, but Pliny and Trajan both make clear that there wasn’t—the name “Christian” simply represented illegal association, just as other unlicensed sodalitates. A much more plausible explanation of Pliny’s ignorance is the actual rarity and insignificance of Christianity altogether in his day. Pliny isn’t pretending to anything on this point. Which flatly contradicts Williams’ claim (pp. 75–76) that everyone must have been well familiar with Christianity then, especially through their government posts. Pliny attests exactly the opposite. Realistic evidence and estimates of Christian growth confirm what he says (see my survey of the scholarship and evidence on this point in Chapter 18 of Not the Impossible Faith). People knew “of” Christians and vaguely that they were somehow “bad,” but actual encounters with them, or any further details about them, were clearly extremely scarce.

Williams then elevates another speculation into a fact: that Pliny says the Christians sing to Jesus “as if” he were a god because Pliny was told or believed Jesus had once been human (pp. 42–43). This is not supported by any evidence, and since there are equally plausible causes of that distancing phrase, we actually cannot support her conclusion over others. It’s still possible—if that’s what the Christians told him, reciting Gospel tales of it (as that is how I conjecture the standard Gospel story reached Tacitus: OHJ, Ch. 8.10). But so are either of three other explanations, which are just as likely as that one:

1. That Christians then were still particular (as Paul had been) in distinguishing Jesus, as a created being subordinate to God, from “God” (honorifically and ontologically, as the Trinity concept did not then exist: see OHJ, Ch. 4, Element 10; not even the authors of the final redaction of the Gospel of John thought they could be declared the same), and Pliny felt it best to represent the vagueness of their position as to his “divinity” (he would see no difference between that and “a” god; but they would have made much of it, leaving him unsure how to represent them, so he chose to be vague); or:
2. That Pliny wasn’t sure of (or comfortable with) granting their target of worship theological reality—in other words, he did not want to declare Jesus officially an actual god in imperial legal correspondence (which would be more in accord with Williams’ other hypothesis about Pliny’s fears in political context; and indeed, the only sense in which it would be correct here); or:
3. That Pliny was struggling to understand whether Jesus was meant to have once been human or not (as even on minimal mythicism they would have said he was, yet then denied his previous presence on Earth, that being not due until the end of days), and so he split the difference.

In the final analysis, though, Williams does not actually argue this “proves” the historicity of Jesus. She spends her time instead discussing what Pliny’s concerns and aims were, and what his letter signifies about the status of Christians and Christianity under the empire, and other ways of contextualizing the letter. She quotes Christian apologist Murray Harris affirming this evinces Jesus existed, but she never defends that take over the more accurate one that, even on his suppositions, all this proves is that Pliny was told he existed—by Christians who could well by then have believed that, seduced by Gospel myths of it. Even that can be doubted (Pliny never actually says this; and what he does say can have other implications instead, as just noted), but Williams never explores these points. She does not even seem to be aware of them. She also is unaware of the text-critical evidence that quasi (to Christ “as if” to a god) might not have been what Pliny wrote, but a medieval corruption—because quotations of this letter in early Church Fathers (Tertullian, Jerome, Eusebius; with variant manuscripts even) have either et or ut, not quasi. So her treatment, though valuable, still suffers a number of deficiencies.

It remains a fact that, as Pliny tells it, he did not consult any books or records on any of what Christians told him (not even his adoptive-father’s History, which we know included a first-person account of the reign of Nero—Williams overlooks this point, a problem I’ll revisit shortly). He only interrogated whoever he had in his court, and only then, he says, learning the details of their bizarre superstition (as he describes it). Which means his testimony is of no use for establishing the historicity of Jesus. Either those Christians just aped the Gospels at him (which does nothing to corroborate them) or didn’t even mention a historical Jesus. We can’t tell from the information he provides; and neither conclusion would be of use to the point anyway.

Tacitus

Williams notes that historicists keep treating Tacitus as “an important source” for Jesus “somewhat illogically,” given that he “produces no new information about Jesus” and “his source” is “unknown” and could easily “have been both oral and Christian” (ECAJ, p. 16, w. n. 76). Quite so. I thus do not consider the passage in question (Annals 15.44) to be “highly inconvenient” as Williams mistakenly avers (p. 67), but conclude, rather, exactly as she does: that the material in Tacitus is unusable because it cannot be established to be independent (as I already noted above). At one point Williams misleadingly says this testimony is “independent,” but not in the sense I here, and historians usually, mean (as in: independent of other sources we already have; here, that would mean, independent of the Gospels and ensuing traditions), but in a completely unrelated sense: “i.e., non-Christian” (p. 54). In other words, Williams means, independently composed, not independently sourced. She simply means this is the first non-Christian testimony to a historical Jesus we can trust is authentically transmitted. Unless it wasn’t. And that is a question she then spends several pages on, the only point where she substantively engages with my work (albeit only my journal article on Tacitus, not my study of historicity).

Minor Issues

Here Williams makes two minor mistakes in dating Tacitus worth mentioning, even though they are largely inconsequential: first, she says the Annals cannot have been completed by the end of 116 A.D., as often claimed, because “four years” isn’t enough time to write them (p. 53 n. 16); second, she says “Christian” references today all ‘mistakenly’ give that as the date of publication. In both points she’s wrong. Four years would be just one scroll (the equivalent of a single chapter in a modern book) per season, plenty of time for any ancient author. Pliny even attests (Ep. 6.16, 7.20, 8.7) that Tacitus was writing histories during his governorship (as even Williams is aware: e.g., ECAJ, p. 66); so she is wrong even to think he could only have begun the Annals afterward. And as to the date, “Christian” references are typically giving the “earliest” (not the latest), for reasons she herself comments on in her introduction: they want to bring this source as close to the life of Jesus as possible.

But I must come to their defense. Christian scholars are correct that it could be that early. Not only because Williams is wrong about the pace of ancient authorship, but also about the reasoning for assigning this date, which for some reason she does not mention: that Tacitus knows Trajan had “just now” taken Mesopotamia in 115 A.D. (Ann. 2.61.2; cf. 4.4–5, where Tacitus describes the empire that was “smaller” than the one he knew, with its then-border at the Euphrates), but does not know that Hadrian had “just then” abandoned it in 117 A.D. (returning to those prior borders). Even in our passage of concern (in the second to last chapter of the Annals), Tacitus does not seem aware of the very pertinent empire-wide Jewish Revolts of 116–117 A.D. Instead, he seems to think that that kind of thing is all in the past (entirely contrary to Williams, p. 96). For an example of both points of reasoning, see Revilo Oliver, “Did Tacitus Finish the ‘Annales’?” Illinois Classical Studies 2 (1977), pp. 289-314 (296–300 & 300–04). Which does not conclusively prove a terminus ante quem. But to not even mention this in a paragraph challenging the date is a bit of a dropped ball, especially if you are going to impugn authors for adopting it. That said, it is certainly possible the Annals were completed in the early 120s as she avers. But that hardly affects anything.

More significant is the question of interpolation, to which Williams dedicates several pages. Here her treatment suffers some minor defects. For example, she mentions C.P. Jones’ reply to B.D. Shaw, but not Shaw’s reply to Jones, and she omits W.J.C. Blom’s reply to me (pp. 55–56 n. 28; there have also been several more articles on the subject since, including one by Drews, and two by Hansen, but Williams can’t have known of them). Likewise, she overlooks the reality of fires at Rome also destroying archives (p. 62 n. 58), even though there had been several, and two quite destructive, under Nero and Domitian. And she does not know (pp. 59–61) that Pontius Pilate was probably both a prefect and a procurator (a form of corrupt double-dealing in imperial administration that is well documented: see my discussion in Herod the Procurator, and the evidence and studies I gather in Chapters 7 and 8 of HHBC). Which is unfortunate, as that fact actually shores up her own (entirely correct) point about Tacitus’s rhetorical intentions in selecting the more ignominious title to mention (ECAJ, pp. 90–93).

Other defects could be mentioned, but like the examples just given, they are usually minor. Otherwise Williams acknowledges the issues, and the diversity of ongoing views, and is mostly correct in her assessments. Indeed, various scholars lately (just as previously) have questioned even more of the text than I have (indicating a revived popularity of doubt here), and it should be noted that I side with Williams against them, making her arguments here useful to consult (ECAJ, pp. 50–74). For example, she rightly and ably shoots down both the apologist’s dreams of Tacitus “having consulted archives” on Jesus as well as bogus mythicist accusations of the passage being “incompetently written.” But in what respects I still disagree with her I’ll now address.

The Literary Argument

Williams begins with praising the literary elegance of Tacitus’s referencing the fate of Christ in comparison to the fate he then describes of (what would then be) his followers, as if that proved authenticity (p. 67). She does not notice this argument is circular: it presumes there were fates to compare; which presumes Tacitus wrote this passage about Christians, and not continuing devotees of Chrestus (see below). If he did not, then this thesis is void. There is then no reason to expect any such line here. On the other side of the equation, there is no other way to gloss Jesus other than credally, as the one executed by Pilate under Tiberius. So we cannot claim any great coincidence between this insertion and the executions to follow. That is inevitable on the interpolation hypothesis. It therefore evinces no literary art.

Indeed, the evidence is more to the contrary: if the line were authentic, Tacitus conspicuously avoided comparing the two. He mentions the Neronian victims were “crucified” (crucibus adfixi; so I cannot explain why Williams thinks Tacitus would never mention crucifixion because it was “beneath” him, pp. 134, 137, as he does so here, and even more explicitly in 14.33). And yet Tacitus gives no such notice for Jesus (he is simply “put to death,” supplicio affectus). The line does not mention how Jesus was executed or even what social rank he held so as to imply how. This sooner suggests the author of the suspect line was not attempting any such comparison at all, directly refuting Williams. So Williams is simply wrong here.

But even apart from that, since the line as written is equally expected on both theories, it can be evidence for neither. A basic principle of logic.

The Comedy Argument

The same problem befalls Williams’ argument that the verbal contradiction between “Chrestiani” and “Christus” is intentional, a side joke against the “mob” who can’t even spell things correctly, as if that argued for authenticity (pp. 68, 79). That is also circular, presuming intentions behind the text that we have no actual access to, in order to argue for those very intentions being there. A hypothesis being used as evidence for itself. A simpler explanation is that the Christus line wasn’t there, because everyone knew who Chrestians were—exactly as Williams herself suggests as an explanation for why Suetonius devotes no space to explaining who Chrestus was in his own reference to the same movement (ECAJ, pp. 201–02). Indeed, we can expect Tacitus already covered the Chrestians in his earlier volume on Claudius, too (when first they became a problem).

So we don’t expect any line here, much less an inside joke. Whereas on authenticity, you do have to adopt the ad hoc assumption that Tacitus is poking fun at the illiteracy of the vulgus, for otherwise there is no explanation why Tacitus would erroneously say Christus explains the name of the Chrestiani. A theory that requires an added assumption not otherwise in evidence is always thereby reduced, not increased in probability. The most we can say for it here is that Williams’ assumption is plausible. But “plausibly the case” does not mean “is the case.” For example, even if we conclude there is a handsome 80% chance that Tacitus was effecting this joke here, that your theory requires assuming he did still reduces your theory’s initial probability. For example, if your theory started as 90% probable, by requiring this supposition, it becomes 90% × 80% = 72%. So the probability of your theory has still gone down, not up.

By contrast, Williams complains that we have no evidence for continuing devotees of Chrestus into the reign of Nero (p. 68), even though we know from Suetonius they did once exist: as the people instigated to riot in Rome some ten years earlier, whom even she admits remained so well known even a century later Suetonius didn’t even have to say more on the matter. Which makes this another circular argument. You have to presume Tacitus is not giving us evidence for this in order to argue that there is no evidence for this. Because he is literally telling us there were folks still around in Nero’s time whom the people were calling Chrestians. So Chrestians definitely existed. The only question is who they were. And when it comes to the possibility they were not devotees of Christus but of Chrestus, we are not just “making up” a Chrestus to explain this. We know there was a very well known Chrestus who could command enough devotion to instigate riots in Rome. And it is not likely his fame would have lasted into the second century if there were not inspired followers of him at least for a time. And in Latin those people would be called Chrestians.

Unlike Williams’ assumption of an unstated joke, this is not an assumption: it is a well-evidence inference. You can say it must still lower the odds a little, but nowhere near as much as the “joke” assumption, which is indeed entirely ad hoc, unlike the assumption that there were still devotees of Chrestus for Nero to round up a mere ten years after they rioted, which is implied by all the evidence we have. So contrary to Williams, it is she, not us, who is inserting a “groundless thesis” to get the result desired. At best this is a wash (either side depends on an assumption, cancelling each other out); at worst, the interpolation theory comes out the better—for we have more evidence for our assumption than she does for hers.

Versus the Improbabilities

And that is really all Williams has. Weak tea. She has more (and more effective) things to say against the theory that the entire section was interpolated (ECAJ, pp. 68–74), but those remarks have no effect on the far simpler theory that a single sentence fragment (as even Williams calls it, a mere “epexigetical clause,” p. 53 n. 18) is actually a marginal gloss (which indeed frequently took exactly that form; and there is nothing uniquely Tacitean about that fragment; e.g., all educated Latinists cherished alliteration, p. 79). This failure to consistently distinguish the two theories trips her up at the most important point: explaining all the improbable silences in other sources, which for me clinches the case for interpolation. These include:

1. The evident silence of Pliny the Elder, Tacitus’s most probable source for this section. Williams knows Tacitus employed Pliny the Elder’s History of the Julio-Claudian era (e.g., p. 54). She also knows Pliny didn’t mention any such persecution in his brief of the event in his Natural History (p. 69), but she rightly explains that as a casualty of his brevity there (p. 70; indeed, he doesn’t mention any culprits there). But she makes no mention of (much less answers, by any apologetic) my actual point, in the very article she claims to be answering: that if there were an account of Christian prosecution and criminality in Pliny the Elder’s History of this event (as there absolutely would have been), the admiring Pliny the Younger would have known a great deal more to report to Trajan than he does (see HHBC, pp. 373–74). The failure of Pliny to remark on this is simply improbable—unless he only knew about an unrelated Chrestian event being recorded there. This is a question that could someday be decisively answered. But for now, the improbability stands. Interpolation readily explains this state of affairs. Authenticity doesn’t.
2. Suetonius’s ignorance of any connection between the persecution of Christians by Nero and the burning of Rome, an even greater improbability if Suetonius wrote after and thus knew the Annals of Tacitus (see HHBC, pp. 375–77; as Williams herself concedes he did, pp. 99, 115, 120). Williams wants to explain this away by suggesting Suetonius was duplicating material (pp. 69, 71), but that is precisely what he does not do here. He does not repeat anything about the Christians in his account of the fire (nor vice versa); he instead files them under a mere slate of moral legislation, evidently unaware of arson charges or any catastrophe to address beyond their being a general moral deviancy to quell. That is simply improbable; unless Suetonius had no knowledge of Christians being connected with the fire. Williams offers no rebuttal to this point. (On the possibility of interpolation here, see below.)
3. Even Christians had never heard of this event; not even the Latinists, even Latinists familiar with this chapter in Tacitus; not even the numerous Christian authors who discuss or narrate a Christian persecution under Nero (see HHBC, pp. 384–89). That is well nigh impossible if the suspect line were in Tacitus at the time (much less actually happened). Williams essentially evades this point entirely, by punting instead to the mere “possibility” of references nevertheless somehow existing, citing absolutely zero examples other than 1 Clement, which does nothing of the sort (something I’ll address later; see below). This reply can’t even address the specific examples of ignorance we have anyway.
4. Even the book of Revelation (HHBC, pp. 389–90), evidently familiar with the fire (Rev. 18), has no idea of any Christian persecution associated with it, but seems rather to credit it as God’s punishment upon Nero for a previous persecution (cf. the framing of Rev. 18 by Rev. 17:12-14 and 19:1-4). Which is also improbable. Williams never addresses this point.
5. Tacitus also improbably says Nero’s victims were “called” Chrestians (past tense: HHBC, p. 382), and assumes the movement he is describing no longer exists. This is quite improbable if Tacitus (especially if through Pliny) knew them to be an ongoing problem. Williams never addresses this point.

To those substantial improbabilities (which stack to an enormous improbability), we can add lesser ones, such as the (at least somewhat) greater probability of everyone already regarding them as “criminals” worthy of killing—and the sort of people you could actually suspect of arson—if they had been involved in infamous riots in Rome not long before (like the devotees of Chrestus), than if they had simply been peaceful lunatics on the invisible fringes of society (like the devotees of Christus). That is relatively weak as evidence, but it weighs in that direction all the same. Likewise the fact that this more easily explains why Tacitus says there were “a huge number” of them, requiring no recourse to explaining this away, as Williams does, as mere hyperbole (pp. 58–59 n. 43), an argument that is plausible, but again, any uncertainty attached to it will not attach to a theory that does not require it. For, even if only somewhat, one would still sooner expect that to be said of a movement so large and violent that the previous emperor had to intervene to put down their city-wide riots, so famously in fact as to stand in popular memory for a century (HHBC, p. 381).

Likewise, the first time any Christian source attests knowledge of this event, in the forged Letters of Seneca and Paul, it appears to be the source that invented the idea (see HHBC, p. 383), attaching Christians to a persecution of Jews for the arson of Rome, suggesting the original story was, indeed, about messianic Jews (the Chrestians), a fact Williams entirely avoids discussing (and thus never answers). Christians often inserted themselves into history where they did not previously exist, from the Domitianic persecution of Jews becoming a story about Christian Saints, to the Rain Miracle of Marcus Aurelius, originally attributed to a pagan sorcerer, becoming attached to a non-existent “Christian legion,” and so on. Indeed even these letters fabricate a Christian involvement in the history of the famed philosopher and statesman, Seneca. And fabrication was the normal mode of Christian literature production (see OHJ, Ch. 5, Element 44; and Moss, Myth of Persecution). That this fabrication evinces no knowledge of the version in Tacitus (agreed by Williams: p. 70 n. 105) supports this conclusion: this appears to be where the idea of Christians being involved at all originated (it’s the first time Christians ever seem aware of it), inspiring Sulpicius Severus to assume Tacitus meant “Christians” and a later scholiast to gloss Tacitus accordingly. Again, this is not a conclusive proof of the theory, but this odd piece of evidence is still somewhat more probable if that theory is true than any other (how else were “Jews” thought to be Nero’s scapegoats?). Interpolation readily explains this. Authenticity doesn’t.

Note that Williams errs here when she says Sulpicius Severus (Chron. 29) provides us with “the first certain reference among the Christian authors to the material now in Annals 15.44″ (p. 70). It conspicuously does not. This is an instance where she forgets and conflates the “one line” interpolation theory with the “whole chapter” theory. The suspect line is not attested there. All we have from him is a belief that Tacitus was speaking of Christians in that narrative, which he could have assumed by merely taking Chrestians as a variant spelling of Christians (more easily done after the forged Letters of Seneca and Paul). He therefore does not confirm the suspect line was there—and that’s already centuries after the fact. It thus could have become inserted centuries later, for all we know. No extant attestation of it predates the 9th century. It is simply too improbable to credit that it took that long for Christians to notice a line about Jesus in Tacitus; even less that they took even half that time to notice he thereby recorded a massive Christian persecution event. I also disagree with Williams on the reading and dating of 1 Clement (ECAJ, pp. 72–73). But I will publish a separate article on that later today.

But overall, though there is much of value in Williams’ treatment of Tacitus, she really just ignores the powerful arguments for the interpolation of the one clause truly in doubt, and tries to answer them with weak and circular appeals to the obvious genius of Tacitus. Her claim that the Chrestian hypothesis is just fanciful handwaving is false; it rests on more and better evidence than her arguments for its authenticity. And as just noted, she never really engages with the contrary evidence. She just handwaves it away with feeble apologetical assertions, rather than any relevant evidence or logic. And because she confuses that theory with the “entire chapter was interpolated” theory, she loses the plot by the end of her chapter on this, with all her arguments that are any good only being applicable to that hypothesis, not the other.

Suetonius

I am delighted to learn Williams rehabilitates Suetonius’s (at least not dismal) reliability as a historian by citing exactly the same example of it that I do (ECAJ, pp. 199–200): his fourth-wall digression on the evidence regarding a dispute over the birth of Caligula and how he resolves that dispute with critical reason (see my more expansive discussion in Not the Impossible Faith, pp. 182–84; cited at OHJ, p. 470). She also agrees with my assessment of the “Chrestus” passage. “Suetonius’s reference in his biography of Claudius to the urban rioter, Chrestus, almost certainly” was “to an entirely different individual” (p. 185). Williams provides two useful appendices discussing various matters usually not correctly understood by arguers over this evidence (pp. 196–98, on why dating the Chrestus riot is made difficult by Suetonius’s thematic style; and 201–05, on how we should be contextualizing his reference to Chrestus), and a substantial discussion in the text (Ch. 6, pp. 98–122). I find her analysis of this conclusive. Any gainsaying of it is mere apologetics at this point.

I am less certain whether the unrelated line only about Christians in another book (his account of Nero) is an interpolation or by his own hand. I would have liked to see more discussion of that question. But Williams seems unaware there is evidence to the point. Which further confirms she did not read my book and its treatment of the passages her book is devoted to. As I wrote in OHJ (p. 349):

It’s also possible that this line was an accidental interpolation of a marginal note summarizing the passage in Tacitus, although arguments made to that effect are not as strong as they sound. [See] Stephen Dando-Collins, The Great Fire of Rome: The Fall of the Emperor Nero and his City (Cambridge, MA: Da Capo Press, 2010), p. 6. To which one might add that the language of the line as we have it is also not in Suetonian style and reflects a Latin idiom that arose after the time of Suetonius, according to K.R. Bradley, ‘Suetonius, Nero 16.2: “afflicti suppliciis Christiani,”’ Classical Review 22.1 (March 1972), pp. 9-10; and although Bradley argues that this means the text was corrupted by emendation and should be restored to align with the paraphrase of Orosius and the known style of Suetonius, that is not the only available explanation of the same evidence.

I conclude ultimately that it doesn’t matter; even authentic, the line affords us no usable evidence for historicity. But the issue should at least have been addressed in Williams. Alas, it is not.

Even more oddly, Williams actually cites Bradley’s article on this (p. 81 n. 43) but she does not seem to know that it argues against what she cites it for: that the extant language of Suetonius here is expected. She cites it in respect to Tacitus, but it is clear she does not know Suetonius uses a different phrase than Tacitus: she didn’t even notice the title of Bradley’s article does not contain the idiom in Tacitus she is citing it for! Bradley actually says the opposite of what she thinks. His entire article is about the fact that “the use of affligere to express the idea of physically inflicted suffering did not become regular until a time considerably later than that at which Suetonius was writing,” which “immediately makes the present reading afflicti suspect” (in fact there are no instances of the idiom in Classical Latin). Bradley concludes the text was corrupted and should be revised back to Suetonian (and Silver Latin) style. But this could as well be evidence of a medieval gloss. This stands contrary to Williams who incorrectly claims “there is not a shred of evidence” for this (p. 119). Bradley’s evidence is far more than a shred. And even Dando-Collins proposes more, which I acknowledged is weak, but even that counts as at least “a shred.”

This is one of the places where Williams’ apologetical bias really shows. Her rash hyperbole here (and consequent concealing of evidence, and thus failure to address it) and her equally false claim that I argued this line was “likely” an interpolation (I clearly did not; even in my article on Tacitus I treat it as authentic, even use it as evidence, and only mention by way of caution the possibility it could be interpolated), all signify she is not being honest or objective when her desired outcome is threatened. But these kinds of gaffes merely sprinkle a survey that is otherwise correct and informative. For example, Williams argues that, like Tacitus, Suetonius also may have been in Pliny the Younger’s circle of friends (p. 120), and thus this line could well have also derived from Pliny, and thence from the Christians he interrogated (if, again, it is authentic). This removes it from the category of independent evidence. It therefore does not help us at all with determining the historicity of Jesus (or even the Neronian persecution).

Lucian and Celsus

I won’t dwell on Williams’ treatment of these later authors because neither of them attest to having any sources of information regarding Jesus beyond the Gospels (and idle folklore, such as that Jesus was ugly: ECAJ, p. 178), so they afford no independent testimony to anything. I didn’t scan everything she says about them, but on a cursory survey I did notice some remarks that warrant caution in relying on her implicitly. There is much that is correct and valuable in these chapters, but you should still always employ her material critically. For example, Williams sometimes makes wildly uninformed assertions like, “by no stretch of the imagination can the brotherhood of all Christians be considered the central tenet of Jesus’s teaching,” for “that, surely” revolved around “the Kingdom of God,” and preparing for the associated apocalypse (ECAJ, p. 135). Williams is attempting to criticize Lucian’s frontloading the brotherhood angle, but there are four enormous mistakes in her remark.

First, because the end had not come in over a hundred years, discussions of the apocalypse by Lucian’s time came to be relegated to obscure back-chapter theological speculation. It was no longer a central doctrine, much less front and center. The urgency of the Gospels’ (and even early Epistles’) imagining of an imminent destruction of the Earth had largely been replaced by a hopelessly confusing millenarianism, which held that “the apocalypse” in a sense was already happening and would take a few more centuries (or even millennia) to finally play out (see this handy guide by Coleman Ford).

Second, the Kingdom of God was to be inherited by God’s “sons,” hence the central ritual of Christianity, baptism, was an adoption by God as one of his sons and thus heirs. This is the sense in which Christians taught they were all brothers (see OHJ, Chapter 4, Element 12 and Chapter 11.10)—it was therefore more central than the apocalypse, which was merely the reason for joining the brotherhood: to secure an escape from it by inheriting the kingdom. Which indicates that Lucian evidently understood the Christianity of his day better than Williams does. This aspect of it has largely bleached out of modern Christianity; so her understanding of it is anachronistic now.

Third, Christianity was ubiquitously marketing itself as the ultimate mystery cult, the only true one, with all the same trappings as its competitors—from communal sacred meals and baptisms, to the worship of suffering saviors and the securing of personal salvation through the learning of ancient secrets, all the way to, again, brotherhood. Fictive kinship was a standard hallmark of these competitors (see OHJ, Chapter 4, Element 11), and Christianity definitely leaned on that fact in the marketplace. Hence this would be the most salient and familiar aspect of it to Lucian and his audience. The fictive kinship of salvation cults was as commonplace then as web browsers and canned beer are now. There has actually been a flurry of studies lately on this aspect of Christianity. As just the latest, on the contextual link, see Ascough, Longenecker, and Kloppenborg; and on the theological and social centrality of brotherhood within Christianity on its own terms, see Hodge, Aasgaard, Peeler, and Burke, as well as Peppard, Bossman, and Punt; even Walters’s chapter on “Paul, Adoption, and Inheritance” in Paul in the Greco-Roman World: A Handbook. So Lucian also understood his society, and the context of Christianity in its religious marketplace, better than Williams.

So keep your eye out for mistakes like that. I get the impression Williams got less meticulous and careful as her chapters proceed. This is an example. Her repetition of the apologetical claim that women were not trusted as witnesses is another (p. 156; a claim so thoroughly false it astonishes me that it still gets repeated: see N.T. Wright Demonstrates the Bankruptcy of Christian Apologetics in Under Nine Minutes). This is a point where her background should have led her to school the apologists on this. It is especially unusual to see a classicist mistake a gendered trope (Celsus’s “hysterical woman”) for a claim about an entire gender. Not all women satisfy the trope, just as not all men satisfy any of the many diverse types they could be pegged to (like “huckster” or “vainglorious soldier” or “shifty slave,” these and more being knowledgeably made much fun of in the modern musical A Funny Thing Happened on the Way to the Forum). Indeed, the fact that they don’t is the point of naming a specific trope (hence it’s “hysterical” woman; not hysterical “woman”). This is made clear by the fact that Celsus immediately goes on to say essentially the same thing of the men (elaborately lambasting Peter and Thomas as either fools, lunatics, or opportunistic liars, in Cont. Cels. 2.55). So he isn’t making any gender-essentialist claims about women. And this is made clear by how Origen defends that woman against the charge of being out of her mind (in Cont. Cels. 2.60). Williams is precisely the sort of scholar who should not have fallen for incompetent apologetical revisionism on matters like this.

Concluding Remarks

Even by her own assessment, none of the sources Williams examines establishes any evidence for Jesus independent of the Gospels and Christians echoing them. The historicity of (a mundane, and hence rather insignificant) Jesus is not challenged by this state of affairs; but neither can it be supported by it. Evidence for historicity simply must be sought elsewhere. Contrary to Williams’ claim that it never occurred to ancient critics that Jesus might not have existed at all (ECAJ, p. 189), we have evidence otherwise (OHJ, Ch. 8.12; see On the Historicity of Jesus and the Rhetoric of Justin’s Dialogue with Trypho). Rather, it was increasingly more impossible to prove, and thus rhetorically ineffective to claim. By the time of Lucian and Celsus (over a hundred years after the fact), thanks to a pervasive earlier disinterest that even Williams grants (pp. 182–89), which ensured that no useful records would survive on the point, there simply was no way to argue to Christians that the Palestinian Jesus was invented. All Lucian and Celsus had were their Gospel representations; and they were very much funnier and easier to lambast when taken literally. They had no reason to waste idle time wondering if the guy didn’t even exist (much less began life in esoteric dreams and visions); and they had no evidence by then on which to argue it (what we can see now, was invisible to them—from Mark’s Use of Paul’s Epistles to esoteric readings of obscure scriptures). But they did have the Christians’ ridiculous myths to poke fun at. Which is why that’s what they did.

Indeed, Celsus chose to frame his attack as a courtroom prosecution of Jesus (ECAJ, pp. 162–63), a pretext that would never lend itself to accusations of not even existing. Even when Celsus leaves that mode to attack Christianity directly, he has thereby already established his line of attack ex concessis (p. 176). Lucian, meanwhile, was entirely sold on the “its hucksters all the way down” motif, paralleling the fraud Jesus with the fraud Proteus, and Christian leadership generally. Challenging the historicity of Jesus would leave his pretext unanchored to any known person he could blame and poke fun at, much less parallel to his targets of satire. Similarly, Tacitus (if he wrote about Christians) needed the myth to be true for his polemic to work. Their exoteric tale was more embarrassing than the esoteric, and too well suited to Tacitus’s own agendas (as even Williams documents), so even if Tacitus knew the esoteric doctrine (and it’s quite doubtful he ever would—who would have told him?), he wouldn’t ruin a good story by mentioning it. Suetonius and Pliny, meanwhile, never mention a historical Jesus either way, nor show any interest in looking into it (e.g., ECAJ, p. 83). So this claim that “they all attest” and thus confirm the existence of a historical Jesus is simply false, even as a matter of basic logic.

This conclusion is already obvious to any critical reader of Williams. So there is no real need of further remark.