Saturday, 24 August 2019

Josephus on Jesus - Review: "A Narrative Anomaly in Josephus: Jewish Antiquities xviii:63" by Paul Hopper




This is one of a few reviews I’m doing of not-so-new academic articles on the same subject: on the first of the passages in Josephus’ works to mention Jesus Christ – the controversial passage.

This review is of Paul Hopper’s essay "A Narrative Anomaly in Josephus: Jewish Antiquities xviii:63.” It was published in Linguistics and Literary Studies: Interfaces, Encounters, Transfers. Eds. Monika Fludernik and Daniel Jacob (Berlin: de Gruyter, 2014), pages 147-169. A copy of the article can be found online.


First things

Josephus was a first century Jewish historian who wrote in Greek. Some of the books he wrote have survived in copies that have been passed down the ages. There are two passages in his works that mention Jesus Christ. Under the microscope here is the controversial mention. It’s found in Josephus’ Antiquities Book 18. It’s a passage sometimes known as the “Testimonium Flavianum” (“TF” for short). I won’t bore you with why it’s called that. (Here, I’m leaving aside the lesser known passage in Book 20. This is about Book 18.) Hopper’s essay argues that Josephus never wrote a word of the book 18 passage about Jesus. I’m going to review his arguments. This is the passage:

Now there was about this time Jesus, a wise man; if it be lawful to call him a man. For he was a doer of wonderful works; a teacher of such men as receive the truth with pleasure. He drew over to him both many of the Jews, and many of the Gentiles. He was [the] Christ.

And when Pilate, at the suggestion of the principal men among us, had condemned him to the cross;7 

those that loved him at the first did not forsake him.

For he appeared to them alive again, the third day:8 as the divine prophets had foretold these and ten thousand other wonderful things concerning him.

And the tribe of Christians, so named from him, are not extinct at this day.


I don't think anything world-shattering depends on the authenticity of this passage, but many scholars have written on it, and it is a fascinating subject to delve into. This is one of three articles on the TF that I am reviewing at the moment. The others are by Gary J. Goldberg and Fernando Bermejo-Rubio. Of the three, Hopper’s has the most problems. There is also a number of telling digressions in his essay, and if I seem to digress here and there, it's to keep up with the essay's digressions. 

Hopper believes he has found a new way to decisively judge the authenticity of the TF, which is interesting because few scholars are claiming that that can be done beyond doubt. In short, what he finds is that the linguistic narrative technique of three passages featuring Pontius Pilate - amongst which the TF is situated - is not matched by the linguistics of the TF. 

From this Hopper somewhat leaps to the conclusion that the TF is therefore entirely un-Josephan. As I will counter, in fact to have the linguistics of one passage nestled within passages with different liguistics does not equate to un-Josephan and Josephan. Rather, if - as is possible - there are other sections in Josephus' works where a passage with linguistics like the TF's is found amongst passages with contrasting linguistics, then this would rather demonstrate that what we actually have is something typical of Josephus. So, the unasked question is, are the results of Hopper's test duplicated in any relevant comparative test? Hopper does not do this test, so we simply don't find that out. (It's unfortunately quite common that this fundamental scientific step is overlooked in tests outside the physical sciences.) 

As such, Hopper does indeed find that the narrative linguistics of the other three passages are not matched by the linguistics of the TF; but he does not establish that this is un-Josephan. 

The remainder of this review will look at this in detail. 


Cross-disciplinary work

It can be welcome to see a cross-discipline contribution to a debate. In this case, Hopper’s usual field of study is not ancient history or ancient texts, nor primarily literary studies. He is a linguist crossing over into literary studies here (and he has previously published work doing a similar crossover). (He’s not purporting to be crossing into historiography per se, but see what follows…) Valuable insight into things can come from another discipline. So, here we have a linguist contributing to discussion of the TF as story, as literature, whereas it is more usually dealt with in the field of the study of ancient history and ancient texts. Sounds good, so far.

To say something about Paul Hopper's area of expertise, and by way of comparison, Fernando Bermejo-Rubio has credentials in history of religions, the historiography of Jesus of Nazareth, early Christianity and Manichaeism. And I’ll mention some more experts on the study of the TF. Those are credentials which command an audience on it. Hopper, as said, commands an audience for different reasons. He is an expert in linguistics, with experience in interactions between linguistics and literature. (And both certainly have superior credentials to myself!) My only reason for contrasting their credentials here is to give some context to some of the strengths and weaknesses that occur in his essay, to which I now turn.

The publisher of the book in which Hopper’s essay appears states, regarding the volume’s intentions:

“This volume explores the interface between linguistics and literary studies. Theoretical and textual analyses help illustrate the common features between everyday discourse and literature, and show the potentials for a collaborative approach between literary scholars and linguists in understanding speech acts and reference; inference, cognitive and cultural background; rhetoric, styles of speaking and writing; as well as perspectives and genres.”

Now there’s a potential problem here, because actually a lot of the scholarly literature on the TF deals with it from the perspective of historians, and Hopper finds himself in the realm of history as well as literature here, whether he intends it or not. I don’t know if, and to what extent, Hopper consulted experts in the field of historiography, but it doesn’t take long before problems emerge.


The “pagan” Josephus and the Pilate Stone: While I acknowledge that Hopper is writing outside his primary area of expertise, some things remain difficult to justify. It is troubling that Hopper categorises Josephus as a “pagan” witness to history. One wonders how much of Josephus’ Antiquities he actually read. I also think that this may be the first time I’ve encountered the use of Wikipedia for a citation in an academic publication: surely an academic source could be cited for the “Pilate Stone”?

Incidentally, the writing in places has an emotive tone that is surely out of place. It may in some way be to indicate his opinion as a linguist, but there is something odd in how he ascribes to the Nicene Creed “the sycophantic tone of the confirmed believer” (page 166) (and calls the Jesus of the gospels “a religious fanatic”). Is there some animus towards religion seeping through and potentially putting the objectivity of Hopper’s thesis at risk? Or is it all just clumsy handling of a field that Hopper is not at home in?


Awareness of scholarly literature: It’s usual in such an essay to begin with a review of scholarly literature on the subject. So it might come as some surprise to note, for example, that in this 2014 essay there is no mention of Alice Whealey’s influential work on the TF. One would think that Hopper ought to have had some knowledge of certain key contributors to contemporary scholarship on the TF. But evidence of that is lacking. If Hopper cites Price and Doherty (he does), then why not Whealey? He cites Feldman but omits his contribution to Josephus, Judaism, and Christianity (Leiden: Brill, 1987). A comparison with the impressive footnotes in Bermejo-Rubio’s article is telling as to the breadth of literature on the subject that Hopper seemingly had not begun to look at.

Corresponding observations apply to Hopper’s bibliography, as it confirms that his research base was worryingly unrepresentative of Josephan studies, and one wonders if Hopper could have known this when he wrote. One also wonders if his research relied on someone who provided him with a problematic reading list. Who would have given him such an unrepresentative list of scholars on the study of Josephus? His bibliography is rather short, and I was surprised to find mythicists’ publications listed but scant reference to academia’s many Josephan experts (although Feldman is mentioned). In this bibliography in a linguistics article, we however find Earl Doherty’s 2005 The Jesus Puzzle; Robert Price’s 2000 Deconstructing Jesus and his 2003 The Incredible Shrinking Son of Man. (And Burton Mack is in there.) Mythicist leanings? If Whealey’s contribution is irrelevant to Hopper’s study, then that surely might apply to Price and Doherty’s just as much. It’s a mystifyingly unbalanced aspect of the essay.


Sceptical or gullible: The essay’s baffling lack of interest in, or awareness of, wider scholarship has awkward effects. His opening gambit is a polarised generalisation about how modern “sceptical” readings stand superior to earlier “credulity” and “gullibility” about texts that are “spurious.” (This to prime the reader to be able to judge the TF wholly spurious.)

Hopper leaps to a comparison, not with a linguistic or literary example as you might expect in this context, but with archaeological fakery relating to the historical Jesus, the so-called ‘James Ossuary’ that Hopper says misled scholars Ben Witherington and John Swales. Their 2003 Brother of Jesus gets a mention, but is this to embarrass them for endorsing the James Ossuary as authentic? I hope Hopper isn't implying that Witherington and Swales are "gullible". Hopper tells us that their book “coincided almost exactly with the publication of decisive linguistic, chemical, epigraphical and circumstantial evidence that the artefact is a fake”. Hopper cites no source for such a wide-ranging assertion, merely stating that it “was quickly disposed of by the experts.” Why is he naming Swales and Witherington but not the experts he has in mind? All in all, one gets a deepening sense that this essay is not entirely as one would expect in such a publication.

Besides, how much has the ossuary got to do with the matter at hand, the linguistics he will apply? Then, Hopper does mention a forged text, the Donation of Constantine, but without letting us know if there are relevant textual comparisons to be made, which seems a shame (page 148). He just wants us to contrast the “sceptical” with “gullibility”, to make sure we expect that he isn’t gullible, and that people who are “gullible” are in his sights.

But I must move on in order to examine the documentary evidence base for Hopper’s analysis.


Evidence base for testing: Textus receptus
Any test is fundamentally dependent on its evidence base. At least four issues arise in this case:

1) Any test of the TF has a problem in that the evidence base is very slim, consisting of just a few lines of text, so some caution needs to be applied in how much we think we can take from it. 

2) There’s also an issue as to which version of it to rely upon. There is quite a bit to say about this, so I am covering this in Appendix 4 below. Potential effects on Hopper’s linguistic test are probably limited enough that we don’t have to worry about abandoning it; so, the potential linguistic problem should not be overstated. But some recognition of the potential problems that I set out in Appendix 4 would have been welcome, especially in regard to some of his non-linguistic assumptions which we see influencing his conclusions. 

Indeed, given the linguistic challenge he sets before himself, it is limiting that Hopper relies solely on the eleventh century Greek textus receptus of Josephus. Other scholars have been examining translations that are older than it, and differ from it. This is germane. As to Hopper’s test, conclusions based on the textus receptus have to be handled with a degree of caution, unless we are confident that: 1) Greek portions analysed linguistically are not compromised portions, or if so, are not accorded more weight than caution allows; and 2) any conclusions drawn about the passage as a whole are unaffected by dependence on compromised portions. Please see Appendix 4 below for more on this problem, and how it relates to Hopper's essay.

3) Another potential issue needs to be mentioned about the evidence base. Hopper a priori excludes from his test the other two stories that are interspersed with the Pilate episodes. He does not test how they compare linguistically with the TF. To omit them from the evidence base for testing is surprising. In them, we don't meet Pilate (interestingly, we barely meet him in the TF as well), but rather feature a young Isis worshipper, and a wealthy Jewish woman respectively. Curiously, in summarily dismissing these two passages and not testing them against the TF, Hopper tells us that "clearly the Jesus episode is to be grouped thematically with the other Pontius Pilate episodes". Well, that's curious, given that Hopper is going to do all he can to argue that the Jesus passage is incompatible with the other Pilate passages. But the issue I'm drawing attention to is that the exclusion of two passages in the sequence from the test prevents him from revealing to us any linguistic similarities and differences to the TF and the other passages in the sequence.

Despite that, Hopper more than implies a couple of times that he has done some degree of testing of the TF against the whole of the rest of the Antiquties - a huge work - but never reveals how we went about that, and it's a suspect aspect of the article.


4) For his analysis, Hopper proceeds with his linguistic test on the basis that the TF entirely stands or entirely falls. He does not do a comparative analysis of partial authenticity.[1] Nevertheless, what Hopper is testing linguistically has little impact on the parts of the TF that are typically thought to be interpolated, because he is mainly focused on grammar in parts that are not normally deemed interpolated.


The thesis

Anyway, to the main point. Hopper’s article, as summarised in its abstract, covers the following ground, to bring the insight of linguistics to the literary study of narrative. He claims that the language of the TF is "incompatible" with the language of the other three Pilate passages. By language, he doesn't mean vocabulary, but rather he means language in terms of the grammar of narrative technique:

“Josephus in the Jewish Antiquities introduces Jesus the Messiah into his history of the Jews, and appears to report events corresponding closely to those of the Gospels, including Jesus’s crucifixion on the orders of Pontius Pilate. A long standing dispute exists about the authenticity of this text. The present article offers a narratological analysis of the passage, comparing the styles of event reporting in the passage with the three other episodes in Josephus’s Pontius Pilate sequence. The study concludes that the uses of the Greek verb forms such as aorists and participles are distinct in the Jesus passage from those in the other Pilate episodes, and that these differences amount to a difference in genre. It is suggested that the Jesus passage is close in style and content to the creeds that were composed two to three centuries after Josephus.”

A point worth repeating: although Hopper refers to "the three other episodes in Josephus’s Pontius Pilate sequence", in so saying he is excluding from reference two other stories in the sequence that don't feature Pilate. Anyway, Hopper is claiming at least two things:

  • that the way in which verbs are used signals that the passage in Josephus has “a difference in genre” that doesn’t belong;
  • that the passage is close “in style and content” to early Christian creeds instead.

In so doing, he is arguing that Josephus was never its author, but that some later Christian authored the whole thing.

How does his argument fare? Well, Hopper’s essay has made little or no impression in academic study of the TF. You might wonder why. In this review, I will try to explain why it might be found unpersuasive. It’s not just a matter of the problems I’ve listed above.[2]


Framework for testing the text

Hopper constructs a framework of linguistic aspects which – via study of narrative as literature - he sees as potentially determinative for authenticity of the TF, and he is focusing on verbs.[3] He chooses to especially factor in:

  • the perfective (Greek aorist) system (used in the text for foregrounding single completed main actions – and sequences of such – especially those of protagonists, things that carry forward the skeleton of a narrative); and
  • the imperfective system (for less focused backgrounding, stage-setting, not so fixed on particular events, nor specific moments in time);
  • and he is assuming that with his method he can identity a protagonist(s), and identify genres. (I found him a bit unclear as to what genre ought to mean here, given his lack of engagement with historiography - perhaps he just means that that an eventive narrative is itself a "genre" in a narratological sense.);
  • and he claims that with his method he can declare the TF an anomaly in terms of literature. All from a narrowly drawn test of a rather short passage.

But this approach does not derive simply from study of ancient Greek literature. He is drawing on categories from Wallis Reid’s work on French linguistics about how perfectives attach to individuals rather than groups, inter alia. Hopper cites other authors on aspects of Greek construction, such as Diver on structure in Homeric Greek, which Hopper presents as comparable.[4] I would like to see a little more work done by scholars on the leap Hopper is making from one to the other, with specific reference to a wider review of Josephus. He considers patterns found in French literature as sufficiently applicable to the ancient Greek text of Josephus, finding one narrative form more or less mirrored by another (relating especially to Greek aorists). He is conducting a finely calibrated test, so some caution is required here, at least with regard to some of the conclusions Hopper tries to take from his findings. His identification of the verb forms and their locations in the passages is uncontroversial in itself, as you would expect.

For those unsure what an aorist is, it's the sort of verb for you have in Greek for saying, "The boxer just punched that other man, knocked him to the ground with one blow, and raised his fist high." It's action: he did it. Heroes and villains tend to behave like that in action films: "The Joker pulled a lever to open the trapdoor. Batman jumped over it. The Joker pulled a gun, and a joke handerkerchief popped out." And so on. You get the idea. Protagonists and antagonists can act that way. Hopper's interest is in the kind of literature where aorists accrue to protagonists primarily.

Now, supposing that this be accepted as a valid framework to use in testing the Greek text of the TF compared to the other Pilate passages, then to what good use does Hopper put it?[5] His approach boils down to identifying a protagonist in events in the TF, and saying whether it is the same protagonist as in the other Pilate passages, and the same or a different genre.[6] We are ready to study his core argument.


The test:

1) Eventive

The other three Pilate passages are eventive in nature, and Hopper uses that as a control for the test. The TF is not eventive; it is more like notes. There’s no reason to doubt that the other three Pilate passages are eventive in nature, but is that evidence base broad enough for the test? Hopper seems aware that it might be insufficient to divorce the TF wholesale from the Antiquities. Perhaps that is why Hopper seems possibly tempted to exaggerate the evidence base that he has tested against (164):

“The Aquifer episode, like the other episodes involving Pontius Pilate, has an event structure. Time in these episodes is kairotic, that is, it is qualitative time (kairos) experienced by individual actors. It is eventive time... By contrast, the temporality of the Testimonium is chronic (chronos)…. It takes place in a more remote perspective of slow changes and general truths… So the Testimonium belongs to a different kind of time from the rest of the Jewish Antiquities.” (emphasis added)

Where has Hopper done this analysis of “the rest of the Jewish Antiquities”? It is unclear – because Hopper doesn’t tell us - what assessment may have been done of the whole of the rest of this huge work to establish this sweeping conclusion. Hopper makes such a claim more than once.[7] Such a large claim needs to be substantiated, but evidence to that effect is lacking, and the impression remains that it is exaggeration. This is potentially rather important to Hopper’s thesis: are there other sections of Josephus’ works where an uneventive passage is embedded amongst eventive ones? It would be strange if that were not so, given how often Josephus digresses and then returns to a subject. Finding any would see this essay's thesis collapse. So it is perhaps understandable why someone might want us to think that this never happens anywhere else in all of Josephus’ Antiquities (and what about in Josephus’ War?). But Hopper’s claim to know that this doesn’t occur anywhere in Antiquities is not convincingly presented. (Also, would Hopper really describe Josephus as a "pagan" witness if he had examined the whole of the Antiquities to the degree he seems to suggest?) Josephan scholars know the Jewish historian’s work in much greater depth and breadth, and it is perhaps unsurprising that Hopper’s thesis has made little or no headway in academia. In short, his claim about the TF against all “the rest of the Jewish Antiquities” has to be set aside. The TF up to this point cannot be divorced from the text of Antiquities.

This large claim is made seemingly with the sole objective of making the TF seem exceptionally anomalous. For me, alarm bells are ringing of a lack of objectivity putting his thesis at risk. This is one of a number of extravagant claims.

But, as said, there is no reason to call into question that the other three Pilate passages are eventive, so there is a basis for testing here. By the by, it’s obvious that for Hopper, what carries more weight is whether particular things found in the longer passages are found in the rather short TF, rather than whether things found in the short TF are found in the longer passages. That’s where he really places his weight.

So with three Pilate passages that are eventive and one that more or less isn’t, Hopper turns his attention to who the protagonists are in the stories. In his sights is the question of whether the TF belongs where we find it, and he thinks identifying a protagonist may help.


The test:

2) Is Pilate the protagonist?

So, on pages 156-60, Hopper selects a number of excerpts from the surrounding three story-driven eventive Pilate episodes as a control for his test (Antiquities 18:55 to 18:89, after prior mentions of Pilate from 18:35 onwards). In order, the TF is the third of four Pilate passages, but of course there are six stories in the sequence of passages, three of which are eventive Pilate passages. So, the TF includes the part of Pontius Pilate, which captures Hopper’s interest. Hopper more or less describes Josephus’ other nearby Pilate passages as following a pattern of (eventive narrative-driven) tales of Pilate versus violent rebels, which is uncontroversial to say. Then, he seeks to persuade us that Pilate is - in a technical literary sense - the clear protagonist in those other passages on account of how aorists in the narrative accrue to his character’s actions. There are other ways to look at whether Pilate is a protagonist, but the point of Hopper’s contribution is to bring linguistics to literary studies.

In short, his test relies on his assumptions that:

  • the location of aorists in those other Pilate passages is a vital determinative of his finding that Pilate is the protagonist in them
  • and that an authentic TF should mirror this: that it too should be an eventive narrative, and that aorists ought to accrue to Pilate, and that Pilate should have this marker for being found to be its protagonist. (Why should it?)

Hopper then lays out the evidence base that, in the TF, aorists do not accrue to Pilate’s single action. No disagreement need be had with that. There is no reason to dispute the basic evidence that aorists do not accrue to Pilate’s single action, and that this leaves the TF’s Pilate without a marker found in the longer passages. These other passages have more than a single action by Pilate, so it is not too surprising that their verbal characteristics include things that are not found in the single action in the TF, but Hopper thinks it matters a lot.

It’s Hopper’s conclusions that are so interesting, and the point of my review is to scrutinise them: as the TF doesn’t fit the wider observed pattern, he says, this supposedly differentiates the TF as not belonging in this piece of literature. As if observing, in this way, that Pilate is not the protagonist, and arguing that a different genre is present (well, the particular genre he is going to suggest anyway...), is so determinative. That, basically, is his argument in a nutshell. Relying on this method, he casts around for another chief protagonist in the TF which he’ll use to help identify another (very particular) genre to buttress his argument that the TF must be inauthentic.

So, that test may potentially be a way of finding that Pilate is not the protagonist in the TF. Whether it is an applicable method or not, it does not upset the ordinary observation that Pilate is not the main protagonist of the TF.

While Hopper places so much weight on findings from linguistics, and has published on its connections with story-telling before, I’m unconvinced that he is going far enough in acknowledging a range of ingredients that go to make literary story-telling, or of the weight that could be accorded to the contribution of other Josephan scholars, or of Josephus’ literary habits.[8] He seems to be applying a narrow analysis with a few digressions, whereas storytelling requires analysis with more breadth (certainly for this kind of judgment). 


The test:

3) Is Jesus the protagonist?

As for verbal forms, Hopper notes that Jesus’ activity is portrayed by use of imperfectives, in only broad strokes, not single actions. Hopper claims that this Jesus is “a passive participant” (page 162). “Passive” seems to me the wrong word for it. This Jesus is portrayed with seeming responsibility for the areas of activity ascribed to him, and only passive in his death. Hopper also pays little regard to the fact that more areas of activity are ascribed to Jesus than to anyone else: “a doer of wonderful works; a teacher ... He drew over to him both many of the Jews, and many of the Gentiles… he appeared to them…”8

But Hopper wants to put his weight on verbs in the aorist aspect; which he seeks to do with the next candidate.


The test:

4) Are Christians as a group the protagonist?

Bear in mind that Hopper is assuming that the TF ought to have a protagonist, and is assuming that aorists are more or less a sufficient tool to identify one. On pages 160-162, Hopper argues that, insofar as the TF counts as a narrative at all, and insofar as there are protagonists at all, it is not Pilate but Christians as a group who fulfil a protagonist role. Well, it is obvious that it is not Pilate, but why Christians? Again, he is relying on observation of to whom aorists accrue. But there is not a lot of aorists to go around in this short passage, so the evidence base is slim: of Christians, the aorists tell us that as to following Jesus, “they have not forsaken”; and “not ceased to exist” What such verbs ascribe to the Christians is that they didn’t do this, and they haven’t become that. And that’s it![9] An aorist with an action is one thing, an aorist without an action is quite another.

Hopper, putting his weight on the location of aorists, argues that the Christians are the overall protagonists of the whole passage. Really? Even with those aorists, the Christians are at most present in broad strokes, in negatives (did not forsake him, are not extinct). No specific eventive actions are ascribed to any Christians whatsoever. This is hardly satisfactory to meet Hopper’s own test. So how can we promote them to chief protagonists? Hopper, perhaps anticipating that this objection is coming, lines up this truism: "Negatives point implicitly to the corresponding affirmative" (page 162). The text merely indirectly conveys that they are still around still following Christ, but the text does not give them the dignity of having this ascribed to them as an action. (It merely implies a process of following, but actionless. It’s a bit like damning with faint praise.)[10]

And that, as far as linguistics go, was Hopper’s attempt to say that the Christians as a group are the protagonist. (He also enumerates participles, inter alia, in the passage, but he doesn't try to put much weight on that.) I don’t find Hopper’s method producing convincing results.

Obviously aware that linguistics don’t cut it here, Hopper seeks to buttress his argument that the Christians are protagonists. As the actionless negative statements - “did not forsake him… are not extinct” - hardly meet Hopper’s test, he finds a way to pull a rabbit out of the hat. He Christianises the phrase “did not forsake him” and hands it back to us as “they haven’t stopped worshipping him”. This is what Hopper wants us to read into the text: a subtle shift closer to an action, one that is quite out of proportion to “not forsaking,” and gratuitously reading Christian meanings into quite ordinary language - not the mark of a balanced linguistic test. It’s precisely the sort of thing that a careful scholar ought not to be doing.[11] Hopper is supplying overtly Christian meaning by recklessly importing into the text what he wants to find. It looks agenda-driven.

His reason for lining up the word “worshipping” is apparent: to argue for the TF being wholly written by and for Christians, as if they are supposedly saying something like, ‘look, we’re strong, and still here worshipping Jesus.’ This is the kind of way in which he makes his leap from different protagonist to different genre. It is intended to amplify his case that the TF is inauthentic. If anything, it serves further to call into question the objectivity of his essay.[12]


What protagonists are there, then?

Linguistics aside, if one were looking at the TF to uncover an overall protagonist for the passage, it is too much of a scrapbook, and one would not necessarily conclude that the Christians, or anyone else, are the overall protagonists on any grounds. The full version of the TF can be broken down into five chunks if we go by whom the spotlight is on. These are, in order: Jesus; Jews and Pilate; Christians; Jesus again; and Christians again. This illustrates why a narrow analysis, such as a linguistic one only, would be insufficient for conclusions that require breadth. Unredacted version:

Spotlighting Jesus: Now there was about this time Jesus, a wise man; if it be lawful to call him a man. For he was a doer of wonderful works; a teacher of such men as receive the truth with pleasure. He drew over to him both many of the Jews, and many of the Gentiles. He was the Christ.

Spotlighting Jews and Pilate: And when Pilate, at the suggestion of the principal men among us, had condemned him to the cross [to bring out the Greek: “on indictment by the first men among us, Pilate having sentenced him to the cross”]

Spotlighting Christians: those that loved him at the first did not forsake him.

Spotlighting Jesus again: For he appeared to them alive again, the third day: as the divine prophets had foretold these and ten thousand other wonderful things concerning him.

Spotlighting Christians again: And the tribe of Christians, so named from him, are not extinct at this day. [emphasis added]

Is “did not forsake him…. are not extinct” really enough to make Christians the chief protagonist of that? Of course not. Not forsaking is not directly an action, and certainly not eventive. Not being extinct isn’t an action either.

The above confirms that Pilate is one of the active players in the TF, but does not manage to grab the main clause of the one sentence where he appears (ditto the Jews): “on indictment by the first men among us, Pilate having sentenced him to the cross…” (emphasis added). It is in a subordinate clause that Pilate has his moment, with grammar as appropriate (page 163). Pilate’s action is not in a passive voice, note. It’s worth bearing that in mind, but it is right to say that Pilate is not the main protagonist.[13]

I would say (and I have studied the elements of story-telling at Masters level, so this is not a comment born of total ignorance) that one cannot settle on the protagonists of an overall piece with merely a single tool such as linguistics, even when reduced to verb study, when other helpful complementary tools are also available to assist the determination. This passage is more like jottings, and it is not warranted to insist that it has to have a chief protagonist, not even if Hopper wants to ascribe such to the actionless Christians as a group. In fact we need to be wary of determining that of any of these passages. This is not Aristotelian drama. We can expand on this point as follows, turning to the other Pilate passages.


Pilate passages: protagonists

Hopper needs us to believe that in the other three Pilate passages, the rebel forces consistently are not main protagonists, but that Pilate consistently is. And he wants us to believe that this is more or less inverted in the TF.

But none of the passages are even close to, say, Aristotelian narrative structure with any clear protagonist or antagonist. The roles are not that clear cut – multiple protagonists and antagonists switching roles here and there would be more like it. It’s not tidy. Pilate in large part is an antagonistic individual who frustrates others’ plans, rather than making his own (at times), while hindered by those acting beyond his instructions, making him seem reactive and even a little weak (at times).

All four Pilate passages feature apparent conflict with him. Pilate reacts to each with bloodshed, killing groups or their leader (Jesus). The waters are too muddy for identifying indisputable protagonists and antagonists in any of the passages. Pilate, like a protagonist seems to weigh pros and cons in his role of keeping the status quo of the Roman peace, but the rebels too have goals (to change the status quo) and Pilate’s main role is resisting and frustrating their attempts to achieve their goals, ultimately becoming villainous in doing so. Linguistics alone do not resolve this, and Hopper may have fallen into a trap of trying to tidy up something that we are best conceding is not asking to be tidy, and into a trap of over-stating the weight of his linguistic analysis (and he probably sensed that we would spot that, and he gives this away with what comes next, below). He seeks to buttress his conclusions in other ways...


Pilate as stereotyped villainous protagonist

Hopper obviously senses that his narrow analysis isn’t as determinative as he wants it to be. So, in his bid to make the TF the odd one out, and make Pilate a strong protagonist of the other three passages, Hopper digresses to turn Pilate into something like a stereotyped villain, which rather misrepresents Josephus’ portrayal. Hopper more or less wants us to see the Pilate of the other three passages as a one-dimensional cold-hearted thug, and the Pilate of the TF as an anomaly, a weak victim of Jewish pressure. He writes:

“Pilate, the decisive Roman boss of the other three Pilate episodes, ruthless scourge of the Jews and despiser of their laws, now appears as the compliant puppet of the Jewish hierarchy [regarding Jesus].”

He seems to be suggesting this radical dichotomy in order to make us think the TF anomalous and inauthentic. But Pilate is not quite as Hopper claims. To ignore Josephus’ more generous comments about him is hazardous. (See Bermejo-Rubio, "Hypothetical Vorlage", page 350, n. 113.) Let’s takes quotes from the first three of the four Pilate passages. In the first, he backs down under Jewish pressure:

story 1: “Pilate was deeply affected with their [the Jews] firm resolution to keep their laws inviolable: and presently commanded the [imperial] images to be carried back from Jerusalem to Caesarea”. [Which Hopper skates over quickly - page 154. So much for the simplistic “despiser of their laws”!]

story 2: “[the soldiers] laid upon [the Jews] much greater blows than Pilate had commanded them”. [So much for “ruthless scourge”!]

story 3: “on indictment by the first men among us, Pilate having sentenced him [Jesus] to the cross”. [This does not seem inconsistent with the above.]

Note patterns in the first three [passages in which Pilate is included:

  • In the first story and the third (the TF), Jewish pressure upon Pilate prevails.
  • In the first two stories, Pilate’s less harsh side – if not relative weakness – is glimpsed. In the first, the Jewish hierarchy successfully prevail upon him to do what he didn’t want to do. In the second, his soldiers go further that he wanted them to go.
  • In the second story and the third (the TF), other actors (certain Romans, and Jews) are more inclined to advance harsh punishment than Pilate is. In the second story, the soldiers are more responsible than he for the bloodshed. In the third, certain Jews seem perhaps more responsible than he for the execution.[14]

We can see that the TF is embedded in patterns and is actually not inconsistent with Josephus’ portrayal of Pilate. Hopper’s stereotyping view reads Josephus with blinkers, if not distorting mirrors.

Only, finally, in the fourth story is it that Pilate ultimately goes too far and is fully responsible for actions that see him deposed. Only this fourth story has Pilate close to meeting Hopper’s description. Whereas the earlier stories undermine Hopper’s position! Let’s recap salient quotes from Josephus:

story 1: “Pilate was deeply affected with their [the Jews] firm resolution to keep their laws inviolable: and presently commanded the [imperial] images to be carried back”

story 2: “[the soldiers] laid upon [the Jews] much greater blows than Pilate had commanded them

story 3: “on indictment by the first men among us, Pilate having sentenced him [Jesus] to the cross”

story 4: “Pilate prevented [the Samaritans] going up, by seizing upon the roads, with a great band … the most potent of those that fled away, Pilate ordered to be slain.”

It is a misreading of the evidence when Hopper claims that “the Testimonium is hard to reconcile with Josephus’s denunciation of Pilate’s crimes against the Jews.” Contra Hopper, only in this fourth story is Pilate’s command absolute and he fully responsible for the worst of this extreme clampdown. (And it serves to bring him down.) In this aspect, this fourth is the odd one out, and for obviously intentional narrative reasons. Hopper overlooks significant similarities between the TF and the first two Pilate stories, and their differences with the last one.

All in all, it is difficult to reconcile this evidence with Hopper’s attempt to treat the TF as anomalous as literature. His caricature of Pilate is untenable, and seems purposed simply to make us fall for a dichotomy and to imagine that the TF doesn’t belong as literature. As if Josephus can't vary genre and protagonists in a sequence, even though Hopper notes that authors can do so - see page 151 (authors who act in a too restricted way can be rather boring). 


Genre: Christian apologetic material?

The genre which Hopper suggests the TF to be, is a Christian genre. He has been leading up to this by magically transforming "forsaking" into "stopped worshipping", etc. The case for the TF’s genre being a Christian apologetic falters once the question of partial authenticity is raised. Hopper wants to convince us that the full TF must be wholly a Christian apologist’s invention. However, as noted earlier, this is Hopper’s judgment only on an unredacted version, whereas most scholars don’t consider an unredacted version determinative for study of the passage.

Virtually all scholars, Christian and non-Christian, accept that there is an element of Christian tampering with the text. After all, the suspect full-length version of the TF looks like it implies truth in the story of the resurrection and the messiahship of Jesus![15] Those bits would be strange for an unbelieving Josephus, but not for a naughty apologist. But take out such alleged interpolations and the TF doesn’t serve an apologetic purpose. As such it is unlikely to be the work of a Christian apologist in toto. Many scholars think a redacted version is merely neutral in tone; although (even worse news for Hopper’s thesis) a very strong case has been made by Bermejo-Rubio that the redacted version of the passage has a sly anti-Christian slant by Josephus himself. But Hopper refuses to countenance a redacted version. In other words, his argument for it being a Christian apologetic is over-dependent on his hasty judgment against partial authenticity. (See Appendix 1 below.) 


Were Roman readers stupid?

Hopper also tries to buttress his case by re-locating the TF’s first audience from the first century to the third/fourth century, one that knew Christian creeds. This leads to the impression that first century Roman readers must have been stupid.

It is interesting to see why Hopper conjectures a later Christian audience - one such as would find the story of Jesus familiar. Hopper’s premise is that Josephus would want to give his readers nothing but eventive stories in this literary sequence (if we allowed he could have had the material to write one). And he argues that since it’s not one, we need to find an audience that could supply its own narrative meaning: a much later Christian audience: “The temporality of the Testimonium derives from its presumed familiarity to its audience, which in turn is more compatible with a third century or later Christian setting than a first century Roman one.”

Hopper is not on secure ground. Firstly, Goldberg’s textual analysis has shown that in fact the TF doesn’t fit in the third/fourth century as a point of origin.

Second, Roman readers were not so stupid as Hopper needs them to be. Hopper rules out, in his conjecture, a late first-century Roman audience because he assumes they would need external information and had none (how does Hopper know this?). That must mean that late first century Romans had never heard of Christ, had never heard of Christians being scapegoated for the 64AD Great Fire of Rome, and frankly must also have been too stupid to understand a fairly simple passage while on the other hand intelligent enough to find Antiquities otherwise understandable. But according to Hopper, assuming that this piece of literature requires an eventive story to function for a reader, argues “The Testimonium could not be understood as a story except by someone who could already place it in its “intelligible whole”” [emphasis added]. Goodness, these educated literature-fed Romans, nurtured on the sophisticated Latin poetry of Virgil and his ilk, and interested in Greek literature (otherwise they wouldn't be reading Josephus' tome in Greek), must have somehow happened to have become really poor readers, if you will forgive me for the sarcasm. Why does Hopper think that he knows that “told as a series of new events to a first century Roman audience unfamiliar with it, the Testimonium would have been a bizarre addition and probably quite unintelligible.” He seems to think that Roman readers would be all at sea, reeling with confusion, if they don't find an eventive story here. This is rather patronising towards sophisticated literate audiences past.

How many people really insist that the following is in itself "bizarre" and "unintelligible" on its own terms? -

“Now there was about this time Jesus, a wise man; if it be lawful to call him a man.

For he was a doer of wonderful works; a teacher of such men as receive the truth with pleasure.

He drew over to him both many of the Jews, and many of the Gentiles. He was [the] Christ.

And when Pilate, at the suggestion of the principal men among us, had condemned him to the cross; those that loved him at the first did not forsake him…” etc.

Schoolchildren read much more difficult passages than that. I acknowledge that Hopper does not claim to be a historian, but what is his knowledge of the literate world of first century Roman readers that makes him think they would have found it beyond their skill to understand? Wrapping his argument around words like "bizarre" and "unintelligible" does not make it so.

Hopper imagines a different audience with prior knowledge of a bigger story. He means a Christian audience. This, he believes, is the first audience that could have understood it. This seems to me to be an unwarranted view, regarding Roman audiences as incapable of comprehending a few sentences that follow chronologically. The TF does not need to be interpreted through a tapestry of Christian knowledge to derive some meaning from it. Josephus was not placing that much weight on it.

The TF is intelligible, while not asking to be understood as an eventive story. It is just a thumbnail sketch of something that has some relevant correspondences and contrasts with the Pilate material it is embedded in. The TF is woven into the passages quite sufficiently. Hopper makes an excessive claim that “without the Gospels the passage is incomprehensible.” We can put this one alongside some of the other extravagant claims in his essay. Hopper has more to say on his idea of the first audience being Christian, and he makes a strained comparison with Christian creeds. I look at what he says about that in Appendix 2 below.


Conclusion

In fact, the TF nicely fits in the to and fro tension in the interplay of power between Pilate and the Jews, and the survival of one movement contrasted with the demise of another. The meaning is all in the context, not extraneous to it.

Hopper is a distinguished scholar, but this does not have all the hallmarks of a distinguished work. I have the sense that all is not as it seems and further work is required to test the validity of the conclusions of his linguistic analysis (for example in relation to the broader base of Antiquities). This essay, however, is deeply flawed. It is doubtful whether authenticity can be proved beyond doubt either way, but all I can say here is that Hopper has not found a way to resolve the question.





Appendix 1: Partial authenticity

It’s worth noting that, generally, scholars who have taught in relevant disciplines in accredited academic institutions who have opined on this subject take the view that the TF is partly authentic, partly interpolated. (For some leads that you might wish to follow up where you can find reviews of the academic literature, here is a handy source list. Also see this article for a degree of summary). The scholarly literature yields important information about both consistencies and inconsistencies between the TF’s content and what we would usually expect of Josephus (useful evidence for partial interpolation), and ideally Hopper could add valuably to this corpus.

Three phrases typically taken by scholars as evidence of tampering are: "if indeed one must call him a man", "He was the Christ"; and "he appeared to them alive again the third day." Most reconstructions remove and/or amend some or all of these phrases, and some nibble away at a few more surrounding words. Those who want to defend some or all of those phrases use differing arguments, interpreting them contextually as either positive, or neutral, or negative. 

However, as if no serious scholar has anything meritous to say for partial authenticity, Hopper forces a false dichotomy on the reader (page 151), positing a clean split between “Christian commentators tending strongly (but not universally) to support its authenticity and religious sceptics seeing it as fraudulent.” (Hopper places himself in the latter camp.) This is similar to his dichotomy between modern “sceptical” readings that stand superior to earlier “credulity” and “gullibility.” But no-one has to buy into a false dichotomy. We do not have to side with Hopper’s characterisation of either “Christian commentators” or “religious sceptics”. We can rather also listen to secular scholars who would not necessarily want to be bracketed in either camp. But their contributions on partial authenticity are somehow erased from the discussion, conveniencing Hopper’s all or nothing approach.

So, his article purports to be only for further discussion on whether the TF is wholly authentic or wholly spurious, ironically, buying into the approach of both some Christian apologists and religious sceptics!

Hopper appears to be something of an outlier in his somewhat peremptory dismissal of the possibility of partial authenticity. For his study, an interesting question could arise. Does one test the linguistic-narrative patterns twice, both on whole and redacted versions (that exclude likely interpolations)? Hopper goes only one way, testing only the full passage whole, but his reasoning for this is limited:

“the syntax of the Testimonium does not display the kinds of discontinuities we might expect to find if substantial changes such as major deletions or insertions had been made. The sentences are well formed, the use of particles such as gar and de is appropriate, the Greek constructions are correct and complete. In short, the passage is linguistically and conceptually integrated, and the assumption of an originally longer text that has been substantially shortened or of a shorter text that has been lengthened does not appear to be warranted on purely internal linguistic grounds.” [emphasis added]

On my points of emphasis:

1) Hopper is giving no attention to the fact that redactions generally suggested by scholars are of just a word here, or a clause there, not “major deletions or insertions”. So this is a moot point.

2) The constructions may be ”correct,” but some of them are frankly strange. Hopper gives insufficient attention to this. Whereas Goldberg focusses on the oddities, why doesn’t a linguist like Hopper? Perhaps it doesn’t fit into his thesis? For example, concerning the resurrection, Hopper first translates literally as “having day again living” (similar to Goldberg) but then goes with Feldman’s dynamic translation “restored to life.” Hopper does not comment on the oddness of the Greek. Surely this is precisely the sort of thing that would catch the eye of a linguist, but not here. Inattention to things like this makes one wonder.


3) His qualification “purely” is something of a get-out: Hopper clearly implies that other disciplines could influence conclusions, potentially in a different direction, but he doesn’t engage with them.

There are linguistic points in favour of partial authenticity, and there is a helpful summary by Tim O’Neill here.



Appendix 2: genre - related to Christian creeds?

We saw a signpost to Hopper's bid to dislocate the TF from the Antiquities, where he Christianised the word "forsaking" out of all proportion into "stopped worshipping". This was towards justifying his saying that this is not a genre that belongs in Josephus' writing, but rather a Christian genre. But what Christian genre is he claiming that this is? Hopper:

“The closest generic match for the Testimonium is perhaps the various creeds that began to be formulated in the early fourth century, such as the Nicene Creed (325 CE). Some credal elements are clearly present: Jesus was the Messiah; he was crucified under Pontius Pilate (passus sub Pontio Pilato, in the words of the Apostles’ Creed); he came back to life on the third day after his death; the movement founded by him – the Christian church – continues to flourish; he performed miracles; the biblical prophets foretold many details of his life.” (page 166) [emphasis added]

Let’s set aside that creeds began to be formulated centuries before he claims. Perhaps that would be nitpicking. What’s more interesting is that Hopper is clearly putting his weight on the unredacted version of the TF. (A redacted version typically lacks some of those elements.) But more importantly, this critique does not convincingly suggest that Hopper has done serious study of the texts of early creeds, not even in regard to aorists (he refers to the Latin creed above, quoting passus sub Pontio Pilato, so he is obviously not entirely in a position to make a linguistic judgement relating to aorists).

It has to be said that the TF and the Nicene creed are notably different in their treatment of Pilate, which Hopper only obliquely signposts towards:

  • Whereas the creed says that Jesus “was crucified for us under Pontius Pilate” – reducing Pilate’s part to a point where no action is ascribed to him - the TF instead speaks of Pilate as actor, “Pilate having sentenced him to the cross”.
  • And whereas the creed gives the crucifixion itself theological significance, “crucified for us”, the TF does not, simply treating it as an effect of government.
  • Unlike the descriptions in the TF, the creed actually has no mention of any of these:
    • Jesus doing miracles;
    • any mention of the church continuing to “flourish” (Hopper’s word for it), only of belief in the church.
  • Also whereas the unredacted TF has the past tense “He was the Christ” (Ho Khristos outos ēn) this is not formulaic credal style material, unlike the actual creed’s notably formulaic address to “Lord Jesus Christ”.

Why does Hopper think that the TF is close to such creeds when it is so different? Curiously, he ascribes to the Nicene Creed “the sycophantic tone of the confirmed believer”, and thinks he sees something like that in the TF (page 166). This seems like a case of distorting mirrors, not perhaps out of character with many things in his essay.

It is difficult to see that his attempt to transfer the genre of the TF into the realm of Christian creeds adds any weight to his case.

By the by, sceptics who speak of the TF as completely interpolated generally refer to it as a fourth century creation appearing first in Eusebius in the fourth century. Hopper, without explaining why, dates such interpolation a century earlier, saying that the TF “reminds third century Christians of events already familiar to them. 


Appendix 3: Other non-linguistic objections to the TF

Separately from linguistics, there is a somewhat one-sided catalogue of familiar looking arrows fired against the authenticity of the TF by Hopper (it’s not clear why Hopper peppers his essay with these well-trodden arguments), so I will just state in brief where discussion on these tends to go:

Patristic citations of the TF: Hopper cites Feldman on the low amount of patristic citations of the TF (page 151), and makes an argument from silence to suggest that the TF is spurious. Bermejo-Rubio’s article dismisses such argument from patristic silence, questioning the significance of it.[16]

Brevity of the TF: Without seemingly being aware of common solutions to what he sees as a problem, Hopper argues that “The Testimonium itself is, when compared to the surrounding episodes, unusually short. Its very brevity is a suspicious feature…” Well, it’s not that suspicious when compared with other examples of Josephus’ moments of brevity. (Which I have addressed in another post.)

Non-mention by Justus of Tiberias: Hopper tries to buttress his case by saying that not only did Josephus not mention Jesus but neither did first century writer Justus of Tiberias. (This is actually putting weight onto a third party, Photius, a ninth century reader of the now lost works of Justus who records that Justus does not mention Jesus.) But arguments from silence are hard to make stick, and we would need to have a copy of Photius’ works to make a meaningful test of this. It would be interesting to know, for instance, whether Jesus belongs to any class of things that Justus normally would record – an essential to this kind of argument from silence – but we don’t have the evidence.

Paucity of first century non-Christian sources on Jesus: Hopper is open that his area of expertise is in other fields, so we can pass over an unawareness of the issues around the general lack of data we have for antiquity. But to illustrate his thoughts: “The activities of a religious fanatic who moved around Galilee and Judaea preaching a gospel of peace and salvation, was said to have performed miracles, was followed by crowds of thousands of adoring disciples, and within the space of a few hours invaded the hallowed grounds of the Temple, was hauled up before the Sanhedrin, tried by King Herod, interrogated by Pontius Pilate and crucified, all amid public tumult, made no impression on history-writers of the period” (page 167). I don’t know how Hopper fits invading the temple into those few hours. But anyway, a previous post I have written touches on the general problem of the paucity of ancient material in general, and can be found here,

Hopper writes “Outside the Gospels, there is no independent contemporary (I,e,, first century CE) account of these events.” This is about trying, mythicist-style, to exclude no fewer than four gospels from consideration as useful to historians of the period. And it completely overlooks New Testament accounts outside the Gospels (such as earlier evidence in the letters of Jesus’ younger contemporary Paul, which Hopper only accords a brief note). New Testament material inside and outside the gospels gives historians more to work with compared to many other significant figures in antiquity. Whatever might be said of his views on this, this is not linguistics.


Appendix 4: reliance on the eleventh century textus receptus 

As mentioned above, any test is fundamentally dependent on its evidence base. Any test of the TF has a problem as to which version of it to rely upon.

To illustrate the issues, other scholars (e.g. Levenson and Martin in their 2014 work) have been examining the TF’s Latin translations that are older than the Greek textus receptus. Any attempt at dissecting the TF linguistically ought to take some account of the work of such scholars in order to potentially get closer to the text of Antiquities as it could have been earlier. Otherwise, while we may be analysing the textus receptus, we may not be analysing the TF as it was known in, say, the fourth century  or earlier.

Alice Whealey has indicated the hazards of relying solely on the textus receptus in her historiographical study The Testimonium Flavianum Controversy from Antiquity to the Present (given at the 2000 SBL Josephus Seminar). In short, ancient translations into Latin and Syriac independently corroborate each other to reveal that there was a different, earlier Greek text of the TF. (She notes that “Latin and Syriac writers did not read each others’ works in late antiquity. Both, however, had access to Greek works.”)

Their shared variant reading - “he was believed/supposed to be the Christ” (significantly different from the textus receptus' “he was the Christ”) - appears both in a Latin translation used by Jerome; and also in a Syriac translation (suspected to be by 7th century James of Edessa, and quoted in a 12th century chronicle, which scholar Shlomo Pines gets the credit for turning up). It is clear that the text changed between "he was the Christ" and “he was believed to be the Christ” in one direction or the other. This is a loose end that could be a key to help resolve the problem of the TF's authenticity. There is no need for “he was believed to be the Christ” to exist unless it goes back to an authentic source - it is the more likely original. By the by, since it is found in these two branches that stem from a Greek original, it would be hopeless to claim that Jerome made it up to doctor the text of Eusebius (as some have done).

Whealey concludes that Jerome “knew the original version of the Testimonium, which he probably found in Eusebius’ Historia Ecclesiastica, which read “he was believed to be the Christ” rather than “he was the Christ.”” Eusebius’ own text has to have been tampered with after those translations obtained, and tampering with the Antiquities consequently happened to harmonise. Thus, prior to tampering, the Greek text read: “he was believed to be the Christ”. This explains why some problematic things in the TF which don’t fit with Josephus' non-Christian outlook are found in the late Greek textus receptus of both Eusebius and Josephus, whereas the earlier translations are free of "he was the Christ" and should be relied on in this regard. Triangulating the evidence leads here, to a redacted version. 


Nothing in Hopper's essay overturns this, since he does not follow the evidence trail away from the textus receptus



[1] As for linguistic points in favour of partial authenticity, there is a helpful summary by Tim O’Neill here. With Tim I disagree on some things, but I respect his diligence!
[2] By and large, in terms of reception of his essay, anything I can find approving of Hopper’s article tends to stem from so-called mythicists who doubt or deny the existence of any historical Jesus, a quite extreme position. Hopper is not wholly in that camp. He says that the crucifixion is the “one true event” in the TF, so Hopper is not an absolute mythicist. However, he cites Price and Doherty (mythicists) as authorities for doubting the historicity of Jesus. (See his footnote 4). Setting aside that the historicity of Jesus is surely not the matter at hand, if Hopper must open up this question, why is he not referencing a more representative range of scholarly contributions?
[3] Those categories can be subdivided in principle; but that is a moot point here as such subdivisions cannot be robustly taken from texts as short as the TF. In longer passages, various verb forms could serve, with complexity potentially increasing, whereas there is less scope for complexity in a passage as short as the TF.
[4]
Reid, Wallis (1977) The Quantitative Validation of a Grammatical Hypothesis: The Passé Simple and the Imparfait.Proceedings of the Seventh Annual Meeting of the North Eastern Linguistic Society
. Eds. Judy Anne Kegl, David Nash and Annie Zaenen. Cambridge, Mass.: North East Linguistic Society. 31533. And

Diver, William (1969) The System of Relevance of the Homeric Verb.Acta Linguistica Hafniensia 12: 4568.

[5] I am noting but setting aside for present purposes the valid question of whether the Pilate passages supply a wide enough Josephan sample for analysis - it is the evidence base for Hopper’s core argument.
[6] To reiterate, Hopper’s results are based on comparing the TF with the other three Pilate episodes. What if the TF were compared more closely with other passages in Josephus? Might we get a different result? But I am parking that here, as this is a reviews of what Hopper does say.
[7] Hopper makes similar sweeping claim on page 165: “The Testimonium is anchored in a radically different discourse community from that of the rest of the Jewish Antiquities. [Emphasis added] The Testimonium reads more like a position paper, a party manifesto, than a narrative.” Again, “of the rest of the Jewish Antiquities”? Has Hopper actually done the full surveys that he seems to be claiming? Perhaps this comment is predicated on his own conclusion that the TF is of the genre of Christian creeds.
[8] For example, it is worth noting that Hopper does not allow that there might be any reason for an author such as Josephus to vary his writing style here. But it is difficult to see why we should deny Josephus the right to do so. In fact, there is a known pattern of reasons why he might. Some of Josephus’ digressions are akin to modern footnotes, a modern literary device unavailable to ancient authors. Footnotes had not been invented. Books were published on long scrolls. Therefore, things that we would add as footnotes to keep a text a smooth read were not so conveniently treated in ancient texts. They pop up obtrusively in Josephus’ writing and that of other ancient writers, interrupting narratives. Hopper, a linguist, seems to have been unaware of this literary convention, which is somewhat unfortunate, since he is trying to convince us of what Josephus ought to write. The TF is not untypical of Josephan digressions. In this light, it is unlikely to be correct that verb forms alone can bear the weight of determining authenticity – it could just be Josephus footnoting, so to speak. But in fact, the TF is integrated into the pattern of narratives in ways that Hopper does not note.

[9] Curiously, Hopper also includes epēgageto as an action of Christians (“many … followed”), whereas other translators find it to be an action of Christ (“he gained a following”).


[10] At most, I and others would read these words as Josephus displaying a modicum of irritation that the Christians have not disappeared, as they possibly should have done after the defeat of their leader.
[11] There is a similar example of Hopper Christianising the meaning of the text: “on indictment by the first men among us, Pilate having sentenced him to the cross…” in Hopper’s hands is taken to indicate that Pilate is a ‘compliant puppet of the Jewish hierarchy’. This interpretation is out of proportion to the actual text, and has to be a meaning imported by Hopper from Christian Gospels such as Matthew, or later Christians influenced by that stream. And tellingly, it's to the latter that Hopper refers, and without such a Christian reference point, it is doubtful that Hopper would have come to such a stark reading, although he tries to locate it squarely in the TF (page 167). This is methodologically unsound, inserting foreign overtones into a text which Hopper is supposed to be analysing linguistically. Rather weakening this giant leap, in this sentence, both the Jews and Pilate are in subordinate clauses. Hopper seems to overlook (but Fernando-Rubio doesn’t) that “on indictment by the first men among us, Pilate having sentenced him to the cross…” sounds like legal process, with a sly hint of Christ being at fault – only the worst people get the cross.
[12] Hopper sides with a familiar motivation for interpolation, “presumably by Christians embarrassed at Josephus’s manifest ignorance of the life and death of Jesus.” (page 167) The idea that Josephus’ “ignorance” is “manifest” is merely a reiteration of Hopper’s judgment. It shows no reflection on the passage mentioning Jesus and his brother James in Josephus’ Antiquities Book 20 (which Hopper never mentions); nor of Origen’s definite perception that Josephus was an unbeliever (and Hopper actually cites Feldman making the latter point. How has Hopper, who has read Feldman, overlooked it?).

[13] This little mention of Pilate in the TF is unlike, say, the most well-known Christian creed that reduces Pilate’s prominence further: “[Jesus] was crucified under Pontius Pilate”.


[14] Also in the second and third passages, we find a tidy ending about the fate of the enemy. The second ends with “And thus an end was put to this sedition”. The third ends with “the tribe of Christians… are not extinct at this day”. It’s a juxtaposition of endings, and deepens the sense of Josephus' irritation that the Christians aren't extinct.  
[15] Scholars often suggest that a Christian hand removed something like “they reported that” from the resurrection wording.

[16] Most notably, third century theologian Origen, doesn’t cite the TF, but does refer emphatically to Josephus as not believing that Jesus was the Christ, which indicates squarely that Josephus did write something which left Origen with that impression. There may be good reason for Origen not discussing the TF where he was engaged in rebutting his pagan opponent Celsus. This could be so if perhaps the original TF was not favourable to Christians (which Bermejo-Rubio explores). The thing is, if Origen were trying to rebut Celsus blow by blow, it might not be very strategic in debate to invoke something that was more useful to his opponent. This could invite his opponent to deliver a blow to which Origen didn't have a matching defence.