Some great debates over ancient texts are like der fliegende Hollander, never coming to rest, but touching shore now and then to make a new appearance before a bewildered audience. A prime example of such an academic poltergeist is the famous (or notorious) Testimonium Flavianum, the supposed "testimony" to Jesus' ministry and death in Flavius Josephus' The Jewish Antiquities (written cat A.D. 93-94). [1] The text, found in Ant. 18.3.3 (Section 63-64), [2] has generated a debate over its authenticity that encompasses every position imaginable, from total acceptance to total rejection. The purpose of this present essay is
To "clear the ground" for the argument over the Testimonium, it will be helpful to review quickly the two other key passages where Josephus is supposed - at least by some - to have spoken of Jesus. One text, in my view, can be dismissed as spurious without further ado; the other needs closer examination. [3]
The clearly unauthentic text is a long interpolation found only in the Old Russian (popularly known as the "Slavonic") version of The Jewish War, surviving in Russian and Rumanian manuscripts. [4] This passage is a wildly garbled condensation of various Gospel events, seasoned with the sort of bizarre legendary expansions found in apocryphal Gospels and Acts of the second and third centuries. Despite the spirited and ingenious attempt of Robert Eisler in the 1920s and 1930s to defend the authenticity of much of the Jesus material in the Slavonic Jewish War, almost all critics today discount his theory. In more recent decades, G. A. Williamson stood in a hopeless minority when he tried to maintain the authenticity of this and similar interpolations, which obviously come from a Christian hand (though not necessarily an orthodox one). [5]
Not so easily dismissed is a reference to James, the brother of Jesus, in book 20 of The Jewish Antiquities. This short passage occurs in a context where Josephus has just described the death of the procurator Festus and the appointment of Albinus as his successor (A.D. 62). While Albinus is still on his way to Palestine, the high priest Ananus the Younger convenes the Sanhedrin without the procurator's consent and has certain enemies put to death. The key passage (Ant. 20.9. [Section 200]) reads:
Being therefore this kind of person [i.e., a heartless Sadducee], Ananus, thinking that he had a favorable opportunity because Festus had died and Albinus was still on his way, called a meeting [literally, "sanhedrin'] of judges and brought into it the brother of Jesus-who-is-called-Messiah [ton adelphon Iesou tou legomenou Christou], James by name, and some others. He made the accusation that they had transgressed the law, and he handed them over to be stoned. [6]There are a number of intriguing points about this short passage. First of all, unlike the text about Jesus from the Slavonic Josephus, this narrative is found in the main Greek-manuscript tradition of The Antiquities without any notable variation. The early 4th-century church historian Eusebius also quotes this passage from Josephus in his Ecclesiastical History (2.23.22).
Second, unlike the extensive review of Jesus' ministry in the Slavonic Josephus, we have here only a passing, almost blase reference to someone called James, whom Josephus obviously considers a minor character. He is mentioned only because his illegal execution causes Ananus to be deposed. But since "James" (actually, the Greek form of the English name James is Jakobos - Jacob) is so common in Jewish usage and in Josephus' writings, Josephus needs some designation to specify which Jacob/James he is talking about. [7] Josephus apparently knows of no pedigree (e.g., "James the son of Joseph") he can use to identify this James; hence he is forced to identify him by his more well-known brother, Jesus, who in turn is specified as that particular Jesus "who-is-called-Messiah."
This leads to a third significant point: the way the text identifies James is not likely to have come from a Christian hand or even a Christian source. Neither the NT nor early Christian writers spoke of James of Jerusalem in a matter-of-fact way as "the brother of Jesus" (ho adelphos lesou), but rather - with the reverence we would expect - "the brother of the Lord" (ho adelphos tou kyriou) or "the brother of the Savior" (ho adelphos tou soteros). Paul, who was not overly fond of James, calls him "the brother of the Lord" in Gal 1:19 and no doubt is thinking especially of him when he speaks of "the brothers of the Lord" in 1 Cor 9:5. Hegesippus, the second-century church historian who was a Jewish convert and probably hailed from Palestine, likewise speaks of "James, the brother of the Lord" (in Eusebius' Ecclesiastical History 2.23.4); [8] indeed, Hegesippus also speaks of certain other well-known Palestinian Christians as "a cousin of the Lord" (4.22.4), "the brothers of the Savior" (3.32.5), and "his [the Lord's] brother according to the flesh" (3.20.1). The point of all this is that Josephus' designation of James as "the brother of Jesus" squares neither with NT nor with early patristic usage, and so does not likely come from the hand of a Christian interpolator. [9]
Fourth, the likelihood of the text coming from Josephus and not an early Christian is increased by the fact that Josephus' account of James' martyrdom differs in time and manner from that of Hegesippus. Josephus has James stoned to death by order of the high priest Ananus around A.D. 62, a good while before the Jewish War actually breaks out. According to Hegesippus, the scribes and Pharisees cast James down from the battlement of the Jerusalem temple. They begin to stone him but are constrained by a priest: finally a laundryman clubs James to death (Ecclesiastical History (2.23.12-18). James' martyrdom, says Hegesippus, was followed immediately by Vespasian's siege of Jerusalem (A.D. 70). Eusebius stresses that Hegesippust account agrees basically with that of the Church Father Clement of Alexandria (2.23.3,19); hence it was apparently the standard Christian story. Once again, it is highly unlikely that Josephus' version is the result of Christian editing of The Jewish Antiquities.
Fifth, there is also the glaring difference between the long, legendary, and edifying (for Christians) account from Hegesippus and the short, matter-of-fact statement of Josephus, who is interested in the illegal behavior of Ananus, not the faith and virtue of James. In fact, Josephus never tells us why James was the object of Ananus, wrath, unless being the "brother of Jesus-who-is-called-Messiah" is thought to be enough of a crime. Praise of James is notably lacking; he is one victim among several, not a glorious martyr dying alone in the spotlight. [10] Also telling is the swipe at the "heartless" or "ruthless" Sadducees by the pro-Pharisaic Josephus; indeed, Josephus' more negative view of the Sadducees is one of the notable shifts from The Jewish War that characterize The Antiquities. [11] In short, it is not surprising that the great Josephus scholar Louis L. Feldman notes: "...few have doubted the genuineness of this passage on James." [12]
As it stands in the Greek text of The Antiquities (the so-called "vulgate" text), the Testimonium reads thus:
At this time there appeared [18] Jesus, a wise man, if indeed one should call him a man. For he was a doer of startling deeds, a teacher of people who receive the truth with pleasure. [19] And he gained a following both among many Jews and among many of Greek origin. He was the Messiah. And when Pilate, because of an accusation made by the leading men among us, condemned him to the cross, those who had loved him previously did not cease to do so. [20] For he appeared to them on the third day, living again, just as the divine prophets had spoken of these and countless other wondrous things about him. And up until this very day the tribe of Christians, named after him, has not died out. [21]At first glance, three passages within the Testimonium strike one as obviously Christian:
At this time there appeared Jesus, a wise man. For he was a doer of startling deeds, a teacher of people who receive the truth with pleasure. And he gained a following both among many Jews and among many of Greek origin. And when [or better: although] Pilate, because of an accusation made by the leading men among us, condemned him to the cross, those who had loved him previously did not cease to do so. And up until this very day the tribe of Christians (named after him) has not died out.The flow of thought is clear. Josephus calls Jesus by the generic title "wise man" (sophos aner, perhaps the Hebrew hakam). He then proceeds to "unpack" that generic designation with two of its main components in the Greco-Roman world: miracle working and effective teaching. [32] This double display of "wisdom" wins him a large following among both Jews and Gentiles, and presumably - though no explicit reason is given - it is this huge success that moves the leading men to accuse him before Pilate. [33] Despite his shameful death on the cross, [34] his earlier adherents do not give up their loyalty to him, and so (note the transition that is much better without the reference to the resurrection) [35] the tribe of Christians has not yet died out.
But even if these deletions do uncover an earlier text, is there sufficient reason to claim that it comes from Josephus? The answer is yes; our initial, intuitive hypothesis can be confirmed by further considerations drawn from the text's history, context, language, and thought.
First of all, unlike the passage about Jesus in the Slavonic Jewish War, the Testimonium is present in all the Greek manuscripts and in all the numerous manuscripts of the Latin translation, made by the school of Cassiodorus in the sixth century; variant versions in Arabic and Syriac have recently been added to the large inventory of indirect witnesses. [36] These facts must be balanced, however, by the sobering realization that we have only three Greek manuscripts of book 18 of The Antiqutities, the earliest of which dates from the eleventh century. One must also come to terms with the strange silence about the Testimonium in the Church Fathers before Eusebius. [37] I will return to this point at the end of the essay.
Second, once we have decided that the reference to "the brother of Jesus-who-is-called-Christ, James by name" is an authentic part of the text in book 20, some earlier reference to Jesus becomes a priori likely. Significantly, in Ant. 20.9.1 Josephus thinks that, to explain who James is, it is sufficient to relate him to "Jesus-who-is-called-Messiah." Josephus does not feel that he must stop to explain who this Jesus is; he is presumed to be the known fixed point that helps locate James on the map. None of this would make any sense to Josephus' audience unless Josephus had previously introduced Jesus and explained something about him. [38] Of course, this does not prove that the text we have isolated in Ant. 18.3.3 is the original one, but it does make probable that some reference to Jesus stood here in the authentic text of The Antiquities.
Third, the vocabulary and grammar of the passage (after the clearly Christian material is removed) cohere well with Josephus' style and language; the same cannot be said when the text's vocabulary and grammar are compared with that of the NT. [39] Indeed, many key words and phrases in the Testimonium are either absent from the NT or are used there in an entirely different sense; in contrast, almost every word in the core of the Testimonium is found elsewhere in Josephus - in fact, most of the vocabulary turns out to be characteristic of Josephus. [40] As for what I identify as the Christian insertions, all the words in those three passages occur at least once in the NT. [41] One must beware of claiming too much, though. Some of the words in the interpolations occur in the NT only once, and some occur more often in Josephus. Still, the main point stands: In that part of the Testimonium which - on other grounds - seems to come from Josephus and not from a Christian, the vocabulary and style jibe well with that of Josephus, [42] and not quite so well with that of the N f. This distinction between the vocabulary of Josephus and the vocabulary of the NT does not hold in the three passages I identify as Christian interpolations. There, the text shows more of an affinity to the NT vocabulary than does the rest of the Testimonium.
This comparison of vocabulary between Josephus and the NT does not provide a neat solution to the problem of authenticity, but it does force us to ask which of two possible scenarios is more probable. Did a Christian of some unknown century so immerse himself in the vocabulary and style of Josephus that, without the aid of any modern dictionaries and concordances, he was able to
An intriguing allied question is whence Josephus drew his information. Thackeray leaves open the possibility that in Rome Josephus had met Luke or read his work. [54] But, as we have seen, the language of the Testimonium is not markedly that of the NT. Of course, it is possible that Josephus had known some Christian Jews in Palestine before the Jewish War; it is even more likely that he had met or heard about Christians after taking up residence in Rome. Feldman notes that, as a-ward of the Flavians, with residence and pension provided for his work, Josephus would have had easy access to the archives of provincial administrators that were kept in the imperial court at Rome. [55] Was there an account of Jesus' trial among the records? An interesting surmise, but impossible to verify. Martin prefers to think that Josephus recounts the common opinion he heard among the educated, "enlightened" Jews of the partially Romanized world he inhabited. [56] Finally, one cannot exclude - no more than one can prove - that, even apart from direct contact with Christian Jews, Josephus learned some basic facts about Jesus in Palestine before the outbreak of the Jewish War. In short, all opinions on the question of Josephus' source remain equally possible because they remain equally unverifiable.
In the end, there is something astounding - some would say incredible - in my modest proposal. Jesus of Nazareth was a marginal Jew in a marginal province at the far eastern end of the Roman Empire. It is amazing that a more prominent Jew of the first century, in no way connected with this marginal Jew's followers, should have preserved a thumbnail sketch of "Jesus-who-is-called-Messiah." Yet practically no one is astounded or refuses to believe that in the same book 18 of The Jewish Antiquities Josephus also chose to write a longer sketch of another marginal Jew, another peculiar religious leader in Palestine, "John surnamed the Baptist" (Ant. 18.5.2 [section 116-19]). Fortunately for us, Josephus had more than a passing interest in marginal Jews. [60]
-----[1] For initial orientation in the vast literature on Josephus, see Emil Schurer, The History of the Jewish People in the Age of Jesus Christ (175 B. C. -A. D. 135) (rev. and ed. by Geza Vermes and Fergus Millar; Edinburgh: Clark, 1973) 1. 61-63 The fullest bibliographies and reviews of literature on Josephus available today have been drawn up by Louis H. Feldman; see his "Flavius Josephus Revisited: The Man, His Writings, and His Significance,"Aufstieg und Niedergang der romischen Welt (ed. W. Haase and H. Temporini; Berlin/New York: de Gruyter, 1984) II/21.2, esp. 822-35; idem, Josephus and Modern Scholarship 1937-1980 (Berlin/New York: de Gruyter, 1984), esp. 679-703 for the Testimonium Flavianum; idem, Josephus. A Supplementary Bibliography (New York/London: Garland, 1986). See also the older bibliography by Heinz Schreckenberg, Bibliographie zu Flavius Josephus (ALGHJ 1; Leiden: Brill, 1968); by following the marginal indication of number 17, one can trace the development of the argument over the authenticity of the Testimonium Flavianum from 1655 (T. Faber) down to 1965 (A. Pelletier). Major studies in Josephus that are also sources of bibliography include Harold W. Attridge, Jr., The Interpretation of Biblical History in the Antiquitates Judaicae of Flavius Josephus (HDR 7; Missoula, MT: Scholars, 1976); Shaye J. D. Cohen, Josephus in Galilee and Rome His Vita and Development as a Historian (Columbia Studies in the Classical Tradition 8; Leiden: Brill, 1979); Tessa Rajak, Josephus. The Historian and His Society (Philadelphia: Fortress, 1984); Louis H. Feldman and Gohei Hata (eds.), Josephus, Judaism, and Christianity (Detroit: Wayne State University, 1987); Per Bilde, Flavius Josephus between Jerusalem and Rome. His Life, His Works, and Their Importance (Journal for the Study of the Pseudepigrapha, Supplement Series 2; Sheffield: JSOT, 1988).
[2] The standard Greek-English edition used in the United States today is that of the Loeb Classical Library (Cambridge, MA: Harvard University; London: Heinemann, 1926-65), encompassing ten volumes (Nos. 186, 203, 210, 242, 281, 326, 365, 410, 433, and 456); the editors through the years have been H. St. J. Thackeray, Ralph Marcus, and Louis H. Feldman. For the key passages in books 18 and 20 of The Antiquities, I have also consulted the critical edition of Benedict Niese, Flavii Josephi Opera (7 vole.; 2d ed.; Berlin: Weidmann, 1955; originally 1885-95) 4. 150-53 and 308-11. English translations of The Jewish War include the Penguin edition by G. A. Williamson, Josephus. The Jewish War (rev. ea.; Harmondsworth: Penguin, 1970); and the beautifully laid out edition with commentary, illustrations, and photographs presenting the archaeological background: Josephus. The Jewish War (ed. G. Cornfeld, B. Mazar, R Maier; Grand Rapids: Zondervan, 1982).
[3] On the question of Christian interpolations into Josephus in particular and Jewish writings in general, see Eva Matthews Sanford, "Propaganda and Censorship in the Transmission of Josephus," TAPA 66 (1935) 127-45; James H. Charlesworth, "Christian and Jewish Self-Definition in Light of the Christian Additions to the Apocryphal Writings," Jewish and Christian Self-Definition. Volume Two. Aspects of Judaism in the Graeco-Roman Period (ed. E. P Sanders et al.; Philadelphia: Fortress, 1981) 27-55. Charlesworth (p. 28) notes that Christian interpolations into Jewish writings range from very minor insertions through blocks of material to massive rewriting of the whole document. As will become clear, I hold that the Christian interpolations in the Testimonium Flavianum are of the first, relatively minor, type; Charlesworth, however, thinks that the Testimonium is "full of Christian additions" (p. 27).
[4] The point of insertion of the Jesus interpolation is J.W. 2.9.2. (Section 169), where Pontius Pilate is introduced into the narrative.
[5] For G. A. Williamson's opinion on the Slavonic version, see his The World of Josephus (Boston/Toronto: Little, Brown, and Company, 1964) 308-9. His position, though not well defended, has had an influence beyond its merits because it was enshrined in an appendix of the Penguin edition of Josephus. The Jewish War. The text of the main "Jesus interpolation" (though Jesus is never mentioned in it by name), along with other pertinent insertions, can be conveniently found in the appendix of Williamson's translation, pp. 397-401. (In fairness, it should be noted that recent Penguin editions of The Jewish War omit Williamson's defense of the Slavonic version.) Another convenient compendium, which supplies the text of all the principal additional passages in the Slavonic Jewish War, can be found in vol. 3 (No. 210) of the Loeb Library edition of Josephus' works. But for a full exposition and defense of the Slavonic version one must have recourse to the once famous, now largely forgotten volumes of Robert Eisler, IESOUS BASILEUS OU BASILEUSAS (2 vols.; Heidelberg: Winter, 1929, 1930); the English translation by Alexander Krappe (The Messiah Jesus and John the Baptist [London: Methuen & Co., 1931]) contains only parts of the German original. On Eisler, see Schurer, The History of the Jewish People, 1. 60-61; E. Bammel, "The Revolution Theory from Reimarus to Brandon," Jesus and the Politics of His Day (ed. E. Bammel and C. F. D. Moule; Cambridge: Cambridge University, 1984) 11-68, esp. 32-37. Bammel thinks that underlying the text in the Slavonic Josephus is a Jewish account redacted by a Christian; the basis of the reference to Jesus is Sanh. 43a in the Talmud; similar developments can be seen in the apocryphal Acts of Pilate and the Jewish Aramaic life of Jesus, the Toledoth Jeshu. Bammel reaches this conclusion about the Slavonic Josephus account of Jesus: "Its evaluation is only possible if its Sitz im Leben in the Jewish-Christian controversy is recognised and the direct link with Josephus and the first centuries abandoned" (p. 33). Paul Winter, in his excursus on Jesus in Josephus in the revised Schurer (The History of the Jewish People, 1. 440), argues that the Slavonic statement on Jesus must have been composed after the Testimonium Flavianum reached its present form. A survey of recent thought on the Slavonic version can be found in Feldman, "Flavius Josephus Revisited," 771-74.
An attempt to salvage something about the original, political Jesus from the Slavonic Josephus was made by J. Spencer Kennard, "Gleanings from the Slavonic Josephus Controversy," JQR 39 (1948-49) 161-70. It was rightly rejected by Solomon Zeitlin ("The Hoax of the 'Slavonic Josephus,"' JQR 39 [1948-49] 171-80), who concludes: "All the Jesus passages are interpolations based on Christian literature." Indeed, Kennard himself admits that portions of the Slavonic text are based on the Church Fathers and the Acts of Pilate. The appeal to supposed oral Jewish traditions reflected in the writings of Augustine simply lacks evidentiary support; worse still, Kennard, according to Zeilin, depends on false translations of the Slavonic text. Kennard himself later admitted his error and recanted his position that the arrest of Jesus in the Slavonic Josephus might contain an element of genuine historical tradition; see his "Slavonic Josephus: A Retraction," JQR 39 (1948-49) 281-83. Zeitlin continued his critique of the use of Slavonic Josephus in "The Slavonic Josephus and the Dead Sea Scrolls: An Expose of Recent Fairy Tales," JQF 58 (1967-68) 173-203. He shows that both external and internal criticism points to the Slavonic Josephus being a compilation of the 11th century. A 10th- or 11th-century date is also accepted by J. Strugnell, "Josephus, Flavius," NCE 7 (1967) 1120-23.
On the other hand, Zeitlin, in my view, dismisses the two Jesus passages in the Greek text of The Antiquities without sufficient reasons (pp. 172-74). Zeitlin's disciple, Ellis Rivkin, accepts the reference to Jesus in Ant. 20 but rejects it in Ant. 18; see his What Crucified Jesus? (Nashville: Abingdon, 1984) 64-67.
[6] The translations from Josephus are my own.
[7] Abraham Schalit (Namenworterbuch zu Flavius Josephus [Leiden: Brill, 1968] s.v.) counts five different people named lakobos who are mentioned by Josephus in his various works.
[8] This is also the form Eusebius himself uses (see 2.23.1), but we need not suspect that Eusebius has foisted his reverential form onto the quotation of Hegesippus. When Eusebius claims to be citing Josephus, he is faithful to Josephus'un-Christian form, "the brother of Jesus" (Ecclesiastical History 2.23.20,22 - the latter passage being a direct citation of Ant. 20.9.1).
[9] Paul Winter, in his excursus on the disputed Josephus passages in the revised Schurer ( The History of the Jewish People, 1. 428-41), also argues that "if a Christian forger had inserted a reference to Jesus, he would scarcely have been content to mention Jesus in such noncommittal fashion" (p. 431).
[10] Douglas R. A. Hare (The Theme of Jewish Persecution of Christians in the Gospel According to St Matthew [SNTSMS 6; Cambridge: Cambridge University, 1967] 34) notes the vagueness of the accusation in Josephus. He suggests that Ananus was acting out of personal animosity and that the "certain others" killed at the same time were not Christians but other personal enemies of Ananus.
[11] On this, see Cohen, Josephus in Galilee and Rome, 237-38.
[12] In vol. 10 (no. 456) of the Loeb Library edition, p. 108 n. a; cf. his treatment in Josephus and Modern Scholarship, 704-7, where the vast majority of scholars affirm authenticity; see the similar judgment of Winter in Schurer's The History of the Jewish People, 430. Even Hans Conzelmann, who doubts the authenticity of the Testimonium Flavianum, accepts Ant. 20.9.1 (Jesus [Philadelphia: Fortress, 1973] 13-14). H. St. J. Thackeray states flatly ("Lecture VI. Josephus and Christianity," Josephus. The Atan and the Historian [The Hilda Stich Stroock Lectures; New York: Jewish Institute of Religion, 1929] 131) that, in his opinion, the authenticity of the two passages in The Antiquities on John the Baptist and on James is "beyond question."
[13] As Ernst Bammel shrewdly notes ("Zum Testimonium Flavianum (Jos Ant 18, 63-64)," Josephus-Studien. Untersuchungen zu Josephus, dem antiken Judentum und dem Neuen Testament [Otto Michel Festschrift; ed. Otto Betz, Klaus Haacker, and Martin Hengel; Gottingen: Vandenhoeck & Ruprecht, 1974] 9-22, on p. 10 n. 17), it is not always clear whether a particular scholar judges the Testimonium spurious or whether he simply thinks that it is impossible to reconstruct the original form of what Josephus said about Jesus.
[14] But see the (mostly uncritical) position of Franz Dornseiff, "Lukas der Schriftsteller. Mit einem Anhang: Josephus und Tacitus," ZNW 35 (1936) 129-55; idem, "Zum Testimonium Flavianum," ZNW 46 (1955) 245-50. Dornseiff maintains that the received Testimonium Flavianum in the Greek manuscripts of Josephus is entirely authentic, yet Josephus remained a Jew, not a Christian. The second article is written against Felix Scheidweiler ("Des Testimonium Flavianum," ZNW45 [1954] 230-43), who accepts a basic core as original, while suggesting that the received form may come from the circle of Paul of Samosata. A. Feuillet ("Les anciens historians profanes et la connaissance de Jesus," Esprit et Vie 87 [1977] 145-53) states that at least the greater part of the Testimonium is authentic; however, his comments on the individual "Christian interpolations" seem to favor the authenticity of the entire text.
More recently, the authenticity of the full Testimonium has been defended by Etienne Nodet, "Jesus et Jean-Baptiste selon Josephe," RR 92 (1985) 320-48, 497-524. His thesis labors under two difficulties: (1) ho christos houtos en must be read as "he was Christ [taken purely as a proper name, not as a set title of the Messiah]." The main problem with this interpretation is the definite article ho. which makes the titular sense ("the Messiah") the more likely interpretation, especially when the predicate nominative is thrown before the copulative verb. (2) All the other phrases that many interpreters judge to be "too Christian" are said to be simply ambiguous, and intentionally so in the mind of Josephus. The neutral tone of Josephus may indeed be purposely ambiguous or guarded, but some of the Christian phrases, especially those referring to the appearance of the living Jesus after his death and the testimony of the divine prophets are too straightforward to be explained in this way. The thesis is not helped by the further claim (p. 524) that Josephus carefully arranged the material in book 18 of The Antiquities for a polemical purpose: to separate John the Baptist from Jesus and thus to deprive the Baptist of his status as the forerunner of Jesus. In Nodet's view, Luke responded to Josephus' polemic in the Third Gospel and the Acts. A great deal is made to hang from very thin threads.
For a precritical defense of total authenticity, see William Whiston, "The Testimonies of Josephus Concerning Jesus Christ, John the Baptist, and James the Just, Vindicated," Josephus. Complete Works (Grand Rapids: Kregel, 1960; Ist ed. 1737) 639-47.
[15] Representative of total rejection - in the line of F. C. Burkitt and E. Norden - is Conzelmann (Jesus, 13-14), who claims that the passage "is constructed in accord with a pattern of the Christian kerygma (and indeed the Lukan one)" (p. 14). Unfortunately, he does not go on to substantiate his claim. One is left wondering what a short summary of Jesus' life by a Jew at the end of the first century would look like instead. Walter Potscher rejects Conzelmann's claim in "Josephus Flavius, Antiquitates 18,63f. (Sprachliche Form und thematischer Inhalt)," Eranos 73 (1975) 26-42, esp. pp. 27-28. Potscher points out the similarities between the Testimonium and "paradoxographic" authors like Apollonius, who wrote similar summaries of the careers of Epimenides and Hermotimus. The whole Testimonium is also rejected as an interpolation by Leon Herrmann, Chrestos. Temoignages paiens et juifs sur le christianisme du premier siecle (Collection Latomus 109; Brussels: Latomus, 1970) 97-98; one is not reassured, however, when one discovers that both the passage on John the Baptist and the mention of James in The Antiquities are likewise declared to be interpolations (pp. 99-104). Also in favor of a complete interpolation, though somewhat tentative in his judgment, is J. Neville Birdsall, "The Continuing Enigma of Josephus's Testimony about Jesus," BJRL 67 (1984-85) 609-22; similarly, Bilde, Flavius Josephus between Jerusalem and Rome. 223. It is wise to remember that Catholic as well as Protestant scholars were once given to total rejection of the Testimonium; Carlo M. Martini ("Il silenzio dei testimoni non cristiani su Gesu," La Civilta Cattolica 113/2 [1962] 341-49) reminds us that both P Battifol and M.-J. Lagrange judged the whole inauthentic (P. 347).
Perhaps symbolic of the shift from the total skepticism of the nineteenth century to the yes-but approach of the twentieth century was the change of opinion by the former "prince" of Josephan scholars, Thackeray. He had earlier held that the Testimonium was a complete forgery; he later came to believe that a core of authentic material could be reconstructed by removing or reformulating Christian interpolations. For the expression of his later view, see his "Josephus and Christianity," 125-53. One regrets that, under the heady influence of the contemporary works of Eisler, Thackeray adopted some emendations on the basis of the Slavonic Josephus, leaning in the direction of Eisler's Jesus the political rebel. Nevertheless, Thackeray does not accept Eisler's views in toto, and Feldman indicates that Thackeray recanted his pro-Eisler views before his death (Josephus and Modern Scholarship, 52). For another expression of the shift from complete skepticism to partial acceptance, see E J. Foakes Jackson, Josephus and the Jews (Grand Rapids: Baker, 1977; Ist ed. 1930) 279; see p. 278 for Jackson's curious view on why Josephus might have inserted the Testimonium into The Antiquities. Granted the notable swing away from total skepticism, it is surprising to read the claim of Michael J. Cook: "The overwhelming consensus among scholars today . . . is that Josephust celebrated testimony about Jesus is spurious" ("Jesus and the Pharisees-The Problem As It Stands Today," JES 15 [1978] 441-60, p. 451). But n. 45 on p. 451 indicates that Cook may be including judgments of partial inauthenticity in his sweeping statement.
[16] In his review of the literature, Feldman ("Flavius Josephus Revisited," 822) says flatly: " .. . the great majority of modern scholars have regarded it [the Testimonium] as partly interpolated, and this is my conclusion as well." C. Martin ("Le 'Testimonium Flavianum'. Vers une solution definitive?," Revue beige de philologic et d'histoire 20 [1941] 409-65) is likewise of the opinion that scholars increasingly favor the judgment of partial interpolation (p. 415). Some critics hold to the view that some parts of the Testimonium come from Josephus and other parts from an interpolator, but consider any attempt to reconstruct the genuine text as sheer speculation; so Zvi Baras, "the Testimonium Flavianum and the Martyrdom of James," Josephus, Judaism, and Christianity, 338-48, esp. p. 339; see also his earlier " Testimonium Flavianum: The State of Recent Scholarship," Society and Religion in the Second Temple Period (The World History of the Jewish People 1/8; ed. Michael Avi-Yonah and Zvi Baras; Jerusalem: Massada, 1977) 303-13, esp. p. 308. For the view that originally the Testimonium contained a hostile reference to Jesus that can no longer be reconstructed, see Clyde Pharr, "The Testimony of Josephus to Christianity," AJP 48 (1927) 137-47. For the view that the Testimonium originally made some unfavorable mention of Jesus at this point in The Antiquities, and that we can at least attempt a reconstruction, see Bammel, "Zum Testimonium Flavianum," 9-22.
[17] See Winter in the revised Schurer, The History of the Jewish People, 432-41; Feldman, "Josephus Revisited," 822-35; S. G. E Brandon, The Trial of Jesus of Nazareth (Historical Trials Series; New York: Stein and Day, 1968) 151-52; Morton Smith, Jesus the Magician (New York: Harper & Row, 1978) 45-46; James H. Charlesworth, Jesus Within Judaism (Anchor Bible Reference Library; New York: Doubleday, 1988) 90-98. In addition to Winter's excursus, a cogent defense of an authentic core going back to Josephus can be found in Louis H. Feldman, "the Testimonium Flavianum: The State of the Question," Christological Perspectives. Essays in Honor of Harvey K. McArthur (ea. R. F Berkey and S. A. Edwards; New York: Pilgrim, 1982) 179-99. See also Martini, "11 silenzio," 347; Wolfgang Trilling, Fragen zur Geschichtlichkeit Jesu (Dusseldorf: Patmos, 1966) 51 -56; A. -M. Dubarle, "Le tEmoignage de Josephe sur Jesus d'apres la tradition indirecte," RB 80 (1973) 481-513, and his update, "Le temoignage de Josephe sur Jesus d'apres des publications recentes," RB 84 (1977) 38-55.
[18] I take the sense of the much disputed ginetai to be "come, appear on the scene, arise," not unlike the sense of egeneto anthropos in John 1:6 (perhaps echoing the Hebrew wayhi, used to begin a narrative; cf. 1 Sam 1:1). A number of translations prefer "there lived," which is possible; but the translation "was born" seems out of place here, Karl A. Rengstorf (A Complete Concordance to Flavius Josephus [4 vole.; Leiden: Brill, 1973-83] s.v.) illustrates the vast range of meanings ginomai can have in Josephus, including "to be born, to arise, to be, to live, to have lived, to have existed." That Josephus uses ginetai de elsewhere in The Jewish Antiquities as an incipit of an narrative is documented by Martin, "Le 'Testimonium Flavianum,"'433.
[19] The Greek phrase ton hedone talethe dechomenon could imply simple-minded enthusiasm, even self-delusion. Yet, while possible, that is not necessarily the sense. We may have here one example of what Josephus is doing throughout the Testimonium: carefully writing an ambiguous text that different audiences could take in different ways.
[20] This is probably the meaning of epausanto; other possibilities are "did not give him up" (i.e., desert him) or "did not cease" (i.e., to exist). The last option would make the final sentence of the Testimonium redundant; and since, in Josephus as in the rest of ancient Greek, pauomai is often completed by a supplementary participle, it is most natural to understand agapesantes (or agapontes) from hoi to proton agapesantes; see William Goodwin and Charles Gulick, Greek Grammar (Boston: Ginn & Co., 1958) 33; also Bammel, "Zum Testimonium Flavianum," 14.
[21] The Greek text, with brackets supplied for those sections I consider Christian additions (see below), is as follows: ginetai de kata touton ton chronon lesous sophos aner, [edge andra auton legein chre,1 en gar paradoxon ergon poietes, didaskalos anthropon ton hedone talethe dechomenon, kai pollous men Ioudaious, pollous de kai tou Hellenikou epegageto; [ho christos houtos en.]kai auton endeixei ton proton andron par'hemin stauro epitetimekotos Pilatou ouk epausanto hoi to proton agapesantes; [ephane gar autois triten echon hemeran palin zon ton theion propheton tauta te kai alla myria peri autou thaumasia eirekoton.] eis eti te nyn ton Christianon apo toude onomasmenon ouk epelipe to phylon. Consult Feldman's notes in the Loeb Library edition (pp. 48-50) for the major emendations suggested by the editors. The translations in the text are my own.
[22] See Andre Pelletier, "Ce que Josephe a dit de Jesus (Ant. XVIII 63-64)," REJ 124 (1965) 9-21, esp. 14.
[23] Dubarle, "Le temoignage de Josephe sur Jesus d'apres la tradition indirecte," 481-513.
[24] It is curious that Martin, though preferring to view edge andra auton legein chre as an interpolation, remains hesitant ("Le 'Testimonium Flavianum'," 450-52), especially since he interprets Josephus as saying that Jesus seduced both Jews and Gentiles and thus brought upon himself the condemnation due an agitator (pp. 445-46).
[25] See Pelletier, "Ce que Josephe a dit," 14.
[26] See Dubarle, "Le temoignage de Josephe sur Jesus d'apres la tradition indirecte," esp. p. 495. Among the many others who favor some reading like legomenos with ho Christos is G. C. Richards, "the Testimonium of Josephus," JTS 42 (1941) 70-71. Williamson (The World of Josephus, 308) tries to save the authenticity of the sentence by claiming that the definite article before Christos is just the untranslatable article before a proper name; hence the sense is not "this man was the Messiah," but rather "the man I have called Jesus was the man commonly called Christ." This explanation ignores the three precise NT parallels, where the sense of houtos estin ho Christos is obviously "this man is the Messiah." If Josephus had wanted to say what Williamson wants to make him say, there were much clearer ways of doing so.
[27] Pelletier, "Ce que Josiphe a dit," 9-21; a variant of this article can be found in "L'originalite de temoignage de Flavius Josephe sur Jesus," RSR 52 (1964) 177-203; his position is accepted by Henri Cazelles, Naissance de l'eglise. Secte juive rejetee? (Paris: Cerf, 1968) 106. Pelletier points to Ant. 17.5.1. (section 87), where Josephus explains the name of the port Sebastos by saying that "Herod, having constructed it at great expense, named it Sebastos in honor of Caesar." Josephus leaves out the logically necessary explanation that Caesar's honorific name is in Latin Augustus, which was translated into Greek as Sebastos. Similarly, it may be argued, Josephus presupposes in Ant. 18.3.3. that the cultured reader knows that Jesus'"second name" was Christos. That cultured Roman readers would indeed have known this equivalence is made likely by the statement of the pagan Tacitus in his Annales (15.44) some two decades later, when he simply uses Christus as the name of the man Pilate put to death, without ever mentioning the proper name Jesus.
[28] On this point, see Martin, "Le 'Testimonium Flavianum,"'452-55.
[29] Curiously, a similar view is found in the theories of Eisler and Thackeray, who read ephane gar autois etc. as oratio obliqua in the original text (Thackeray, "Josephus and Christianity," 147).
[30] Dubarle, "Le temoignage de Josephe sur Jesus d'apres la tradition indirecte," esp. p. 499. I do not mean to disparage the great contribution Dubarle makes in providing easy access to so many patristic citations of the Testimonium. He himself admits the variety of the indirect tradition and the liberty the patristic writers take with the text (p. 502). I do not share his confidence that taking a majority vote among the diverging patristic witnesses gives us a good chance of restoring the original text. Despite his protests at the analogy, one would never think of going about NT text criticism in this way; I do not think the method is more valid when applied to the Testimonium.
[31] Strange to say, the position I am advocating is a relatively rare one in Josephan research. One of the few well-known scholars who approximates my position is Joseph Klausner, Jesus of Nazareth. His Life, Times, and Teaching (New York: Macmillan, 1925) 55-60. But Klausner does not supply detailed argumentation for his view, and he remains hesitant as to whether something else originally stood in the place of the three interpolations. I instead feel convinced that the deletion of these three passages restores the Testimonium to its original form. The scholar who comes closest to my own view is Martin ("Le 'Testimonium Flavianum'," 409-65), whose work was not available to me until this article was almost finished. I am encouraged by finding independent support for my own stance. At the same time, I find a number of Martin's positions curious or improbable; I will note my disagreements at the appropriate places.
[32] Martin ("Le 'Testimonium Flavianum'," 440) misses this point by making en gar paradoxon ergon poietes merely a parenthetical explanation of sophos aner, instead of seeing it as coequal with didaskalos anthropon as the two-part definition of sophos aner. He is quite correct, however, in emphasizing that in Josephus, as in many other Greek authors, sophos could convey the idea of a possessor of supernatural knowledge and power (pp. 442-44).
[33] The causal link would be clearer if we could be sure that epegageto meant "seduced, led astray. " That is a possible, though not necessary, meaning of the verb. Once again, Josephus may be engaging in a studied ambiguity.
[34] As Martin points out ("Le 'Testimonium Flavianum'," 446), the genitive absolute epitetimekotos Pilatou has the force of a concessive clause: "Although Pilate condemned him to the cross, nevertheless those who had loved him previously did not cease to do so."
[35] Indeed, one could argue that the te in eis eti te nyn demands ouk epausanto hoi to proton agapesantes just before it. It is almost impossible to make sense of the telltale te following upon the reference to the resurrection appearance and the OT prophecies; cf. Martin, "Le 'Testimonium Flavianum'," 441.
[36] For a discussion of the manuscript tradition and the confusing references in the Church Fathers, see Feldman, "the Testimonium Flavianum," 181-85; also David Flusser, "Der lateinische Josephus und der hebraische Josippon," Josephus-Studien, 122-32. On the Latin translation Feldman makes an intriguing observation: "Despite the obvious importance of the Latin version, since it antedates by half a millennium the earliest extant Greek manuscript, it has not been edited, except for the first five books of the Antiqurties and the treatise Against Apion since 1524.... The main manuscript of the Latin version, the Ambrosianus papyraceus, dates from the ninth century, two centuries before the earliest of the Greek manuscripts" (p. 289 n. 12). For the history of the tradition of the Latin translation, see Franz Blatt, The Latin Josephus. I. Introduction and Text. The Antiquities: Books 1-V (Acta Jutlandica 30/ 1, Humanities Series 44; Aarhus: Universitetsforlaget, 1958) 9-116.
On the Arabic and Syriac texts see Shlomo Pines, An Arabic Version of the Testimonium Flavianum and Rs Implications (Publications of the Israel Academy of Sciences and Humanities, Section of Humanities; Jerusalem: The Israel Academy of Sciences and Elumanities, 1971); Dubarle, "Le temoignage de Josephe sur Jesus d'apres IB tradition indirecte," 48:5-86. An English translation of the Arabic text of Agapius, with a brief discussion, is readily available in the Cornfeld edition of Josephus. The Jewish War, 510. Feldman (Josephus and Modern Scholarship,, 701) believes that Agapius used both Josephus and other sources and combined them: "We may . . . conclude that Agapiust excerpt is hardly decisive, since it contains several elements, notably changes in order, that indicate that it is a paraphrase rather than a translation." Nodet ("Jesus et Jean-Baptiste," 335-36) thinks that Agapius represents a deformed tradition of the Eusebius text found in the Ecclesiastical History. Personally, I am doubtful that this 10th century Arabic manuscript preserves the original form of the Testimonium, especially since it contains sentences that, as I have just argued, are probably later expansions or variants of the text. Indeed, Bammel ("Zum Testimonium Flavianum," 21) thinks that the text of the Testimonium was much more subject to summaries and expansions in the East than in the West. But, at the very least, the text from Agapius does expand our knowledge of the indirect witnesses to the text. Zvi Baras ("the Testimonium Flavianum and the Martyrdom of James," Josephus, Judaism, and Christianity, 338-48) judges (p. 340): "Although it is not considered the genuine text of Josephus, it provides us with an example of another more moderate attempt at the Christianization of the original text, probably unfavorable to Christians, and one which may have been known to Origen and Jerome." As will become clear, I do not share the latter part of this judgment.
[37] One possible explanation of this silence would jibe well with my reconstruction of the Testimonium and my isolation of the Christian interpolations. If until shortly before the time of Eusebius the Testimonium lacked the three Christian interpolations I have bracketed, the church fathers would not have been overly eager to cite it; for it hardly supports mainline Christian belief in Jesus as the Son of God who rose from the dead. This would explain why Origen in the third century affirmed that Josephus did not believe Jesus to be the Messiah ( Commentary on Matthew 10. 17; Contra Celsum 1.47). Origen's text of the Testimonium lacked the interpolations; and, without them, the Testimonium simply testified, in Christian eyes, to Josephus' unbelief - not exactly a useful apologetic tool in addressing pagans or a useful polemical tool in christological controversies among Christians. Indeed, if something like the text I reconstruct was not in Origen's copy of Josephus, one is left asking the question: What was it in the text that caused Origen to state apodictically that Josephus did not believe Jesus to be the Christ? The James passage in book 20 is not a sufficient reason.
[38] So, e.g., Richards, "the Testimonium of Josephus," 70-71. Some translators even take lesou tou legomenou Christou (Ant. 20.9.1) to mean "Jesus, the aforementioned Messiah,"-not a necessary meaning, though a possible one in this context (see Feldman, "the Testimonium Flavianum," 192). If this were the correct translation, it would help argue for some reference to messiahship in the original form of the Testimonium: hence, some hold that krome's reading, "credebatur esse Christus" ("he was believed to be the Messiah"), reflects part of the original text. However, since by the time of Josephus Christians had made Christos practically a second proper name of Jesus, Josephus may be presupposing in book 20 that the Jesus mentioned in book 18 is generally known as Christos. Of course, none of this speculation is necessary. Josephus may be introducing the title for the first (and only) time in The Antiquities in book 20. In this case, the phrase does mean "Jesus, who is called the Messiah," perhaps intimating why James (whose activity is otherwise not described) is a source of aggravation to the high priest.
[39] Here we face a basic methodological problem. Since we do not know who the Christian interpolator(s) were, or whether indeed the interpolations are anything more than random marginal notes that entered the text at different times, it is impossible to compare Josephus' vocabulary and style with any one Christian author. The NT is chosen for the sake of providing some fixed point of contrast; it is one corpus of first-century Christian works, most of which were written within the span of five decades, just as Josephus supplies us with a corpus of first-century Jewish works, all written within the span of three decades. The choice of the NT is not totally arbitrary, since presumably a Christian theologian or scribe of the first few centuries would be steeped in NT thought and vocabulary and would naturally narrate the story of Jesus out of his NT background.
[40] As Thackeray states, " . . . practically the whole of the language [of the Testimonium] can be illustrated from Josephus" ("Josephus and Christianity,', 141). For Thackeray, it is the criterion of style that turns the scale in favor of authenticity: "If the text has been mutilated and modified, there is at least a Josephan basis" (ibid.). For an examination of Josephus' vocabulary, the indispensable tool is Karl H. Rengstorf, A Complete Concordance to Flavius Josephus (4 vole.; Leiden: Brill, 1973-83), along with Schalit, Namenworterbuch zu Flavius Josephus. A study of the vocabulary and style of the Testimonium as compared to that of both the NT and Josephus can be found in the appendix at the end of this article. For a treatment of the philological argument from a slightly different angle, see Martin, "Le 'Testimonium Flavianum'," 432-39.
[41] In what I identify as the three Christian interpolations, the occurrence of the words in the NT is as follows: (1) In the first interpolation, edge 5x; aner 216x; chre 1x. (2) In the second interpolation, christos 531x (versus the one occurrence in Josephus outside the Testimonium); indeed, the whole second interpolation seems lifted directly from the NT. (3) In the third interpolation, phaino 31x, mostly in passive deponent forms, and often in the aorist passive; ephane is used of angelic appearances in Matt 1:10; 2:13,19; of Elijah in Luke 9:8; and of the risen Jesus in Mark 16:9. te trite hemera is the common NT way of referring to the resurrection on the third day. What is unusual is the construction triten echon hemeran, which lacks a precise parallel in Josephus as well. Both the NT and Josephus do use echo with cardinal numerals to express duration of time and age. Two NT texts that partially echo the usage here are Luke 24:21, triten tauten hemeran agei; and John 11:17, heuren auton tessaras ede hemeras echonta en to mnemeio. The rest of the vocabulary in the third interpolation offers no problem: zao, 140x, often as a present participle; palin, 141 x; theios, 3x; prophetes, 144x; myrias, 8x, myrioi (though not the form myria), 3x; thaumasios, 1x. When we consider the number of words and constructions in the core of the Testimonium that are not found in the NT, the total agreement of the interpolations with the vocabulary of the NT is striking. Of course, there are some stylistic quirks (e.g., the triten echon hemeran), pointing to a Christian interpolator or interpolators who naturally drew upon NT vocabulary as he (they) wrote about Jesus in his (their) own style. It must also be admitted that Josephus likewise uses all the words found in the three interpolations. In a few cases, the usage is more Josephan than that of the NT (e.g., theios used 206x, including with prophetes [though not in the plural]; thaumasios 31x). Hence, in the case of the three interpolations, the major argument against their authenticity is from content. Still, the difference from the core text is clear: in the core, not only are the vocabulary and style overwhelmingly Josephan, but at least some of the vocabulary is absent from the NT and some of the content is at variance with what the NT says.
[42] For an opposing view of the argument from "Josephan" vocabulary, see Birdsall, "The Continuing Enigma," 619-21. In my view, Birdsall applies the measuring rod of Josephan vocabulary and style a little too mechanically. The undisputed epistles of Paul have their share not only of hapax legomena but also of Pauline words and phrases that Paul uses in a given passage with an unusual meaning or construction. Especially since Josephus is dealing in the Testimonium with peculiar material, drawn perhaps from a special source, we need not be surprised if his usage differs slightly at a few points.
More specifically, I do not think all of Birdsall's claims concerning divergent style or meaning stand up to close scrutiny. For example, Birdsall claims that talethe only once elsewhere in Josephus carries the abstract sense of truth; the sense usually is the objective truth of a statement or report about a man's character. Actually, this is a fine distinction that does not always work well in practice. More to the point, there are a number of cases in Josephus where talethe means truth in some global or general sense: e.g., J.W.1.15 (Section 16); Ant. 8.2.1 (Section 23); 14.1.1 (Section 3). In another usage as well ("receive with gladness"), Birdsall has to admit that there is at least one other example in Josephus of the phrase being used in the same sense as in the Testimonium. I fail to see how this warrants a judgment of inauthentic.
Furthermore, Birdsall presumes that he knows the precise tenor or nuance of all of the phrases in the Testimonium. But that is far from certain. It is quite possible that the old fox Josephus purposely wrote a neutral or ambiguous statement about Jesus that could be read in more than one way. Simply as a matter of fact, the Christian interpolator(s) read it one way, while Eisler and his followers read it in a totally different way.
Finally, if Birdsall wishes to attribute the whole of the Testimonium to the pre-Nicene patristic period, then he must explain why some notable phrases from the core do not appear in any of the Church Fathers before Eusebius - e.g.,(1) sophos aner used to describe Jesus; and (2) phylon used to describe Christians. Indeed, while G. W. H. Lampe lists applications of sophos to God, OT heroes such as Daniel, Esther, and Judith, the authors of the wisdom books, St. Paul, and others possessed of Christian wisdom, he catalogues no application to Jesus; see his A Patristic Lexicon (Oxford: Clarendon Press, 1961).
[43] Notice that Josephus does not say outright that Jesus was a teacher of the truth; rather, he was a teacher of people who receive the truth with pleasure. As Pelletier points out from similar passages in Josephus ("Ce que Josephe a dit," 18), the phrase "receive the truth with pleasure" refers to the subjective "good faith" of those listening, not necessarily the objective truth of what the speaker propounds.
[44] Some critics suggest that Eusebius, or least a source only a few decades earlier than Eusebius, is the author of the Testimonium; see Zeitlin, "The Hoax of the 'Slavonic Josephus'," 171-80; and Kennard, "Gleanings," 161-63, who thinks that the received text of the Testimonium dates from a few decades before Eusebius. Pelletier, citing C. Martin, mentions Origen as a possible author of the Christian interpolations ("Ce que Josephe a dit," 14-15); Pelletier also suggests the possibility of a Christian polemicist of the fourth century.
[45] In light of the Testimonium's diffusion among the "main stream" Fathers of the Church from the fourth century onwards, it seems highly unlikely that it would have been composed by some remnant of an Ebionite or other suspect group.
[46] The question, of course, may be put the other way round: If the three statements I have bracketed are later Christian interpolations, what were the intention and the theology of the later interpolator? Actually, I think it more likely that we are dealing with random Christian glosses which secondarily crept into the text; they need not all be from the same hand or the same time. Hence to ask about the one purpose or theology of the three interpolations may be a mistake from the start. For the many marginal glosses in the manuscripts of Josephus, see Sanford, "Propaganda and Censorship," 132: "The process of interpolation [into the text of Josephus] was furthered by the intrusions into the text of glosses and adversaria by Christian students studying and copying Josephus. No existing manuscript of his work is free from such interpolation."
[47] One should not make of the royal official in John 4:46-53 a pagan centurion (contra John Marsh, The Gospel of St John [Harmondsworth: Penguin, 1968] 235-41); that is to read into John what only the Synoptics say in a notably different version of the miracle story. The Samaritans of chap. 4 serve as a half-way house between the Jewish and pagan worlds.
[48] Dr. Shaye J. D. Cohen, of Jewish Theological Seminary in New York, was kind enough to read this essay and make some comments. He saved me from more than one embarrassing slip. Dr. Cohen felt that the particular argument I make in point (a) is weak because later Christians had no hesitation in projecting back on Jesus the theology and ideology of Paul: of course Jesus had come to bring salvation to the Gentiles - and this is what the Testimonium says. How many premodern (and preecumenical) Christians would have said otherwise? Cohen shrewdly pointed out to me that what would have been unusual for an ancient Christian author was the statement that Jesus won over many Jews as well as many Gentiles. Cohen's point is well taken, but I think there is a difference between an anachronistic attribution to Jesus of an intention to save both Jew and Gentile and a statement about what Jesus de facto did during his public ministry, i.e., a description of his public activity (and its results) that simply flies in the face of what the four canonical Gospels say.
[49] Hence, Thackeray's claim ("Josephus and Christianity," 148) that phylon is distinctly disparaging and cannot have come from a Christian hand is not correct. See instead Solomon Zeitlin, "The Christ Passage in Josephus," JQR 18 (1927-28) 231-53, esp pp. 238-39, 253; Zeitlin points to Eusebius' Ecclesiastical History 3.33.3: to ton Christianon phylon.
[50] This point is stressed especially by Nodet, "Jesus et Jean-Baptiste," 320-48, 497-524.
[51] For example, by changing sophos aner to sophistes (used of leaders of factions Josephus dislikes) or talethe to taethe (the unusual, the abnormal, the bizarre); so Eisler, IESOUS BASILEUS, 1. 46-88; S. G. E Brandon, Jesus and the Zealots (Manchester: Manchester University, 1967) 359-68; similarly, Smith, Jesus the Magician, 45-46; Potscher, "Josephus Flavius, Antiquitates 18,63f.," 26-42. The changes Eisler introduces can be found in vol. 9 (No. 433) of the Loeb Library edition of Josephus' works, p. 48 n. a.
Baras ("the Testimonium Flavianum and the Martyrdom of James,', 340) correctly observes that if the original text of the Testimonium had been derogatory or ironic, it would have called forth a strong denunciation from Origen. Hence I am also very skeptical of the theory that originally the Testimonium had a derogatory reference to Jesus' virginal conception; for such a position, which in my view fails for lack of any direct evidence, see Pharr, "The Testimony of Josephus to Christianity," 37-47. For an attempt to bolster Pharr's suggestion by appealing to the fourth-century pseudo-Hegesippus (and even this demands reading into the evidence), see Albert A. Bell, Jr., "Josephus the Satirist. A Clue to the Original Form of the Testimonium Flavianum?" JQR 67 (1966-67) 16-22. Bammel ("Zum Testimonium Flavianum," 18-22) thinks that the Testimonium originally expressed a negative view of Christians (less so of Christ), but that good method demands that the modern scholar keep conjectural emendations to a minimum. However, his own corrections (apatethentes or autos apatontes for agapesantes, apegageto for epegageto, the omission of en after ho christos houtos, and the insertion of phaskontes hoti) are based on tenuous arguments, and even they are much more complicated than the simple procedure of bracketing the three obviously Christian statements.
[52] I have purposely not spent any time on the objection that the Testimonium breaks the thread of the narrative in book 18; if one is interested in such a line of argumentation, one should see Thackeray, "Josephus and Christianity," 140-41 (where, in my view, he is much too dependent on Eisler). Perhaps the best insight in Thackeray's whole explanation is the simple observation: "Josephus was a patchwork writer" (p. 141). Cohen is blunter: "We have emphasized another aspect of Josephus' work: his inveterate sloppiness. Texts suitable for tendentious revision as well as passages which contradict his motives are sometimes left untouched. The narrative is frequently confused, obscure, and contradictory" (Josephus in Galilee and Rome, 233). In the present case, one wonders whether any greater link need exist for Josephus than the fact that the account of Jesus (who is crucified by Pilate) is preceded by a story about Pilate in which many Jews are killed (Ant. 18.3.2 [section 60-62]) and is followed by a story in which tricksters are punished by crucifixion (Ant. 18.3.4 [section 65-80]). Hence I think the lengthy attempt of Bammel ("Zum Testimonium Flavianum," 15-18) to explain the connections with what precedes and follows the Testimonium is beside the point. For a detailed refutation of Norden's claim that the Testimonium supposedly disrupts the narrative flow and thematic unity of the larger context, see Martin, "Le 'Testimonium Flavianum'," 422-31.
[53] As Thackeray observes, in view of the pervasive Josephan vocabulary and style, "alterations should be reduced to a minimum" ("Josephus and Christianity," 143). In my view, Thackeray himself undertook too many alterations.
[54] Ibid., 127-28; Thackeray, however, says no more than that there is "nothing improbable" in this. He also notes that Mark's Gospel would probably have been in circulation in Rome when Josephus wrote The Antiquities.
[55] Feldman, "the Testimonium Flavianum, 194-95.
[56] Martin, "Le `Testimonium Flavianum'," 450.
[57] Needless to say, such an answer cannot claim absolute certitude, but only that high degree of probability which is commonly called moral certitude.
[58] It is important to stress here that the basic fact of Jesus' existence and at least some stray facts about him remain unquestioned whether one accepts my reconstruction or any of the many others offered across the spectrum of scholarship in the twentieth century. If one prefers the position of the later Thackeray, Dornseiff, Dubarle, Pelletier, or Feldman, the basic point I make in the text remains the same.
[59] Hence I do not find the silence of Christian writers in the first three centuries as powerful an objection to authenticity as does Feldman ("Flavius Josephus Revisited," 822). In my opinion, it is unfortunate that Martin, in his otherwise fine article, tries to launch the hypothesis that Origen himself was the author of the interpolations ("Le'Testimonium Flavianum'," 458-65). Any attempt to name the author of the three interpolations seems futile since we must take seriously the possibility that each of the three interpolations was scribbled into the margin by a different Christian at a different time.
[60] It is a pleasure to express my sincere thanks to Profs. Shaye J. D. Cohen, Harold W. Attridge, Jr., and Myles M. Bourke for their generous help in the preparation of this article.
I. The Testimonium and the New Testament
To take only the most obvious examples of the differences between the usage of the NT and that of the core of the Testimonium: poietes is never used of Jesus in the NT; hellenikos is used only 1x, of the Greek language, in Rev 9:11; hedone is used only 5x in the NT, and always in a pejorative sense of sensual or illicit pleasure; epago occurs 3x, but never in the middle voice with the sense of "win over" or "gain a following"; the crasis form talethe is never used in the NT; nothing even vaguely like the phrase anthropon ton hedone talethe dechomenon is ever used in the NT to describe the disciples of Jesus or those who decide to follow Jesus; endeixis occurs 4x, always in Paul, and always with the general sense of showing, manifesting, demonstrating, proving, never with the formal judicial sense of accusation, charge, denunciation, indictment; protoi andres never occurs in the NT (the closest analogue is Luke 19:47, hoi protoi tou laou, designating not all the Jewish leaders but only the elders as distinct from the high priests and the scribes; even here the phrase occurs outside the Passion Narrative; cf. Acts 13:50; 17:4; 25:2; 28:7); epitimao is always used in the NT of verbal rebukes (most often in the mouth of Jesus), never of any physical punishment, and certainly not of crucifixion; onomazo is never used in the NT in the form of the perfect passive participle and never in reference to the naming of Christians (cf. the statement of Luke in Acts 11:26); epileipo is used in the NT only once, in the transitive sense of "fail," never in the intransitive sense of "disappear, die out"; phylon is never used in the NT; the NT instead uses phyle 31x, in the sense of either (1) a tribe or the twelve tribes of Israel or (2) all the nations or peoples of the earth; the first sense is sometimes used symbolically of Christians (the clearest case in Jas 1:1), but phyle is never used directly with Christianon or any equivalent (the disciples, the brethren, etc.). No one of these differences means all that much; but the accumulated evidence of all these differences may point to an author who is not taking his material from the NT.
II. The Testimonium and Josephus
The usage of Josephus presents a totally different picture. With regard to "wise man" (sophos aner), Josephus calls Solomon (Ant. 8.2.7 [section 53]) and likewise Daniel (Ant. 10.11.2 [section 237]) a "wise man," while no NT writing ever predicates the adjective "wise" (sophos) directly of Jesus. Kata touton ton chronon and similar time references with kata + chronon are frequent in Josephus (see, e.g., Ant. 13.2.3 [section 46]; 18.2.4 [section 39]), but the NT never uses kata + chronon in any connection; paradoxos occurs 50x in Josephus (1x in NT), including two cases where it modifies erga, as here (Ant. 9.7.6 [section 182]: thaumasta gar kai paradoxa epedeixato erga [of Elisha the prophet]; Ant. 12.2.8 [section 63]: ergon kainon kai paradoxon); paradoxon touto poion (Ant. 12.2.11 [section 87]) also occurs. Poietes is used elsewhere in Josephus only in the sense of "poet"; but Josephus (or his secretary, according to Thackeray) has a fondness for resolving a simple verb into two words: a noun expressing the agent and the auxiliary verb (e.g., krites einai for the simple krinein). Moreover, Josephus uses such cognates as poieteos, "that which is to be done," poiesis, "doing, causing" (as well as "poetry, poem"), and poietikos, "that which causes something" (as well as "poetic"). Didaskalos occurs 16x in Josephus (59x in NT); anthropos, needless to say, is extremely common in Josephus. Hedone is used 128x by Josephus (5x in NT), who uses it in both positive and negative senses; hedone in the dative is used in the sense of "with pleasure," "gladly," as here. (Thackeray sees in this "receive with pleasure" locution the hand of a particular secretary of Josephus, "the Thucydidean hack" ["Josephus and Christianity," 141, 144]; I prescind from theories of secretaries, since my only concern is to show that the language is Josephan. Moreover, Thackeray's position about helpers in the composition of The Antiquities has not been universally accepted. Heinz Schreckenberg [Rezeptionsgeschichtliche und textkritische Untersuchungen zu Flavius Josephus (ALGHJ 10; Leiden: Brill, 1977) 173-74] stresses Josephus' command of Greek and the unity of his style in The Antiquities. The differences in style that exist in The Antiquities are to be attributed to Josephus himself, who over many years wrote The Antiquities, using various sources and influenced by various classical authors he was reading.) Alethes occurs 150x in Josephus (26x in NT); more importantly, Josephus frequently uses the crasis forms talethes and talethe, which never occur in the NT (note the phrase talethe kai ta dikaia labon in Ant. 8.2.1 [section 23]). Dechomai is used 221x in Josephus (56x in NT). Indeed, language very similar to the Testimonium is found in Ant. 17.12.1 (section 329): ton anthropon to hedone dechomenon tous logous; moreover, hedone is used remarkably often with dechomai in Josephus (never in the NT). Hellenikos occurs in Josephus 25x (NT 1x). Epago is used 106x in Josephus (3x in NT); unlike the NT, Josephus does use it in the middle voice to mean "attract, persuade, motivate, induce"; see especially J.W. 7.6.1 (section 164): pollous eis apostasian epagagetai; and Ant. 17.12.1 (section 327): Ioudaion hoposois eis homilian aphiketo epegageto eis pistin. Endeixis is used once in Josephus in the sense of "denunciation, charge, accusation" (Ant. 19.9.16 [section 133]); but the NT never uses it in this sense. Hoi protoi andres occurs some 6x just in The Antiquities (allowing for different word order); slightly variant forms of the same phrase occur three more times in the same work. Especially interesting is the fact that there is a high concentration of this phrase or very similar phrases in book 18: 18.1.1 (section 7); 18.4.4 (section 98,99); 18.5.3 (section 121); 18.9.9 (section 376). While there is no exact equivalent to the full phrase ton proton andron par' hemin, Ant. 18.9.9 reads: ton proton par' hekaterois andron, "the leading men on both sides"; closer still are Ant. 20.1.1 (section 2): chords gnomes tes ton proton par' autois, "without the consent of their leaders"; Ant. 20.9.1 (section 198): hoper oudeni synebe ton par'hemin archiereon, "which happened to none of our high priests"; andAg. Ap. 2.12 (section 136): peri de ton par'hemin andron gegonoton, "but concerning our own men of past history." (For further examples, see Eisler, IESOUS BASILEUS, 73.) Epitimao is used 13x in Josephus (29x in NT), including a number of instances where it means "to sentence" or "to fix (a punishment)," a sense never found in the NT; note the perfect active forms in Ant. 18.5.6 (section 183); Ag. Ap. 2.32 (section 239). The one difference from the other texts is the unusual accusative of the person and dative of the punishment, but such a construction would be unusual in any Greek author; and amid so many examples of typical Josephan style some minor variations are to be expected. Eisler (ibid., 74), noting the similarity to Tacitus' reference to Jesus' death (Annales 15.44), suggests a literal translation of a technical Latin juridical phrase. Agapao in Josephus can mean "to like, love, value, welcome; to content oneself; to tolerate, accept." Passages dealing with God, parents, children, justice, etc. make it clear that Josephus does use the word in the sense of affective love; vaguely reminiscent of our passage is J.W. 1.1.12 (section 30): tois ge ten aletheian agaposin, alla me pros hedonen anegrapsa. Hence it is hardly justified to claim that agapesantes is "specifically and technically Christian" (so A. von Gutschmid, quoted by Bammel, "Zum Testimonium Flavianum," 11 n. 22). Thackeray ("Josephus and Christianity," 147) prefers here the sense of "to be content with," but only because he insists on seeing an originally hostile sense in the Testimonium. Pauo occurs 170x in Josephus, in both active and middle, the meanings of the middle including the sense witnessed here, "to stop (doing something), to give up"; the NT has only one example of the active and 14 of the middle, the middle regularly meaning "to stop (doing something)." The words eis, eti, te, and nyn are all very common in Josephus, but their conjunction here is unusual. The Testimonium is the only passage in Josephus where Christianoi appears; but then it occurs only 3x in the NT. Onomazo occurs in Josephus 62x (10x in NT). Josephus often uses the phrase onomazo apo tinos in the sense of "to name something after someone," which occurs in the passive as well as the active voice; see, e.g., Ant. 1.15 (section 241); 1.6.1 (section 123); note the exact form of the perfect passive participle in Ag. Ap. 1.26 (section 245). The onomazo apo tinos construction does not occur in the NT. Hode is used frequently by Josephus. While epileipo occurs only 1x in the NT, in the sense of "time will fail me" (Heb 11:32), it occurs 40x in Josephus, including the sense of "stop, cease, become extinct, die out"; note Ant. 2.9.3 (section 210): dedios hyper tou pantos ethnous, me spanei tes epitraphesomenes neotetos epileipe. While Josephus uses phyle more often than phylon, he does use the latter 11x, usually in the sense of "tribe, people, nation"; note J.W. 2.15.4 (section 397): anairesein de pan hymon tophylon.
III. Conclusion to Appendix
The upshot of all this is that, apart from Christianon, not one word of what I identify as the original text of the Testimonium fails to occur elsewhere in Josephus, usually with the same meaning and/or construction. As indicated in the first part of this note, the same cannot be said of the NT. One final caution: Word statistics are used here simply to indicate whether a word occurs fairly frequently either in Josephus or in the NT. The statistics should not be used simplistically to contrast the two bodies of material taken as a whole. Josephus represents a larger corpus of works than does the NT, and he is one well-known author as opposed to numerous NT authors, many of whom are anonymous.