The Testimonium Flavianum Canonicum: Josephus as a Witness to the Biblical Canon, 1566–1823

It therefore naturally, or rather necessarily, follows (seeing that with us it is not open to everybody to write the records, and that there is no discrepancy in what is written; seeing that, on the contrary, the prophets alone had this privilege, obtaining their knowledge of the most remote and ancient history through the inspiration which they owed to God, and committing to writing a clear account of the events of their own time just as they occurred) – it follows, I say, that we do not possess myriads of inconsistent books, conflicting with each other. Our books, those which are justly accredited, are but two and twenty, and contain the record of all time. Of these, five are the books of Moses, comprising the laws and the traditional history from the birth of man down to the death of the lawgiver. This period falls only a little short of three thousand years. From the death of Moses until Artaxerxes, who succeeded Xerxes as king of Persia, the prophets subsequent to Moses wrote the history of the events of their own times in thirteen books. The remaining four books contain hymns to God and precepts for the conduct of human life. From Artaxerxes to our own time the complete history has been written, but has not been deemed worthy of equal credit with the earlier records, because of the failure of the exact succession of the prophets. We have given practical proof of our reverence for our own Scriptures. For, although such long ages have now passed, no one has ventured either to add or to remove, or to alter a syllable; and it is an instinct with every Jew, from the day of his birth, to regard them as the decrees of God, to abide by them, and if need be, cheerfully to die for them. Time and again ere now the sight has been witnessed of prisoners enduring tortures and death in every form in the theatres, rather than utter a single word against the laws and the allied documents. What Greek would endure as much for the same cause?¹

—Josephus, Contra Apionem I.37–44

Even after the heady discovery of the Dead Sea Scrolls in the mid-twentieth century and their subsequent study, the lines above still contain the earliest extant explicit witness to a closed canon of Jewish Scripture.² One might call this passage the Testimonium Flavianum Canonicum. Unlike the Testimonium Flavianum, a text long ascribed to Josephus that has likewise fascinated scholars for centuries, this testimony is not spurious.³ It presents a genuine historical testimony to the Jewish biblical corpus in the decades before and immediately following the destruction of the Second Temple in Jerusalem, and therewith the Bible of the first Christians. As a witness to the history of Jewish and Christian Scriptures, it is all the more valuable in that it predates the normative formulations of Scriptural canons of Rabbinic Judaism and of the early Church.

This essay takes a close look at a set of scholars in different cultural contexts and intellectual traditions and of different scholarly temperaments and confessional loyalties across the long early modern period, and asks what role this passage played in their biblical scholarship, and in particular, their study of the canon. What weight did Josephus’s description carry? What kind of witness was he taken to be? These scholars are Sisto da Siena (1520–1569), Robert Bellarmine (1542–1621), William Whitaker (1547/8–1595), John Cosin (1594–1672), Johannes Hoornbeek (1617–1666), Richard Simon (1638–1712) and Johann David Michaelis (1717–1791). They are each too sophisticated to be taken as neat representatives of a larger group, much less as points on a graph that connect to an unmistakeable linear trend. The selection is driven rather by a wish to present a wide panorama of ways Josephus was read across three centuries by scholars with a particular question in mind, to understand the weight he was given as a historical witness and to reconstruct the scholarly ends to which that testimony was put. Given that the passage in question consists of a mere seven lines in Josephus’s final work, this approach offers a way to trace one thread in the long and expansive tapestry of Josephus’s reception and, given the magnitude of modern scholarship on Josephus, a way to make virtue of a necessity.⁴

Contra Apionem and the Canon of Jewish Scriptures: Imprecisions and Ambiguities

‘The whole of Josephus’s narrative is permeated by the ambivalence which inevitably arose from [his] complex political career, first as a defender of Jerusalem, then an apologist for the regime that had destroyed it.’⁵ The same cannot be said of Against Apion. While the Jewish War may deserve as its subtitle ‘du bon usage de la trahison’ (Pierre Vidal-Naquet’s fine phrase), Against Apion is a meticulously argued and decidedly non-ambivalent defence of Judaism.⁶ Its style, structure and vocabulary serve a deliberate forensic rhetoric, carefully constructed to refute the claims of such detractors of the Jews as Manetho, Chaeremon, Lysimachus and Apion of Alexandria.⁷ These men had pointed to the fact that Greek literature contains no evidence for the antiquity of the Jews, and they had argued from that silence that the Jews must have no antiquity to speak of. They had aimed all manner of defamation against Jewish texts as dependable historical records and against Josephus’s own histories. Against Apion is a frontal counter-attack against these accusations, and a robust defence of the superiority of the Jews’s constitution and of the antiquity, accuracy and veracity of their historical records. As such, it is as much a treatise on historical method as an apology for Judaism.⁸ In the text itself, Josephus calls it an ἀντίρρησις (counter-statement) and an ἀπολογία (defense).⁹ John Barclay has called it ‘the finest sample of Judean apologetics from antiquity [that] stands as a brilliant finale to [Josephus’s] long literary career’, while given the extraordinary and incomparable conditions of Josephus’s life, Martin Goodman has argued against it being an exemplar of any wider phenomenon. No other Jew lived the life Josephus lived, and the work, like the life, is sui generis.¹⁰

Yet for all its meticulous construction and detailed argument, Josephus’s description of the Jewish scriptures is remarkably imprecise, and it raises several major questions.¹¹ Firstly, and most importantly, what were the twenty-two books in his canon? The Talmudic list of Biblical books (b. Baba Batra 14b–15a) will list not twenty-two but twenty-four, an enumeration consolidated by later rabbinic authorities.¹² Medieval Jewish scholars will turn the number itself into a name for the Hebrew Scriptures, speaking of ha-Esrim ve-Arba’ah (or Arba’ah ve-esrim), ‘the twenty-four.’ Was the Jewish canon in the first century CE smaller than the rabbinic Scriptures by two books? If so, by which two? Or did the canons of twenty-two and twenty-four books comprise the same books, differently counted?

Secondly, exactly what did Josephus mean by prophecy, if in his reckoning the prophets were historians, their books accurate, trustworthy histories from Moses to Artaxerxes? In Rabbinic tradition, there are eight prophetic books (Joshua, Judges, Samuel, Kings, Jeremiah, Ezekiel, Isaiah and the Twelve Minor Prophets numbered as one). Given that Josephus identified Xerxes’s son Artaxerxes with the biblical King Ahasveros (Antiquities XI.184), he would seem to have included Esther among his thirteen unnamed prophetical books along with Daniel, Chronicles and Ezra (and Nehemiah), all of which the Rabbis consign to the non-prophetic writings (Ketuvim, Hagiographa). Josephus leaves his final four books unnamed, describing them merely as ‘hymns to God and precepts for the conduct of human life.’ In Rabbinic tradition, the Ketuvim comprise eleven books (Ruth, Psalms, Job, Proverbs, Ecclesiastes, Song of Songs, Lamentations, Daniel, Esther, Ezra, Chronicles). Several of these, including Ruth, Daniel and Chronicles, Josephus could not conceivably have described as hymns but must have considered, by his own criteria, as prophetic histories. If Josephus’s final four comprised the Psalms, Proverbs, Ecclesiastes and the Song of Songs (which is likely), the remainder of his Scriptural canon – including Job – must somehow fit into the unnamed thirteen books he calls prophetical, and identified only by their terminus post quem non (Artaxerxes). If Daniel, Ezra and Nehemiah, Chronicles, Esther and Job all belonged to Josephus’s prophetic category, that still leaves six books in the third group rather than four. In Antiquities V Josephus mentioned Ruth and Samson together and in Antiquities X Jeremiah with Lamentations, couplings that will recur in early Christian authors, including Origen, Epiphanius, Jerome and Cyril. On this ground, S.Z. Leiman has argued for the probability that Josephus coupled Judges with Ruth and joined Lamentations to Jeremiah and that his twenty-two-book canon comprised all twenty-four books of what would become the Rabbinic canon, organized in as many numbers (twenty-two) as the Hebrew alphabet has letters.¹³ Probability, however high, is not certainty, and John Barton has insisted that ‘it is impossible to know exactly which books [Josephus] included, or how he counted them.’¹⁴

Thirdly, Josephus raises the question of the division of the canon. Like the tripartite Jewish TaNaKh, the Old Testament of most Christian traditions also comprises three parts, though categorical arrangements as well as order, both pregnant with theological significance, differed widely. In Rabbinic Judaism, the organisation of the parts must have stabilized at a relatively late date. If indeed Josephus’s canon included Daniel, for example, that book must have belonged among his prophets, while in the Talmudic canon it will belong among the Ketuvim. This is no trivial matter: the question whether Daniel was to be read prophetically – and if so, how – was of immense moment for both Jewish and Christian readers.¹⁵ This was certainly the case in the sixteenth and seventeenth centuries, when the question of canonicity would take on renewed urgency and different Christian confessions rearranged the books of the Old Testament to reflect their traditions and practices of prophetic reading.¹⁶

Canon, Crisis and Confessional Erudition

The authority and accuracy of the Christian Scriptures belong among the most fraught issues in the early modern rupture of Western Christianity. The rapid expansion of knowledge of Greek and Hebrew, the proliferation of new Latin and vernacular translations of the Scriptures, the enormous growth of biblical literacy, the continuous collection, study and dissemination of variant readings among Vulgate manuscripts, the rediscovery of ancient, non-European versions of Scripture: all these developments, accelerated and transformed by the new technology of print, contributed to a profound crisis surrounding the very texts that formed the cornerstone of the Church.¹⁷

But which texts were they? Logically prior to uncertainties about versions, readings and the particular manuscripts with which early modern scholars struggled stood an uncertainty about the constituent parts of Scripture itself. Which books were in, and which were out? In case of doubt, what authority determined canonicity? Were the Scriptures true and sacred because the Church said so, or were the Scriptures in and of themselves sufficient witness to their truth and holiness such that the authority of the Church derived from them instead? The question of the biblical canon had implications and consequences for the authority of tradition and of the Church far beyond this particular matter. To some at the frontlines of confessional battles, it seemed the most important question of all. As William Whitaker, the Cambridge divine to whom we will return below, reflected in 1588:

The question, therefore, between us and the papists is, whence [the Scriptures] have received such great authority, and what it is, and on what this whole weight of such divine dignity and authority depends. The subject is difficult and perplexed; nor do I know whether there is any other controversy between us of greater importance.¹⁸

Debates across the early modern period such as that about the authority of the Hebrew text vis-à-vis the Greek were bound up in the larger questions of whether texts not or no longer extant in Hebrew (Tobit, Judith, Wisdom, Ecclesiasticus, I and II Maccabees, Baruch) could be canonical at all. These questions, in turn, were compounded by the fact that tradition itself was not univocal. Church Fathers had disagreed about them; there was no consensus Patrum.

The denial of a book’s canonicity therefore entailed a denial of the Church’s authority and constituted sufficient grounds for the denier to be declared anathema. And it is not hard to see why, especially when a passage in the book in question contained a proof text for doctrine. For example, Christian tradition held II Maccabees 12:43–5, which describes Judah the Maccabees’s atonement for the dead in light of their resurrection, to be the proof text for the doctrine of purgatory. Luther’s Widerruf vom Fegefeuer (Disavowal of Purgatory, 1530) consequently argued from the non-canonicity of I and II Maccabees.¹⁹ As the Latin Church broke apart in the sixteenth century, the instability of the text, the instability of the canon and the instability of doctrine compounded each other.

Awareness of the canonical instability of the Old Testament never disappeared in the Latin West. Augustine’s On Christian Doctrine (II:8) left no doubt as to the contents of the Old Testament: it included Tobit, Judith, Wisdom, Ecclesiasticus and I and II Maccabees.²⁰ But Jerome took a diametrically opposed view, spelled out in his preface to the Books of Kings (including Samuel). He called this his ‘helmeted prologue’ (prologus galeatus), arming himself against the attacks he must have anticipated on account of the fact that he limited the canon of the Old Testament to twenty-two books, divided into ‘Law’, ‘Prophets’ and ‘Hagiographa’ (that is, TaNaKh), while relegating to a fourth category (‘inter apocrifa’) Tobit, Judith, Wisdom, Ecclesiasticus and I and II Maccabees (as well as the Shepherd of Hermas). ‘Non sunt in canone’, the saintly translator asserted.²¹

The canonical disagreement between Augustine and Jerome was not merely about one set of books versus another: it was about two completely different kinds of criteria for canonicity. For Augustine, without the Church to call it sacred, the Bible was essentially a collection of stories like any other. The criterion for sanctity was tradition. ‘In truth’, Augustine declared, ‘I would not believe the Gospel unless the authority of the Catholic Church moved me to.’²² For Jerome, by contrast, the criterion, at least for the Old Testament, was that the book in question should be part of the Hebrew Bible, not a book extant in a Greek translation only much less one written in Greek to begin with. Without mentioning Josephus by name, Jerome’s ‘helmeted prologue’ makes the analogy between the twenty-two letters of Hebrew alphabet and the twenty-two books of the Hebrew canon.

As William Horbury has shown, Jerome’s view, as well as the very term ‘apocrypha’ in the sense in which he used it, became widely known through medieval commentaries both on scripture (the Glossa Ordinaria) and on canon law (the glosses to Gratian’s Decretrum).²³ True, Isidore of Seville ascribed Jerome’s opinion to the Jews themselves rather than to the translator, while identifying that of Augustine with that of the Church. Nonetheless, both comparative study of patristic sources and familiarity with the ubiquitous Glossae left the question wide-open.²⁴ Two distinct canonical traditions emerged intact at the end of the Middle Ages, their irreconcilability unresolved: one Augustinian, inclusive of the entire Greek Old Testament and based on the authority of the Church, and one Hieronymian, in which only the books of the Hebrew Bible sacred to Jews were considered canonical parts of the Christian Old Testament.²⁵

This canonical instability pitted the authority of the Church against that of Jewish tradition, and it became one of the most inflammatory issues of the early Reformation. Did one or didn’t one need the Church to determine the authority and canonicity of a biblical book? Luther set the Apocrypha apart in his epochal 1534 German translation of the Old Testament, considering them unnecessary for salvation and merely ‘useful and good to read.’ Other Reformers followed his example.²⁶

The Roman Church countered the Reformers resolutely. In its session of April 8, 1546, the Council of Trent decreed the Latin Vulgate to be the only version of Scripture to be held publically authentic in the Church. But before it could do so, the Council needed to re-establish which books belonged to the Christian Scriptures in the first place. Hence, a Decretum de Canonicis Scripturis precedes the Decretum de Editione et Usu Sacrorum Librorum. Against Luther, Calvin and their followers, but ironically also against Jerome whose translation they placed above all other versions, the Council decreed the Apocrypha to be canonical.²⁷

In the wake of Trent, two rival canons emerged consolidated. The incommensurability of the Reformed and Counter-Reformed canons was but one among many cracks in a breaking Church, but it was a very deep one, branching out in countless directions. With the Tridentine decree of 1546, canonicity became one of the fault-lines of Western Christendom at large and of confessional scholarship in particular. The question of the authentic contents of the Old Testament would exacerbate confessional polemic for several centuries. It is in the context of this polemic that early modern scholars turned to the seven lines in Contra Apionem in which Josephus describes and defends the twenty-two-book canon of Jewish Scriptures.

Ecclesiastical history provided a critical backdrop for renewed interest in Josephus and his century. As the Council of Trent drew to a close, the writing of histories of the early Church became one of the primary arenas for confessional gladiators. Some of the most accomplished European scholars welded their erudition into interpretations of the Christian past that validated the historical and theological claims of their confession.²⁸ As an eyewitness to Palestine in the first Christian century and as a cultural intermediary between Jerusalem and Rome, no source was comparable to Josephus. Early modern historians and antiquarians turned to Josephus for different kinds of invaluable information: about the geography of the Holy Land, the topography of Jerusalem, the architecture of the Temple, the chronology of the Jewish Revolt in which Josephus himself had fought until his capture in 67, the military history of Rome under the Flavians, the constitution of the ‘Hebrew Republic’, as well as material for forgeries.²⁹ Within the large library of Josephus’s works, select passages drew particular attention and became charged with historical significance and scholarly interest.

As questions of biblical authority, textual stability and scriptural canonicity gained urgency throughout the sixteenth and seventeenth centuries, scholars turned to Josephus for another reason: as an extra-biblical source for knowledge about the history of the Bible itself. In so doing, they imbued one passage with singular importance for determining the contents of the Old Testament in the first Christian century: Contra Apionem I.37–43. The question of the formation of the canon of the Scriptures of the early Church welded biblical philology to ecclesiastical history in ways that would make this brief passage one of the central pieces of evidence in one of the most contested debates in early modern biblical learning.

In 1566, the Dominican censor Sisto da Siena published his Bibliotheca Sancta, arguably the most important Roman Catholic manual of biblical study to appear in the immediate aftermath of the Council of Trent.³⁰ In the opening sections of his book, Sisto surveys the different ways in which biblical books had been canonized, ordered and organized before ordering them anew himself. The very first witness he brings forward is Josephus’s description in Against Apion.³¹ Seeking to resolve the irreconcilable Augustinian and Hieronymian positions in the wake of the Council, Sisto introduces new categories of his own. The books of the Hebrew Bible he termed protocanonical; those Jerome had called apocrypha Sisto termed deuterocanonical, while the remainder fell under the rubric apocrypha (which by the early eighteenth-century would come to be called pseudepigrapha).³² To Sisto, the neologism deuterocanonical did not mean that they came after the canon, but rather that they had become canonical later than the others. Book four of the Bibliotheca Sancta consists of a long bio-bibliographical lexicon of ‘Catholic interpreters of Sacred Scriptures’ followed by a short ‘catalogue of Hebrew/Jewish or Rabbinic commentators, who commented on the sacred scriptures in the Jewish language and according to Hebrew traditions.’³³ Strikingly, Josephus’s place is in the first list. How had the Jewish priest, Jewish warrior and Jewish historian become a Father of the Church?

Sisto’s new classification – if not its terminology – was taken up by a work that Wilhelm Schmidt-Biggemann has called ‘by some length the most important work of theological controversy of the seventeenth century.’³⁴ Robert Bellarmine's Disputations on the Controversies of Christian Faith against the Heretics of this Time, the flagship of the Counter-Reformation armada, first began to be published in 1586, but Schmidt-Biggemann’s description is accurate in the sense that it elicited attacks, counter-attacks and new editions throughout the seventeenth century.³⁵ Bellarmine addresses the composition of the canon in the first part ‘On the Sacred and Apocryphal Scriptures’ of the first book, titled On the Word of God, in the fourth chapter. Here he sets Josephus alongside Melito of Sardis (as recorded in Eusebius’s Ecclesiastical History) and Hilary of Poitiers’s Preface to the Psalms as witnesses to the contents of the Old Testament Canon.

As Piet van Boxel shows in his contribution to this volume, Sisto and Bellarmine did not agree on every point, including the validity of Josephus as a historical witness to the canonicity of the additions to the Book of Esther found in the Greek Bible, sections of which occur in Josephus’s Antiquities, too. As readers of Jewish texts from antiquity and more recent times, Sisto and Bellarmine also represented a change in the Catholic Church’s attitudes toward Jewish learning in the wake of the burnings of the Talmud in the 1550s and 1560s and its inclusion in the indices of Paul IV (1559) and Pius IV (1564). Where Sisto had been sent to Cremona to burn the Talmud, Bellarmine, as van Boxel has shown elsewhere, was a pivotal figure in the shift from burning to censorship.³⁶ If the Rabbis come to Bellarmine mediated through censorship and expurgation, Josephus, as we shall see, is transmitted in part through a very different, patristic funnel, though, as van Boxel shows, that did not keep Bellarmine from disregarding the trustworthiness of parts of Josephus’s oeuvre that did not make it into Eusebius’s History.

The Church of England sought to forge a middle way. Following Luther it did not consider the Apocrypha (Sisto’s deuterocanonical books) to be canonical, yet when the thirty-nine articles were formulated in 1562, the sixth article followed the Würtemberg Confession, avoiding the term ‘apocrypha’ altogether. After listing those books about whose authority there had never been any doubt, the article speaks simply of ‘the other books’, a difference that could not but question their authority.³⁷

Among the anti-Catholic controversialists that Bellarmine’s Disputationes provoked into polemic was the above-mentioned Puritan William Whitaker, Regius Professor of Divinity at Cambridge and Master of St John’s College there.³⁸ It is significant that Whitaker’s 1588 attack on Bellarmine should focus entirely on Scripture, and that the first of his book’s six parts be an argument ‘On the number of the canonical books of Scripture.’³⁹ Whitaker mobilizes his witnesses in the service of two syllogisms: (1) All canonical books of the Old Testament are written by prophets; none of the disputed books is written by a prophet; therefore, none of the disputed books is canonical; and (2) The Old Church of the Jews received and approved all the canonical books of the Old Testament; the Church did not receive these books; ergo, they are not canonical. Adding more testimonies to his case, Whitaker summons the very sources that Bellarmine had cited at the relevant places in his Disputationes: the Council of Laodicea, Melito of Sardis and Origen as recorded by Eusebius, Athanasius, Hilary of Poitiers, Gregory of Nazianzen, Cyril of Jerusalem, Epiphanius and of course Jerome’s ‘helmeted prologue’, all giving the books of the Old Testament as twenty-two in number, ‘as the letters of the Hebrew alphabet’, as well as others, including Rufinus and Gregory the Great, who had written on the non-canonicity of Apocrypha. ‘To these authorities of the ancient fathers I shall subjoin the testimony of Josephus, which exactly agrees with them, as it lies in his first book against Apion the grammarian, and is transcribed by Eusebius in the tenth chapter of the third book of his Ecclesiastical History.’ At this point Whitaker cites the passage from Contra Apionem I to conclude that ‘Assuredly it is plain enough from this testimony of Josephus, what was the judgment of the Israelitish church concerning these books.’⁴⁰

Whitaker furnishes us twice with the reason that Josephus had come to take on the weight of a patristic source: in his Ecclesiastical History, Eusebius had collected numerous ancient writers who had testified to the contents of the biblical canon, primarily Melito of Sardis, Origen and Josephus. In citing Josephus’s description of the canon in Against Apion verbatim, Eusebius’s History dressed Josephus’s testimony in the robes of patristic authority, on which, in turn, was draped the lustre of his inclusion among Jerome’s Illustrious Men. Bellarmine had referred to all these sources, too: Whitaker fought the Jesuit from Montepulciano using the latter’s own weapons. It was a lesson Bellarmine never forgot. It is said, Whitaker’s biographer records, perhaps betraying his own sympathies, ‘that Bellarmine placed his [Whitaker’s] picture above his desk, rather as Montgomery was to keep an eye on Rommel.’⁴¹ In turn, other Jesuits would come to Bellarmine’s defence, to the point of disregarding patristic authority and associating Josephus’s Judaism with the ‘heresy’ of men like Whitaker. ‘We should not be amazed’, the Ingoldstadt Jesuit Jakob Gretser would sneer, ‘if a heretic listens more closely to the words of a Jew than to the consensus of the Christian and Catholic Church.’⁴²

Whitaker had argued his case in Latin, in seventy-seven pages of densely packed small type, with truncated references to patristic sources. It was a form suited for his intended readership, and numerous seventeenth-century theologians and biblical critics turned to his treatment of canonicity, including John Bois, Whitaker’s student and one of the Cambridge translators of the Apocrypha for King James.⁴³

The question of the canon, knowledge of its ancient sources and of the confessional debates about it reached a potentially much larger readership when they were taken up in the most elaborate early modern exploration of the development of the biblical canon written in English. The High Anglican John Cosin (1595–1672) wrote his Scholastical History of the Canon of the Holy Scripture (1657) during the Royalist exile in Paris, after he had been Master of Peterhouse (1635–1644) and Vice-Chancellor of the University of Cambridge (1639), while he was dean of Peterborough (1640–1660) and before his return and appointment as Bishop of Durham (1660–1672) and his famous 1662 revision of the Book of Common Prayer.⁴⁴ The vast range of sources cited in the Scholastical History – helpfully organized and catalogued in a set of thematic indices – shows Cosin participating directly and critically in the controversies discussed above. They also suggest he had access to a fine and up-to-date library – possibly even that of Cardinal Mazarin – while in Paris, an excellent if delicate vantage point from which to write an anti-Catholic book.⁴⁵

Cosin’s essay, written a century after the mid-sixteenth century upheavals, begins forensically. Before writing a word himself, he presents the two irreconcilable positions through their now classic formulations: the sixth of the thirty-nine articles and the Tridentine decree, though he gives the latter the polemical title ‘A New Canon of Scripture.’ Over the next 214 pages, he announces to his reader,

I give an Accompt of the Canonical and undubitate Books of Holy Scripture, as they are numbered in the VI Article of Religion set forth by the Church of England and have been received by the Catholick Church all severall Ages since the time of the Apostles, till the Church of Rome thought fit to compose and dresse up a New Additional Canon thereof for themselves in their late Councel of Trent … which they call an Oecumencial Councel, as if all the Bishops in Christendom had been there present and voted in it.

Working indefatigably, Cosin collected a very large number of ‘witnesses’, and arranged them according to century in an elaborate index, in the fashion of the Magdeburg Centuries.⁴⁶ Cosin’s second chapter, ‘The Testimony of the Ancient Judaical Church’ turns to the Jewish Scriptural canon. The ‘witnesses’ for their number and order are not Jewish but Christian: Jerome’s Prologus Galeatus and Sisto da Siena’s Bibliotheca Sancta.

In the meantime, another voice had entered the debate. In his Chronographia of 1581, the French Catholic biblicist and Hebrew scholar Gilbert Génébrard (1537–1597) had suggested the possibility that, besides the canonization of the Hebrew Scriptures under Ezra, Jews in Egypt at the time of the Septuagint translation could have determined a second, different canon, and that yet a third canon could have been formed at the time of Hillel and Shammai that would have canonized the Maccabees.⁴⁷ There is not ‘any probability or likelihood in it all’, Cosin asserts,

[W]hen all the world knowes, that the Jews (who have always been both religious and superstitious observers of their Fathers Traditions,) never yet admitted, never acknowledged, nor never heard of any such Second or Third Canon of Scripture among them, having most exactly kept themselves to The First, as it was consigned and delivered to them by the Prophets. Which is so fully attested not only by the Modern and Ancient Jews, but confirmed likewise by the Greek and Latin Fathers of the Church, as it is most an end so freely acknowledged by the writers in the Roman Church itself, that it would be too importune and superfluous a labour to recite here all their Depositions on this purpose. It will be enough to produce only the testimony of Josephus who lived in the Time of the Apostles, & wrote the Antiquities of the Jews (of whom he was one himself) in a most exact and diligent manner. His Testimony so great in this matter, that it is repeated by Eusebius & put in his Ecclesiastical History full at length.⁴⁸

Here Cosin gives the precise reference to Eusebius’s Ecclesiastical History, and then cites text of Josephus’s Against Apion I:37–42 in English and in Greek (partially) and Latin (entirely), adding: ‘Agreeable whereunto we have the Testimony also of Philo, who lived in the same Age with Josephus, “That the Jewes would rather have suffered a Thousand deaths, then that any thing should be once altered in all the divine laws and Statutes of their Nation.’’’

The testimonies of Philo and Josephus, two first-century Jews, have, by patristic osmosis (Cosin’s reference for Philo is Eusebius’s Praeparatio Evangelica 8), become part of ancient Christian literature such that one can mobilize Eusebius, as Cosin does here and as Whitaker had done before him, to argue for the care with which Jewish tradition protects the Law. But Cosin departs from the precedent set by Bellarmine and Whitaker, dispensing altogether with the patristic vessel through which Josephus has been transmitted. For Cosin, to prove that the canon consists of the books of the Hebrew Bible and no others ‘it will be enough to produce only the testimony of Josephus.’ Not only a necessary witness, Josephus has become a sufficient witness.

If Josephus’s list of canonical books could be mobilized in inter-confessional polemic between Protestants and Catholics, it was also put forward in early modern religious disputation between Christians and Jews. Scholars of seventeenth-century European erudition have long drawn attention to the unique set of circumstances that enabled the flourishing, in the Dutch Republic, of the Christian study of ancient, medieval and contemporary Hebrew and Jewish texts.⁴⁹ One of the forms that Dutch Calvinist engagement with Jewish learning and contemporary Jews took was one of the classic genres of Christian literature: the refutation of Judaism. Johannes Hoornbeek’s Teshuvat Yehuda (‘Judah’s Answer’) is a fist of a book, written in part against the most famous Jew of his age, the Amsterdam Rabbi Menasseh ben Israel. It was published in 1655 in Leiden, for which Hoornbeek had recently left Utrecht to take up a chair of theology.⁵⁰ Laying out a method for debating Jews, Hoornbeek asserts that Sacred Scripture, which is to say, the Sacred Scripture of the Old Testament, is the point of departure for all disputations:

Verily, what the Old Testament is concerned, and its authenticity as well as the books that constitute the canon, there is no discrepancy between us.⁵¹

Taking his cue from the controversialist literature discussed above, Hoornbeek begins his explication with Josephus’s description of the canon in Against Apion. But Hoornbeek does not set Josephus among patristic witnesses, but rather among medieval and contemporary Jewish ones. In Holland, the centre of the European printing and book-trade in his day, Hoornbeek had access to a wide range of relatively recent Jewish literature, including David Gans’s Zemah David (translated by Willem Vorstius), Solomon ibn Verga’s Shevet Yehuda (translated by Georg Gentius), Abraham Zacuto’s Sefer Yuhasin, Azariah de’ Rossi’ Meor Einayim and the works of Menasseh ben Israel himself. As he read through these works, Hoornbeek recorded evidence for Jewish awareness of the Apocrypha.⁵² Hoornbeek did not speculate, as Génébrard had, that there might have been additional instances of Jewish canonization in the pre-Christian or early-Christian Hellenistic eras. Instead, he looked for examples of Jewish awareness of the Apocrypha in the course of the Christian era. He believed that the Jewish conviction that the Christian Old Testament was very different from the Hebrew Bible presented one of the major stumbling blocks in the conversion of the Jews. If only they could be shown that the Jewish Scriptures and the Christian Old Testament contained the very same books, solid ground for a Religionsgespräch would be established.

If the Dutch Calvinist Hoornbeek departed from Protestant tradition by embedding Josephus among other Jewish sources, the French scholar and priest Richard Simon departed from Catholic tradition of the Jesuit controversialists.⁵³ Simon was too radical in his account of the history of Scripture to remain an orthodox Oratorian in Bossuet’s France, yet arguably too Catholic to be at home in the Dutch Republic where his works were freely published.⁵⁴ In his monumental study of the textual history of the Old Testament, Simon read Josephus’s account of the canon in a completely new and innovative way. His concern is not the search for a witness corroborating the canon and doctrine of his confession. Rather, Simon takes it as an important piece of evidence for the long history of Jewish scribal practice on both sides of Artaxerxes’s rule, that is, before and after the historical moment when Josephus claims prophecy to have ended. ‘Given that the power of public scribes … has always been the same throughout the period of the Jews’ Republic's subsistence’, Simon argues,

It must not surprise us that in the collection of canonical scriptures, there are those which were written after Ezra, and that Ezra was not the last compiler of Sacred Books. It is of little importance that these later scribes did not have the name of Prophets, given that they had the same authority. For it is certain that the Jews preserve after Ezra the acts of all that was of significance in their state, as we can see from the end of I Maccabees. Josephus, nonetheless, writing against Apion, testifies to the fact that the books of the Jews which were written after the reign of Artaxerxes do not have the same authority as those written prior to that time because there was then no longer a certain succession of prophets.⁵⁵

Like Hoornbeek, Simon was deeply conversant with post-biblical Jewish literature and was keen to point out those places where the rabbis mention books of the Greek Old Testament that did not survive in Hebrew. In the chapter immediately following, reviewing the different traditions of ordering biblical books, Simon notes the two traditions of counting them as twenty-two and as twenty-four, and the different interpretations of these numbers. Here Simon explains the numbering of twenty-two books in Against Apion as ‘according to the letters of the Hebrew alphabet.’⁵⁶ But Josephus himself said no such thing; it is reported by Eusebius, who ascribes it not to Josephus but to Origen (History VI.25) whence it is taken up by Jerome and others. Completely immersed in this tradition, Simon conflates it: Josephus’s mediation through Eusebius into patristic literature becomes total assimilation.

Johann David Michaelis, while entirely conversant with the literature discussed above, belonged self-consciously to a different, eighteenth-century world.⁵⁷ For the German Pietists who educated him the canonicity of a biblical book had ceased to be a burning question, relegated to a neat corner of the academy, Dogmatic Theology. After studying in Oxford with Robert Lowth, Michaelis became a key figure in what Jonathan Sheehan has called the forging of the Enlightenment Bible, shifting scholarly focus from confessional polemic to the ancient Near Eastern setting of the Hebrew Scriptures. ‘I shall here avoid entering into those disputes, which have been conducted with so much heat, and so much perplexity, with respect to determining the canon.’⁵⁸ Michaelis also sought to defuse, historically, the critical doctrinal question with which those canonical arguments had been bound up: the authority of the Church. ‘The Church of the eighteenth century can testify, that the sacred books at present in use are the same, which existed in the seventeenth century, this again with respect to the preceding and so on to the fourth century; further is the testimony of the church of no value.’⁵⁹ To Michaelis, the fierce disputations across the confessional divide about the authority of the Church were a hall of mirrors:

Whoever appeals to the evidence of the church to determine a book to be canonical, not to mention that it has condemned at one period, what it has approved at another, must first decide this difficult question, What is the Church, and who are heretics? If we answer, The true church is that which maintains the doctrines delivered in the sacred writings of the NT and if in answer to the question, how do you know that those writings are inspired, we reply, because the true church has determined them to be inspired, we manifestly argue in a circle.⁶⁰

And with that, Michaelis left two and a half centuries of controversy behind him. He anticipated the objection ‘But we appeal to the canon of the Jews with respect to the Old Testament: shall the Christian Church then have less authority than the Jewish synagogue?’; and he answers with something approaching historicism:

The difference is too visible to need explanation, and the bare testimony [dass blosse Zeugnis] of Josephus for the divine inspiration of the Old Testament is of more weight, than the decision of the Christian Church for the Divinity of a book of the New, even were all the sects in Christendom united to constitute that church. The writings of the Old Testament are confirmed not only by St Paul, but by Christ himself: on their authority therefore we rely, and not on that of the synagogue. But we have no Apostle to vouch for the canon adopted by the Christian Church, since the collection of canonical books was made after the death of the Apostles.⁶¹

In Michaelis, and in his English student and translator, Herbert Marsh (1757–1839), who carried his work into the nineteenth century, Josephus is not cloaked in patristic authority nor in confessional polemic, nor is he set among later Jewish sources on the Apocrypha.⁶² Perhaps in this case alone, Michaelis's historical criticism trumped his racialist hatred of post-biblical Jews, and ‘mere Josephus’ stands alone, the bare but reliable witness to the biblical canon in the first-century CE.