JEWISH PSEUDEPIGRAPHA AND CHRISTIAN APOCRYPHA:
(HOW) CAN WE TELL THEM APART?
University of St. Andrews
An earlier version of this paper was posted here as an online lecture for my Old Testament Pseudepigrapha course on 22 February in the spring semester of 2002. The current, revised version is scheduled to be presented in the New Testament, Apocrypha, Pseudepigrapha and Dead Sea Scrolls Seminar at the British New Testament Conference in Cambridge, England, on 5-7 September 2002. This is still very much a work in progress!
The problem addressed in this paper is one central for dealing with the Old Testament Pseudepigrapha: the widespread assumption (especially frequently made by biblical scholars seeking New Testament "background") that Pseudepigraphic texts that lack explicitly Christian content or elements, or whose explicitly Christian elements can be easily excised, were originally Jewish compositions. In other words, if it doesn't look especially Christian, it must be Jewish. In two articles published in recent years, Robert Kraft has challenged this assumption. Most of these texts have been transmitted by Christians in Church languages and survive in rather late Christian manuscripts. Therefore, with texts in this category, our starting point should be these manuscripts. We should try to understand them initially as Christian works ("Christian apocrypha"), since this was their function in the forms in which they are actually preserved - they must have meant something to their Christian tradents, whatever the origins of these works. This is not a conclusion but a starting point: we need to understand the work in the historical, linguistic, and cultural context of the earliest surviving manuscipts and to work backwards from there only as required by needs of the particular case. When working backwards we shall conclude in some - perhaps many - cases, that a given work is a "Jewish pseudepigraphon" (i.e., that it was composed originally in Jewish circles, primarily for a Jewish audience). In other words Kraft wishes to reverse the burden of proof: rather than assuming that a given work is a Jewish composition until demonstrated otherwise, it behooves us to consider it a Christian composition unless demonstrated to be Jewish. Kraft's methodology has been used profitably in a number of recent works, for example Ross Kraemer's book on Joseph and Aseneth and David Satran's book on the Lives of the Prophets.
I take Kraft's position as a starting point in this paper and wish to explore its implications and develop it. Given that Jewish pseudepigrapha transmitted by Christian tradents must have meant something to these tradents; that it can be argued that no literary work is perfectly consistent in itself or perfectly understood by anyone, including its author; and that readers often can tolerate a great deal of cognitive dissonance when reading a text, especially one with a spiritual or canonical authority (witness the Christian exegesis of the Pentateuch); what sorts of things tell us that we need to work backwards from the manuscripts we have started from? More specifically, what positive evidence might indicate to us that the Christian apocryphon in our hands originated as a Jewish pseudepigraphon? What factors might lead us to consider a work Jewish, Christian, some mixture of the two, or something else? What is the range of possible categories of authorship from the second temple period through late antiquity? The purpose of this paper is to work out the range of possibilities, to propose some methodological principles for sorting them out, and to explore the limits of what we can know.
My approach will be somewhat roundabout, and will review some heavily trodden ground, but it aims to formulate criteria for making explict how we know what we know, so that we can also make clear what we do not know. Starting with the principle of positive verification, I will endeavour first to formulate external criteria for identifying Jewish compositions, then draw on the initial corpus and the scholarly literature on the "parting of the ways" to formulate internal criteria.
External Criteria and a Preliminary Corpus
I proceed from the assumption (which I do not have time to defend here, but which most would accept) that Judaism developed out of the religion and culture of Iron Age Israel and Persian period Judea, and that the bulk of the collection now known as the Hebrew Bible was composed and redacted into various forms, some surviving today, during this same period. If we want to know how to identify ancient Jewish literature, the logical first step is to identify verifiably ancient Jewish manuscripts, manuscripts that survive from the second temple and Hellenistic periods.
There are some very obvious positive criteria that can establish with virtual certainty that a given ancient work is Jewish: if a work with clear Jewish themes and content (i.e., pervasive knowledge of and interest in material and themes in the Hebrew Bible) survives in manuscripts copied in the pre-Christian era; if a work survives from the pre-Christian era or even the Hellenistic/Roman periods in Hebrew, the language of the Jewish scriptures; or if a copy of the work is preserved in a physical context that is undoubtably Jewish (the Qumran library is the obvious example). If we use these obvious criteria to establish a preliminary corpus of ancient Jewish literature, we come up with something like the following.
The Dead Sea Scrolls recovered from caves near the Wadi Qumran
The texts recovered from Masada
The texts of the Bar Kokhba era recovered from caves in the Judean Desert
Jewish inscriptions in Palestine and the Diaspora (a category that has its own problems and will not be considered here)
Texts preserved in highly fragmentary manuscripts among the Judean Desert manuscripts but fully in later primary or secondary translations. These include the Book of the Watchers; the Astronomical Book; the Animal Apocalypse, and the Epistle of Enoch (all of which survive in 1 Enoch), as well as Jubilees, Tobit, Ben Sira/Ecclesiasticus, and the Epistle of Jeremiah. None of these survive intact in ancient copies and only the last survives complete in its original language, but the spot checking allowed us by the ancient (and in the case of Ben Sira, medieval) fragments of the original confirms that they have been accurately transmitted in translation.
Tannaitic rabbinic material, especially in the Mishnah. Such works are preserved in late manuscripts but are written largely in Hebrew and transmitted exclusively in Jewish circles.
I have imposed a somewhat arbitrary cut-off point of the early third century C.E., which takes into account the fact that most allegedly Jewish pseudepigrapha are thought to have been composed by this time, but also keeps the preliminary corpus manageable in size.
Another external criterion is less conclusive, but worth taking into account. It is reasonably to take it at least as a working hypothesis that works by a named and dated Jewish author which are extensively quoted with attribution by numerous later Christian authors are in fact Jewish. I have in mind, of course, the large corpora of texts attributed to the first-century C.E. historian Josephus and the slightly earlier philosopher Philo of Alexandria. Ultimately, these corpora are accepted as Jewish on internal grounds (on which more anon) as much as external. But the overwhelming evidence for the existence of these works in antiquity in much the same form we have them today should give us reasonable confidence that they are mostly genuine works of the attributed authors. For any specific passage we must, of course, pay close attention to issues of textual criticism before accepting a given text as original
In addition, quotation as Jewish in a certainly Jewish text, or transmission of a work in second temple or postbiblical Jewish circles, supports its genuineness as an originally Jewish text. This criterion is potentially useful as collateral evidence but is not in itself decisive. Jewish writers sometimes quote with approval or make use of non-Jewish works. For example, the fifth-century Jewish Aramaic archive recovered from Elephantine (in Egypt) included a copy of the obviously (by its content) non-Jewish Story of Ahiqar, and Ahiqar himself is mentioned with approval in Tobit 1:21-22 and even claimed as Tobit's nephew. If we take into account quotations in Philo of Alexandria and Josephus, we could add a number of works to the list, such as the Letter of Aristeas and 1 Maccabees.
The Problem of "Common Judaism"
Taking this preliminary corpus as a starting point, can we then say that it provides evidence for a cohesive thought world that we are justified in thinking of as "common Judaism," a Judaism shared by the writers of all the texts, whatever their differences? It has been argued at length by a number of writers (such as E. P. Sanders and James Dunn) that such a common Judaism existed in the centuries on either side of the Common Era, although Jacob Neusner has disputed this point and has preferred to speak of a greater or lesser number of "Judaisms" instead. We may take Dunn's "four pillars of second temple Judaism" as representative of the case in favour of common Judaism. Dunn's pillars are:
Monotheism - there is only one God, who rules the world, and the many gods of the more tolerant polytheists of the ancient world are idols who may not be worshiped and who are unworthy of worship, either because they do not exist at all or they are angels subordinate to God or they are really aspects of God's character.
Election - God has chosen the people of Israel alone as his own covenant people to whom he has given the Promised Land.
Covenant and Torah (Sanders's "covenantal nomism") - The covenant between God and Israel is centrally focused in the Torah, the revelation and instruction of God to Israel. By the period that concerns us, Torah was embodied in the written Pentateuch of Moses, on whose text extensive oral and written commentary was already being composed. The Torah gave Israel distinctiveness, a sense of privilege, and focused laws and rituals that marked them off from others (chiefly circumcision, observation of the Sabbath, and food laws). I would add the laws of ritual purity to this list
Land focused in Temple - The Temple in Jerusalem stood at the centre of the national worship of God. It had political, economic, and religious functions, especially notably the priestly and sacrificial system which, I would add, made it the focal point for the concept of ritual purity.
If we compare Dunn's pillars to the evidence of our preliminary corpus of Jewish literature we find a close degree of correspondence, but also a number of difficulties. Most of the texts do consistently share precisely the concerns laid out by Dunn as common Judaism, but one corpus in particular does not. This leads me to the work of Gabriele Boccaccini, particularly in his book Beyond the Essene Hypothesis. Boccaccini has attempted to demonstrate that the Judaism of the second temple period included two major trends or streams, which he calls "Zadokite" and "Enochic" Judaism. According to Zadokite Judaism the world is a good place (as per Genesis 1) and evil arose from the sin of Adam and Eve in the Garden of Eden (as per Genesis 2-3). The hero of Zadokite Judaism is Moses, and the Mosaic Torah or divine instruction in the Pentateuch is centrally important. The Jerusalem temple is a microcosm of the universe, and its ritual cult and the Zadokite priesthood that maintains it are also centrally important. Enochian Judaism presents evil as arising when fallen angels corrupted the world and seduced human women, as in the myth of the giants, found in the Book of Giants, the Book of the Watchers, and elsewhere (cf. Gen 6:1-4). The hero of the Enochians is the pre-Flood patriarch Enoch (cf. Gen 5:18-24). The Enochians may have had priestly connections and they certainly show interest in priestly matters, but they are hostile to the Second Temple and to the Zadokite priesthood who administer it. They show little interest in the Mosaic Torah, focusing instead on the divine wisdom revealed to Enoch in his visions. Boccaccini argues that the Dead Sea Scrolls were collected and preserved by an extremist sect of Enochic Judaism (which had meanwhile accepted the importance of Mosaic Torah alongside Enoch's revelations), which, if true, means that our major library of certainly Jewish literature was collected by a rather untypical group with somewhat fringe views. Presumably it would contain both extremist sectarian literature (i.e., the Community Rule) and older foundational literature (i.e., biblical and Enochic books) but would leave out the rival Zadokite literature. Thus the sample of works in the Qumran library is skewed both in what it contains and what it omits. By and large, the elements of Dunn's common Judaism (apart from monotheism) apply more readily to Zadokite than to Enochic Judaism. Indeed, George Nickelsburg has argued persuasively and at length that Enochic Judaism (as represented in the Enochic works in our preliminary corpus) rejected the Mosaic Torah and was actively hostile toward the Jerusalem Temple in the period in question.
In short, we come to an impasse: a "monothetic" definition of common Judaism in antiquity does not seem to work. That is, no definition of Judaism based on a sine qua non or core essence can be formulated. But Jonathan Z. Smith, in his article, "Fences and Neighbors: Some Contours of Early Judaism," has pointed to a way forward. He proposes that classification of religions should follow a "polythetic" rather than a monothetic approach. Polythetic classification is an idea borrowed from biology. Rather than attempting to find an essence common to every member, it is based on a broad grouping of characteristics or properties. A member of the class being defined must have many of these characteristics, but no single characteristic is possessed by every member. Most members share characteristics with many other members, but some members have nothing in common with others. There will also be some borderline cases, which have a few of the characteristics but not enough to justify accepting these cases as members of the class. With a polythetic classification, the best we can do is to look for some general trends shared in antiquity by many or most Jews. Some important ones would include: worship of the God of Israel alone; acceptance of certain books as Jewish scriptures given as revelation by this God; the following of Jewish customs, laws, and rituals; participation in or support of the Temple cult in Jerusalem; self-identification with the Jewish nation; membership in and acceptance by a particular Jewish community; and acceptance of Palestine as the holy land. One can perhaps reasonably speak of a "common Judaism" and frequently shared elements, but there is no sine qua non. Some types of Judaism, notably the Enochic form, had very little overlap with common Judaism.
Early Christianity and the Parting of the Ways
The next question is, what is Christianity? The problem is similar but perhaps we can find something of an essence at least in antiquity: identification with Christianity or the Jesus movement involved belief in Jesus as a divine being with a redemptive function. The specifics of both elements were worked out in different ways. Jesus was the messiah or the incarnate god or an angel or a descending/ascending redeemer, etc., who functioned as eschatological redeemer, atoning sacrifice, or revealer of gnosis, etc. (Any comprehensive definition of modern Christianity would be still more diffuse: perhaps simply self-identification as a Christian along with the belief that Jesus was somehow important.) The Jesus movement originated in Palestinian Jewish circles, but within a generation it began to be coopted into a gentile movement. For our purposes here, the Apostle Paul is centrally important: he established a gentile version of the Jesus movement which featured Jesus as a redeemer through the atoning sacrifice of his death, and which lifted the requirement of following Jewish ritual law. Broadly speaking, the early Jesus movement thus developed two streams: a strictly Jewish one, the Jesus movement as a Jewish party or sect, and a gentile movement pioneered and spread by Paul and his followers which adopted the Jewish scriptures but taught the veneration of Jesus and the abandonment of Jewish ritual law, and for obvious reasons quickly lost interest in Jewish national identity. The Jewish Jesus movement mostly died out in the early centuries C.E., while the gentile movement is ancestral to all subsequent forms of Christianity (although of course with vast and rapidly increasing differentiation built around the basics summarized above).
I realize that I am making massive generalizations about a very complex set of historical questions, and some may want to challenge me on some of them, but in support of this general picture I can cite the article by L. V. Rutgers, "Jewish Literary Production in the Diaspora in Late Antiquity: the Western Evidence," in which he argues for the Jewish origin of an anonymous Latin treatise composed roughly in the late fourth century, the Collatio Legum Mosaicarum et Romanarum. To make his case he surveys Christian patristic literature through the fourth century for the views of the fathers on the Mosaic law. He finds a general agreement that the ten commandments, the "first law" that was given to Moses on his first ascent of Sinai, were still valid for Christians. However, the rest of the Pentateuchal laws, including the ritual laws so central to Judaism, were given during Moses' second ascent of Sinai after the golden calf episode (Exod 32). This "second law" was given to the Jews as a punishment to keep them in line after their idolatrous sin, and it was not to be followed by Christians. This widespread patristic view confirms the general picture I have outlined above. Christians accepted Jesus as a divine redeemer and adopted the Jewish scriptures, but rejected the Jewish ritual law.
In general, and this has been shown in numerous publications by many scholars in the last century, it can be shown that early Christianity, particularly in its majority and (retrospectively) orthodox forms, accepted the idea of monotheism (at least by its own lights), rejected the temple and ritual cult, rejected the Torah apart from the Ten Commandments, and allegorized the national identity of Israel and the priesthood to apply to itself.
The Range of Possibilities
It seems clear, then, that we can think at least in broad terms of some widespread theological differences between Judaism and Christianity in antiquity. I will come back to the question of what exactly they are, but first I want to explore the general implications of their existence for the question of how to tell Jewish pseudepigrapha from Christian apocrypha. Given the relationship between Christianity and Judaism outlined above, and working from first principles, what possible configurations of Judeo-Christian, explicitly Christian, and explicitly Jewish theological and cultural views might exist in a given pseudepigraphic text? The range of possibilities seems to me to be approximately the following. I am assuming that the works in question are Old Testament pseudepigrapha, not New Testament apocrypha (i.e., that they are meant to be read as works written in the Old Testament period, generally by Old Testament figures, rather than in the New Testament period by apostles, Jesus, etc.). I also assume in all cases that, whatever their origins, the works were transmitted in Christian circles.
1. A Jewish composition could be written (a) either containing what I will call Jewish "signature features," elements and views expressed that are characteristic of Judaism but not of Christianity, or (b) lacking such signature features.
2. A Jewish composition may have undergone minor Christian redactional changes during its transmission. These may include (a) additions with Christian signature features; (b) additions that do not include any Christian signature features; (c) inadvertent or deliberate deletions of Jewish signature features; and (d) other deletions. In other words, a Jewish work can be changed by Christian tradents to look either more Christian or less Jewish.
3. A Christian composition could be written containing numerous Christian signature features.
4. A Christian composition could be written containing only a few Christian signature features, such that it is indistinguishable from 2a.
5. A Christian composition could be written without any Christian signature features (for example, to given an air of verisimilitude to a work written in the name of an Old Testament character).
At least superficially, works in the categories 1b and 5 could be read as either Jewish or Christian works, whatever their actual origins.
6. A Jewish composition may be heavily reworked by Christian tradents, in which case it may be indistinguishable from 3 above.
7. Works may be composed by observant Jewish believers in Jesus ("Jewish Christians"). Such works could (but need not) contain both Jewish and Christian signature features and could be impossible to distinguish from 2 or 6 above (as well as 1 and 3-5!). It would be worthwhile to examine certainly Jewish Christian literature for signature features that could help distinguish 7 from the other categories (e.g., explicitly Jewish Christian polemics against other Jews).
8. Works may be composed by Judaizing Gentile Christians. Such Christians were the bane of John Chrysostom and are mentioned by other early Church fathers. These people were believers in Jesus but also had an active interest in Jewish culture and some chose to observe Jewish practices such as synagogue worship, the Sabbath or Jewish festivals, kosher laws, and even circumcision. We cannot rule out that such people might have written Old Testament pseudepigrapha that were sympathetic to Jewish matters rejected by the majority of Christians.
9. Works may be composed by non- or quasi-Jewish Israelites (Samaritans and Galileans). Signature features, if possible to isolate at all, would have to be worked out for these groups separately.
10. It is not impossible that works with biblical overtones could be composed by writers who are neither Jews nor Christians (e.g., proselytes and Judaizing "God-fearers," philosophers, or polytheists). Gentiles attracted to Judaism formed a continuum from pagans with some interest in Jewish traditions to proselytes who converted fully to Judaism. One could mention the Poimandres in the Hermetic corpus is an example of a polytheistic work influenced by Jewish tradition. It also has many polytheistic/indigenous religious elements but it is possible in principle that such elements might not be included, in which case works in this category could look like works from categories 1 or 5.
11. There are various highly unlikely scenarios that remain possible in principle: a composition by a Gentile (Christian or otherwise) who for some reason has learned Hebrew and writes it in that language; an apparently Christian work actually written by a Jew who for some reason intends to fool Christians with it and succeeds; etc.
If the analysis above is anything like correct, two main conclusions can be drawn from it. First, it will not always be possible to isolate originally Jewish works when they have been transmitted in gentile Christian circles. In some cases Christian compositions may not show any Christian signature features; in some cases Jewish works may lack Jewish signature features; redacted Jewish works may be indistinguishable from Christian compositions; and other, less likely scenarios may occur. But second, in some cases it will be possible to limit the options of authorship and origin and perhaps even to be confident that a composition is Jewish. When Jewish signature features are present we can reasonably postulate either Jewish authorship outright or at least significant Jewish (i.e., Judaizing) influence on the writer, however the text may have been transformed during its transmission. In some cases we may be able to postulate Jewish authorship and accurate transmission with some confidence. In other words, positive criteria may isolate texts more likely to be Jewish in origin but negative criteria (such as lack of Christian signature features) have much less, if any weight. In all cases, as is normal for ancient history, we must speak in terms of balances of probability rather than mathematical proof.
A Way Forward: Methodological Proposals
The goal then, is to formulate a set of criteria - signature features - which are positive signs of Jewish authorship. These can be worked out from the survey of the state of the question given above and verified against the preliminary corpus of Jewish texts also established above. The object is to apply these signature features to pseudepigraphic texts transmitted in Christian circles which are preserved in manuscripts from late antiquity or the Middle Ages, composed in or translated into languages used by Christian churches, using the criteria for positive assessment to isolate works that are likely to have a Jewish origin. I propose the following as a preliminary set of signature features. All of these are internal criteria, that is, they are based on the contents of the texts rather than facts external to them.
i. Jewish content and strong internal evidence that the work was composed in the pre-Christian era.
ii. Compelling evidence that the work was translated from Hebrew, which, as far as I can tell, was used only by Jews in antiquity. Compositions in Hebrew by Jewish Christians remain a possibility. Jews also wrote works in Aramaic and Greek, but Hebrew was the Jewish sacred language.
iii. Sympathetic concern with the Jewish ritual cult (especially priesthood, temple, ritual purity, calendar, festivals, sabbaths, circumcision).
iv. Sympathetic concern with Jewish law/Torah and halakhah (overlaps to a large degree with iii above).
v. Concern with Jewish national interests, particularly polemics against gentile polytheistic religions and internal Jewish polemics.
A warning: beware of Christian allegorical treatments of issues relevant to iii-v, which do not count as evidence of Jewish composition. Also distinguish internal Jewish polemics from Christian polemics against Judaism as well as Christian polemics against gentile polytheistic religions.
Using these criteria, it should be possible in many cases (although by no means in all) to isolate composition categories 1a (Jewish compositions with Jewish signature features) from 5 (Christian compositions with no Christian signature features); 2a (Jewish compositions with minor Christian additions including Christian signature features) from 3-5 (Christian compositions of various types); 2b (Jewish compostions with minor Christian additions, not including signature features) from 5, and 6 (Jewish compositions with extensive Christian reworking) and from 3 (blatantly Christian compositions). Category 9 (compositions by non- or quasi-Jewish Israelites) presumably would frequently be distinguishable from Jewish works, but not always. It is still unclear to me to what degree it is possible to isolate 7 (Jewish Christian compositions) and 8 (compositions by gentile Christian Judaizers) from categories 1-6, especially 2a and 6, although solid internal evidence that a work was composed before the rise of Christianity would be one conclusive feature. Compositions by especially dedicated pagan Judaizers (10) might also be difficult to distinguish from 1, 2b, and 5. Nevertheless, we can say at least that these last three categories (7, 8, and elements of 10) are directly in touch with real Jewish traditions.
Based on all the above, I propose the following principles as starting-point working hypotheses when we analyse pseudepigrapha.
A. Works with extensive or pervasive positive Jewish signature features and no Christian signature features provide the best case for belonging in category 1 (Jewish compositions) and it is reasonable to regard them as such. Even these could, in principle, belong in categories 7, 8, or 10, although a close reading of these works could in some cases eliminate these categories as possibilities, but the burden of proof would be upon those who wished to argue that the works belonged in these latter categories rather than category 1.
B. Texts with signature features of both Judaism and Christianity could in principle be regarded as either 2, 6, 7, or 8, and further analysis may or may not narrow down the possibilities.
C. All other pseudepigrapha, including those with neither Jewish nor Christian signature features, should be regarded (as an initial working hypothesis) as Christian compositions. They were, after all, treated as Christian works by the tradents, so this is the best place to start.
These three principles are imperfect for a number of reasons. Principle A may treat as Jewish pseudepigrapha Jewish works that have been altered by Christian tradents who nevertheless did not add Christian signature features. It may also categorize works by Jewish Christians or very dedicated Christian Judaizers as Jewish, assuming (as we cannot rule out) that these groups wrote pseudepigrapha and did not always include Christian signature features in them. Principle B leaves us with a frustratingly large range of possibilities. And principle C may lead us to treat real Jewish works that lack Jewish signature features as Christian compostions. But nevertheless, the use of these principles would force us to apply a methodological standard that is considerably more rigorous than the intuitive one generally applied at present in the study of these works.
Two more criteria for separating Jewish from Christian compositions are worth mentioning, although both are more subjective than the first five.
vi. Comprehensive eschatological scenarios, especially those explicitly involving eschatological redeemers or divine mediators, are unlikely to be Christian compositions if Jesus does not figure in them.
vii. A general but highly subjective criterion, to be tried especially if none of the other criteria apply. Which results in a greater exegetical payoff: reading the text as a Jewish composition or as a Christian composition?
The following are some limitations to the analysis:
The sources to be analyzed need to be well preserved. Highly fragmentary works are less likely to preserve the positive features needed for even a preliminary analysis.
Genre should be taken into account. For example, Jewish wisdom literature is less likely to contain distinctively Jewish material than halakhic literature, and Christian apocalypses with eschatological interests, are more likely to include direct references to Jesus than Christian rewritings of Old Testament stories.
Texts set in the pre-Mosaic period may deliberately avoid reference to Torah related issues, even if by Jews.
Texts set in the pre-Abrahamic period may deliberately avoid reference to Torah-related and nationalist issues, even if by Jews.
Analysis of translated texts (i.e., most of the Old Testament Pseudepigrapha) is highly dependent on the reliability of the translation.
Gentile Christianity of antiquity encompassed a wide geographical range and a vast span of time and came in many flavours, types, and movements. Christian apocrypha could have originated in many different social circumstances and works in the Old Testament Pseudepigrapha need to be studied on case-by-case basis with this in mind.
A Preliminary Application of the Method
A broad-brush application of the method proposed here, one that would have to be refined greatly for any given text, yields the following results for texts in the Charlesworth corpus which have not yet been evaluated in this paper. I limit the treatment to substantial works that are fully or mostly preserved. I consider only works probably composed or translated into Greek (and sometimes translated in turn into other languages) but not accepted into the major canons. I do not discuss works preserved exclusively in Armenian and Slavonic, since I am not trained in the relevant languages or cultures, and works from the Nag Hammadi corpus, since they deserve a separate treatment of their own.
1. Works containing Christian signature features and no Jewish signature features: Sibylline Oracles 6, 7, 8; Greek Apocalypse of Ezra; Vision of Ezra; Questions of Ezra; Apocalypse of Sedrach; Apocalypse of Elijah; Apocalypse of Daniel; Testament of Abraham; Testament of Jacob; Testament of Solomon; Testament of Adam; Martyrdom and Ascension of Isaiah. I consider all of these to be Christian compositions with the possible exception of the Testament of Abraham.
2. Works with pervasive and extensive Jewish signature features and no Christian signature features: 4 Ezra; 2 Baruch; Testament of Moses; Pseudo-Philo; 3 Maccabees; 4 Maccabees; Psalms of Solomon. I consider all of these to be Jewish compositions transmitted with minimal Christian interference.
3. Works containing both Jewish and Christian signature features: Sibylline Oracles 3-5 (possibly two very small Christian interpolations?); 4 Baruch; Testaments of the Twelve Patriarchs; Hellenistic Synagogal Prayers. In principle, these could be Jewish pseudepigrapha with Christian interpolations, Christian apocrypha making extensive use of Jewish sources, or works composed by Jewish or Judaizing Christians.
4. Works containing no Christian signature features and possibly a few Jewish signature features: Sibylline Oracles 11-12, Pseudo-Phocylides. These could be Jewish works, Christian works incorporating Jewish material, or works by Jewish Christians or Judaizing pagans or Christians.
5. Works containing Christian signature features and possibly also a few Jewish ones: Sibylline Oracles 1-2, 14; Lives of the Prophets; 3 Baruch; Testament of Isaac. These could be Christian works or Christian works incorporating Jewish material or Christian adaptations of Jewish works or works by Jewish Christians or Judaizing Christians.
6. Works containing neither Christian nor Jewish signature features: Similitudes of Enoch; Apocalypse of Zephaniah; Sibylline Oracles 13; Testament of Job; Joseph and Aseneth; Life of Adam and Eve; History of the Rechabites (Syriac version); Prayer of Manasseh; Sentences of the Syriac Menander. These could be Christian works (and this is the default position) but a wide range of other possibilities remain, to be argued, again, on the basis of positive evidence. For example, Syriac Menander could well be a pagan work adopted by Christians.
1. The term "Old Testament Pseudepigrapha" deserves a long discussion of its own, one not possible here. Let the following serve as a preliminary definition: the Old Testament Pseudepigrapha consists of the ancient literary remains of the anonymous/pseudonymous revelatory stream of tradition that originated in response to (or sometimes alongside of) the Hebrew Bible/Old Testament accepted in the major Jewish and Christian canons. These literary remains are frequently attributed to characters from the Old Testament and deal with themes in the Old Testament and legends about second temple Judaism. Any attempt to suggest a chronological cutoff point for including works in the corpus is to some degree arbitrary. Most works traditionally included are alleged to have been composed within a few centuries on either side of the Common Era. But in any case, they were adopted, preserved, developed, and augmented, mostly by early Christianity, through late antiquity and beyond. For the sake of argument in this paper, I will treat the term Old Testament Pseudepigrapha as synonymous with the corpus collected by James H. Charlesworth in The Old Testament Pseudepigrapha (2 vols.; New York: Doubleday, 1983 and 1986), although more texts could be added and some could be deleted.
2. Robert A, Kraft, "The Pseudepigrapha in Christianity," in Tracing the Threads: Studies in the Vitality of Jewish Pseudepigrapha (ed. John C. Reeves; SBLEJL 6; Atlanta: Scholars Press, 1994), 55-86; idem, "The Pseudepigrapha and Christianity Revisited: Setting the Stage and Framing Some Central Questions," JSJ 32 (2001): 371-395.
3. Éric Junod, "Apocryphes du NT ou apocryphes Chrétiens anciens?," ETR 58 (1983): 409-21; idem, "'Apocryphes du Nouveau Testament': une appellation erronée et une collection artificielle," Apocrypha 3 (1992): 17-46.
4. Ross Shepard Kraemer, When Aseneth Met Joseph: A Late Antique Tale of the Biblical Patriarch and His Egyptian Wife, Reconsidered (Oxford: Oxford University Press, 1998); David Satran, Biblical Prophets in Byzantine Palestine: Reassessing the "Lives of the Prophets" (Leiden: Brill, 1995).
5. E. P. Sanders, Paul and Palestinian Judaism: A Comparison of Patterns of Religion (London: S.C.M., 1977); idem, Judaism: Practice and Belief, 63BCE-66CE (London: S.C.M., 1992); James D. G. Dunn, The Partings of the Ways between Christianity and Judaism and their Significance for the Character of Christianity (London: S.C.M., 1991).
9. George W. E. Nickelsburg, "Enochic Wisdom: An Alternative to the Mosaic Torah?" In Hesed Ve-Emet. Studies in Honor of Ernest S. Frerichs (ed. Jodi Magness and Seymour Gittin; Atlanta, Ga.: Scholars Press, 1998), 123-32.
11. L. V. Rutgers, "Jewish Literary Production in the Diaspora in Late Antiquity: the Western Evidence," in The Hidden Heritage of Diaspora Judaism (2nd ed.; Contributions to Biblical Exegesis and Theology, 20; Leuven: Peeters, 1998), 235-84.
12. The classic work on the subject is Marcel Simon's Verus Israel: A Study of the Relations between Christians and Jews in the Roman Empire (135-425) (Oxford: Oxford University Press, 1986. Translation of French original by H. McKeating (Paris: Editions E. de Boccard, 1964). See esp. pp. 85-91, 163-73.
14. Corresponding Christian signature features are somewhat more straightforward and would include such things as mention of Jesus, the virgin birth, the crucifixion, the church, the apostles, and the Trinity; hostile references to the falling away of the Jews; and quotation from or clear allusion to the New Testament. See also Kraft, "Setting the Stage and Framing Some Central Questions," 386.
15. Texts with signature features of both Judaism and indigenous polytheistic (i.e., "pagan") religions would presumably fit somewhere in category 10. Offhand, I know of no such cases. Analogous to principle A, works with signature features of indigenous polytheistic religions which have been transmitted in Jewish circles (e.g., Ahiqar) or Christian circles (e.g., Syriac Menander), if they contain no Jewish or Christian signature features, may reasonably be regarded as indigenous polytheistic works as a first hypothesis.