Introduction
The previous papers have concentrated on fragmentary witnesses to the canonical text — Hebrew witnesses to the Old Testament, Greek witnesses to the New, and the methods by which both are read and reported. The present paper turns to a different kind of evidence. Many of the fragments recovered from antiquity, including some of the most discussed in popular literature, do not contain canonical text. They contain non-canonical writings: works composed alongside the canonical books, sometimes in conscious imitation of them, sometimes in deliberate competition with them, sometimes in independent development of related themes. They also include fragments that have not yet been identified with any known work at all — pieces of writing whose place in the broader literary record remains an open question.
This material lies at the edges of the canonical question, and the edges are precisely the places where the temptations described in earlier papers run especially strong. The temptation to overclaim — to treat a non-canonical fragment as though it were a serious challenger to the canonical text, or to treat an unidentified fragment as though its meaning were already known — is real on one side. The temptation to underclaim — to treat the non-canonical material as though it had nothing to teach the careful reader, or to treat the unidentified material as though it could simply be dismissed — is real on the other. Honest fragmentology requires that both temptations be resisted.
The aim of this paper is to survey the principal categories of non-canonical fragmentary evidence, to set out the biblicist’s appropriate discrimination among them, and to commend the disciplines of humility appropriate to fragments that remain unidentified. The argument throughout assumes the doctrine of canon set out in earlier papers and seeks to apply it, with the hermeneutics of humility commended in Paper Three, to a particularly demanding portion of the evidentiary record.
I. The Non-Canonical Fragmentary Record
A. Second Temple Jewish writings
The body of Jewish literature produced between the close of the Old Testament writings and the apostolic age is substantial, and a considerable portion of it survives only in fragmentary form. The Dead Sea materials, surveyed in Paper Four for their biblical witnesses, also include a large body of non-biblical writings: sectarian rules of community life, hymnic and liturgical texts, commentaries on biblical books, apocalyptic visions, retellings and expansions of biblical narrative, calendrical documents, and others. Together with material preserved in other languages and in other forms — Greek translations, later Christian transmission of Jewish works, citations in early Christian writers — this corpus offers a view into the world in which the apostolic writings were produced.
For the present paper, several features of this corpus are worth naming. The first is its diversity. The literature of Second Temple Judaism was not a single body of writing with a uniform theological program; it represented multiple communities, multiple theological tendencies, and multiple genres. The second is its fragmentary character. Many of these writings are known only in part, sometimes only in small part, and the gaps in the surviving evidence are often substantial. The third is its relationship to the canonical books. Some Second Temple writings show extensive engagement with Old Testament Scripture — quotation, allusion, paraphrase, expansion. A few are engaged in turn by New Testament writers, who quote or allude to them in ways that have been the subject of careful study.
The believing reader need not be alarmed by the existence of this literature, nor by the fact that some of it intersects with the canonical record. The Holy Spirit, in inspiring the apostolic writers, did not isolate them from the literary world in which they wrote. That an apostle might allude to a non-canonical Jewish writing in the course of his teaching no more compromises the inspiration of the apostle’s writing than the apostle’s use of common Greek vocabulary does. The fragmentary record of Second Temple Judaism is part of the historical context of the New Testament and is, in that respect, useful background; it is not, on that account, a competitor for canonical authority.
B. Early Christian apocryphal works
A second category of non-canonical fragmentary evidence is the body of early Christian writings that present themselves, in various ways, as accounts of Christ, of the apostles, or of related matters, but that were not received by the church as canonical Scripture. This body includes works commonly known as the Gospel of Thomas, the Gospel of Peter, the Gospel of Mary, the Acts of various apostles in non-canonical forms, several apocalypses, and a number of other writings of various kinds. Some of these are preserved in substantially complete form; many are preserved in fragments, sometimes a single page, sometimes a few lines.
These works have received considerable attention in popular discussion in recent decades, often presented as suppressed or forgotten alternatives to the canonical Gospels. The careful worker is in a position to evaluate these claims more soberly than popular discussion has tended to do, and several observations are worth setting down.
The dating of these works, where it can be assessed, generally places them later than the canonical Gospels — in many cases substantially later. They do not, in the main, represent independent first-century traditions about Christ that the canonical Gospels failed to record. They represent, rather, second-century and later compositions whose relationship to the canonical writings is, in many cases, demonstrably one of dependence. Where they contain material not found in the canonical Gospels, that material frequently reflects theological tendencies — gnostic, encratite, and others — that were not received by the broader church.
The fragmentary state of much of this material has, on occasion, been used to suggest that the works in question were once more substantial and more widely read than the surviving evidence indicates. This is possible in principle but is not strongly supported by the evidence in most cases. The fragmentary character of the surviving copies is, in the main, what one would expect for writings that circulated in narrower communities, were copied less frequently, and were not preserved by the broad transmission that carried the canonical books across the centuries.
The biblicist’s response to this material is not to ignore it. Some of it is genuinely interesting as historical evidence of the diversity of religious thought in the early centuries; some of it has independent value for the study of the period. The biblicist’s response is, rather, to receive this material for what it is — non-canonical writing, generally later than the canonical writings, often theologically divergent from them, of historical interest but not of canonical authority.
C. Quotation, allusion, and intertextual overlap with canonical books
A particular kind of complication arises where non-canonical fragmentary writings quote, allude to, or otherwise overlap with the canonical books. The complication runs in both directions.
In one direction, non-canonical writings often quote or allude to canonical Scripture. A Second Temple commentary on a biblical book quotes the book it comments on; an apocryphal Gospel, in seeking to present itself as a record of Christ, reproduces sayings or narratives that overlap with the canonical Gospels; a sectarian writing builds its theological argument on verses from the prophets. Such quotations and allusions are, in themselves, fragmentary evidence for the wording of the canonical text in the period in which the non-canonical work was composed. They function, for textual purposes, somewhat like patristic citations, with the same limitations: a quotation may have been adjusted to fit the rhetorical purpose of the citing work, may have been drawn from memory rather than from a text, may have been subject to its own transmission history. The careful worker uses such evidence with appropriate caution.
In the other direction, the canonical books occasionally engage non-canonical material. A New Testament writer may quote, allude to, or echo a Jewish writing not received as Scripture, sometimes for the purpose of agreement and sometimes for the purpose of correction. The handling of such cases requires careful interpretation. It is one thing for an apostle to use a phrase or image from a non-canonical writing in the course of his own inspired teaching; it is another thing entirely to claim, on that basis, that the non-canonical writing must therefore be regarded as authoritative. The latter claim does not follow from the former. The apostle’s use of a source does not confer canonical status on the source any more than his use of a Greek poet’s line, in addressing a Greek audience, confers canonical status on Greek poetry.
The biblicist worker, encountering intertextual overlap, examines the cases on their own terms. He notes what is quoted, how it is quoted, and to what end. He weighs the evidence for the wording of canonical and non-canonical texts in the light of the citation. He resists the temptation, present in some quarters of contemporary scholarship, to treat intertextual overlap as evidence for canonical instability or for the equivalence of canonical and non-canonical sources. The overlap is real and is worth study; it does not establish what those who exaggerate it sometimes claim.
II. The Biblicist’s Discrimination
A. Canon as received, not constructed in modernity
A foundational point for any biblicist treatment of non-canonical material is the doctrine of canon presupposed in earlier papers. The canon, on this view, is not a construction of modern scholarship, nor a product of late ecclesiastical politics, nor a fluid category subject to revision in the light of newly discovered fragments. The canon is the body of writings that God breathed out and that the believing community received, under the providence of God, as Scripture.
This doctrine of canon does not require the believer to claim that there was no historical process by which the recognition of the canon took place. There was such a process, and it can be traced to some extent in the historical record. The doctrine claims, rather, that the process was a process of recognition rather than of construction — that the church, under the leading of the Spirit, came to acknowledge as Scripture the writings that God had given as Scripture, and that the eventual settled recognition reflects not the imposition of authority by ecclesiastical fiat but the convergence of the believing community’s reception around the writings that bore the marks of inspiration.
This doctrine has a direct bearing on how the biblicist receives the fragmentary record of non-canonical material. The discovery of a previously unknown apocryphal writing, or the publication of new fragments of a known one, does not pose a threat to the canon, because the canon is not a list assembled by modern scholarship that might have to be revised in the light of new evidence. The canon is the body of writings that the church has received as Scripture, on grounds that do not depend on the comprehensiveness of the modern scholar’s archival knowledge. New evidence may illuminate the history of the period; it does not reopen the question of which writings are Scripture.
B. Why the existence of non-canonical fragments does not destabilize the canon
The point just made can be restated more directly. The existence of non-canonical fragmentary evidence does not destabilize the canon, because the canon is not stabilized by the absence of such evidence. The canon is stabilized by the inspiration of the writings God has given and by the recognition of those writings by the community to which they were entrusted. Non-canonical writings existed in antiquity; they were not received as Scripture; their continued existence in fragmentary form, and their occasional re-emergence through new discoveries, is exactly what one would expect for writings that were known but not canonical.
This is worth stating directly because popular literature has, on occasion, presented the publication of non-canonical fragments as though it were a kind of revelation that overturns settled understandings of Christianity. It is not. The early church knew of these writings. The fathers wrote about them, sometimes at length, sometimes in works that themselves survive only in fragmentary form. The non-reception of these writings as Scripture was not a failure of the early church to consider them but a considered judgment, sustained across the believing community over time. The recovery of physical fragments of these writings in the modern period gives us better access to texts the early church knew of but did not receive; it does not give us access to texts the early church somehow missed.
The biblicist may therefore receive the non-canonical fragmentary record with calm. He may study it, learn from it, and benefit from the historical illumination it provides. He need not regard each new publication as an occasion for anxiety, nor as a challenge to the doctrine of canon. The doctrine of canon was not formulated in ignorance of the existence of such writings, and it does not require ignorance of them in order to be sustained.
C. How fragmentary apocryphal material can illuminate background without competing for authority
Non-canonical fragmentary material can, when handled with appropriate discrimination, illuminate the historical and literary background of the canonical writings in ways that aid careful study.
The vocabulary, idioms, and theological categories of Second Temple Jewish writings often shed light on the language of the New Testament. A New Testament word whose precise sense in apostolic usage is debated may be illumined by parallel uses in roughly contemporary Jewish writing, fragmentary or otherwise. The conceptual world of first-century Judaism — its expectations about the Messiah, its categories for thinking about Scripture, its forms of community life — is, in part, accessible through the surviving Second Temple literature, and this access can refine the modern reader’s understanding of the world in which the apostolic writings were produced.
Similarly, early Christian apocryphal writings, even where their theology is not received, are useful for understanding the ways in which biblical themes and vocabulary were taken up, developed, distorted, or contested in the centuries following the apostolic age. They provide evidence for the spread of Christian discourse in regions where direct historical records are otherwise limited. They illuminate the backgrounds against which orthodox writings of the same period were composed and the controversies in which they were involved.
The discipline here is to receive this illumination without confusing background for authority. The non-canonical material is useful as background; it is not, on that account, on a level with the canonical material. The biblicist does not preach from the Gospel of Thomas as he preaches from Matthew, nor does he treat a Second Temple commentary as he treats Isaiah. He may read the former with profit; he reads the latter as the Word of God. The distinction is not one of mere personal preference but of the doctrine of canon, applied with appropriate consistency.
III. Humility Regarding the Unidentified
A particular category of fragmentary evidence remains to be considered: fragments that have not yet been identified with any known work. This category requires its own discipline of humility.
A. Fragments not yet placed in a known work
Among the fragments recovered from antiquity, a portion remains unidentified at any given time. A small piece of papyrus may contain a few lines of Greek that do not correspond to any known biblical or non-biblical work. A scrap of parchment may preserve a few words of Hebrew or Aramaic that fit no recognized composition. These fragments are not necessarily insignificant; some of them may, on further study, prove to belong to known works whose other portions have not survived in the same hand, or to works lost altogether and now known only from such fragmentary remains.
The unidentified status of such a fragment is, in itself, a fact about the present state of knowledge rather than a fact about the fragment. A fragment unidentified today may be identified tomorrow, when a related portion is recovered or when a careful comparison reveals a connection that earlier study had missed. Honest reporting acknowledges that unidentified fragments are pending cases rather than closed ones.
B. The discipline of provisional identification
When a worker proposes an identification for a previously unidentified fragment, the proposal is properly provisional. It is offered as a hypothesis: that the fragment belongs to such-and-such a work, that it represents such-and-such a portion of that work, that it stands in such-and-such a relationship to other surviving witnesses. The hypothesis may be more or less strongly supported by the evidence. It remains, until adequately tested by other workers and by further evidence, a hypothesis.
The discipline of provisional identification has several practical features. The worker presents his proposal with the evidence on which it rests, in a form that allows other workers to evaluate it. He distinguishes between the features of the fragment that strongly suggest the identification and the features that are merely consistent with it. He notes alternative identifications that the evidence does not exclude. He revises his proposal when subsequent evidence requires.
This discipline is important because identifications, once published, tend to enter the literature and to be repeated. An identification proposed tentatively in one publication may be cited as established in another, and as the basis for further argument in a third. By the time a careful reader examines the chain, the original tentativeness may have been forgotten, and a hypothesis may have acquired the appearance of settled knowledge. The pattern is the same one named in Paper Six in connection with reconstructions, and the remedy is the same: clear marking of the status of claims, and willingness to revise as evidence requires.
C. The danger of forcing fragments into theological agendas
A particular danger besets the handling of unidentified fragments and, indeed, of fragments at the canonical edges generally: the danger of forcing fragments into theological agendas. The danger runs in every direction.
In one direction, a worker hostile to the historic faith may treat an unidentified or non-canonical fragment as though it constituted evidence against the canonical witness — as though its existence, or its content, called the canonical record into question. Such treatments often press the evidence beyond what it can bear. A small fragment, whose content is partial and whose context is unknown, becomes the basis for sweeping claims about early Christian diversity, about suppressed traditions, about the constructedness of the canon. The claims may attract popular attention; they rarely survive the careful examination of specialists who have actually done the work.
In the other direction, a worker friendly to the historic faith may be tempted to dismiss non-canonical and unidentified fragments out of hand, or to read them as straightforward confirmations of his prior commitments. This too is a forcing of the evidence, and it does its own damage. The unidentified fragment is what it is; it is not yet known what it is. Treating it as already known, in the service of any agenda, falsifies the evidence in the same way that overreading does.
The biblicist commitments of this series do not require the suppression of evidence; they require the honest examination of it. A fragment whose contents seem to challenge a comfortable understanding may, on closer examination, prove to fit comfortably enough; or it may prove to be evidence of something the worker had not previously considered, in which case the worker must learn from it. The doctrine of canon does not depend on the suppression of inconvenient evidence, because the doctrine is sound and the evidence, on careful examination, does not actually challenge it.
The right posture is the posture commended throughout this series: confidence in what God has spoken, modesty about what we have figured out, honesty about what we do not yet know. This posture allows the worker to handle unidentified fragments with patience, to examine non-canonical material with discrimination, and to refuse the various pressures, in any direction, to make the evidence say more or less than it actually says.
Conclusion
The fragmentary record at the edges of the canonical question — Second Temple Jewish writings, early Christian apocryphal works, fragments not yet identified with any known composition — is real, substantial, and worth careful study. It does not destabilize the canon, because the canon is not stabilized by the absence of such evidence but by the inspiration of the writings God has given and the believing community’s recognition of them. It can illuminate the historical and literary background of the canonical writings in ways that aid honest interpretation, provided the distinction between background and authority is consistently maintained. It calls for particular humility in the handling of unidentified material, where the discipline of provisional identification and the refusal to force fragments into theological agendas are especially important.
The biblicist need not approach this body of evidence with anxiety, nor with the dismissive confidence that pretends the evidence is not there. He approaches it as he approaches the rest of the fragmentary record: with confidence in the Word of God that has been given and preserved, with modesty about the state of his own knowledge, with the honest disciplines of triple distinction and calibrated reporting commended in the previous papers, and with willingness to receive correction from the believing community across time.
The next paper turns from this consideration of fragments at the edges to the practical chain by which fragmentary evidence reaches the believing reader: the path from fragment to critical apparatus to translation to pulpit and pew. The disciplines and discriminations developed in the present paper, like those of the earlier papers, will inform the pastoral and apologetic considerations that the next paper will address.
