Abstract
The similarities between 1 Peter and certain New Testament epistles (e.g., Romans, Ephesians) have long been recognized. According to the modern consensus, these affinities developed out of shared oral traditions which circulated within early Christianity and not through any direct literary borrowing. What has been overlooked in this discussion, however, is that the methods used to assess the linguistic and thematic parallels, by their very nature, have (virtually) excluded literary dependence as a potential explanation, and thus the denial of a literary relationship has been (essentially) guaranteed from the start. While not seeking to resolve the question of 1 Peter’s dependence on the Pauline corpus, this article attempts to expose some of the methodological problems surrounding the consensus view and to establish reliable criteria which can be used to determine literary relationships. The conclusions – while focused on 1 Peter – are applicable to the discussion of intertextuality more broadly.
Keywords
1. Introduction
It has long been recognized that 1 Peter is an ‘epistle of tradition’ (Spicq 1966: 15). The Petrine author weaves together a variety of source materials to create a patchwork of citations and allusions upon which he builds his paraenesis. One source from which he draws extensively is the Jewish scriptures (primarily from the text of the LXX). This is clear enough from the fact that, on occasion, the letter employs quotation formulas to indicate the presence of scriptural citations (cf. 1 Pet. 1.16, 24-25; 2.6-8, 25; 3.10-12). Even when such formulas are absent, numerous allusions are posited. By some estimates, the letter contains as many as 30 or 40 references to specific passages from the Old Testament. 1 Beyond this, many consider 1 Peter to be filled with intertextual echoes (or ‘biblicisms’), which mark a connection to the language and themes of the scriptural text. When the count is extended to include these references, then ‘scarcely a verse in this epistle would be exempt’ (Carson 2007: 1015).
Scholars are much more cautious when it comes to drawing literary connections with early Christian materials. Although it is acknowledged that 1 Peter contains numerous linguistic and structural affinities with other New Testament texts, most interpreters stress that the evidence is insufficient to posit a literary relationship. This was not always the case, however. Throughout the nineteenth and early twentieth centuries, the parallels with certain Pauline letters (especially Romans and Ephesians) led most scholars to conclude that 1 Peter was not only aware of the Pauline corpus, but also borrowed from it. 2 For this reason, much of early critical research on 1 Peter approached the letter as ‘an imitation of Paul’ (Semler 1783: praef. Abs. 3).
It was during the mid-twentieth century that the sources underlying 1 Peter began to be reassessed. While theories of literary dependence were still popular, some began to give serious consideration to the possibility that oral source materials might underlie these texts, a view which received strong impetus from the earlier introduction of form and tradition criticism. Many of the insights from these new methods were brought to Petrine studies through the monumental commentary of Selwyn (1947). Building on the pioneering efforts of Seeberg (1903: 86-96) and Carrington (1940: 23-29), Selwyn devoted over 100 pages to analyzing the similarities between 1 Peter and other early Christian writings (1947: 363-466). The conclusions that he reached were paradigm-altering: the numerous affinities represented a body of shared tradition which circulated within the early church and from which multiple authors drew.
Selwyn’s tradition-historical analysis laid the groundwork for future discussions. In the years that followed, many would extend his efforts even further (e.g., Shimada 1966). Within a few decades, some already felt that the discussion had reached the point where it was necessary to liberate 1 Peter from its ‘Pauline bondage’ (Elliott 1976: 248). Subsequent studies would do just that, insisting that the Pauline connections were not as strong as previously suggested and stressing the distinctive character of 1 Peter (Shimada 1998a; 1998b; Herzer 1998). These efforts led to the formation of the modern consensus: ‘Rather than literary dependence, the trend now is to view the affinities between 1 Peter and Pauline literature, as well as parallels with other New Testament literature, as arising from the use of common tradition in early Christianity’ (Dubis 2006: 209; cf. Webb 2004: 380).
Over the last decades, a few formidable challenges have been leveled against this position. 3 Nevertheless, most remain firmly committed to the standard tradition-historical construal. One of the reasons why the consensus has become so entrenched within the fabric of Petrine scholarship is because of what appears to be its strong methodological foundation. This view has arisen from a close analysis of the letter’s shared affinities, and it is informed by criteria designed to produce a sound conclusion. A closer look, however, reveals a number of reasons for concern. Most notably, one could point to an evaluative bias which has heavily influenced intertextual decisions. The purpose of this article is to expose some of the problems surrounding the consensus view and to propose a new methodological direction for the treatment of source materials in 1 Peter.
2. Exposing the Problems with the Modern Consensus
Few today would posit a literary relationship between 1 Peter and certain (deutero-) Pauline writings (e.g., Romans, Ephesians). As discussed above, this view fell out of favor many years ago, and it continues to be dismissed within the scholarly discussion. What has gone unnoticed, however, are the methodological shortcomings which mark this approach. The methods used to assess the linguistic and thematic parallels, by their very nature, have (virtually) excluded literary dependence as a potential explanation, and thus the denial of a literary connection has been (essentially) guaranteed from the start. In what follows, we will explore three factors which have contributed to this situation.
2.1. The Problem of Terminology
When it comes to the topic of intertextuality in 1 Peter, the problems begin at a very basic level: terminology. Literary dependence is often so narrowly defined that the possibility of its occurrence is virtually eliminated by its very definition. In some cases, interpreters suppose that Paul’s letters may have exerted an indirect influence on 1 Peter, which they unyieldingly distinguish from literary dependence. A case in point is the treatment by Elliott. Although one of the most vocal opponents of a literary connection with the writings of Paul, Elliott readily admits that ‘there are discernible similarities between 1 Peter and the Pauline letters’, noting not only the epistolary form and structure, but also ‘a significant amount of vocabulary, ideas, themes, and even formulas’ (2000: 38). Three of these parallels, in fact, are noteworthy enough to be singled out because of their exclusivity to 1 Peter and the Pauline epistles: the expression ἐν Χριστῷ, the reference to χαρίσματα and the combination of specific passages from the Jewish scriptures (viz. Isa. 8 and 28 with Hos. 2) (2000: 38).
Given these important similarities, it is instructive to note that Elliott also concedes that the Petrine author was exposed, in some way, to Pauline ideas. He states, ‘It is possible, if not probable that the Petrine author was familiar with one or more of Paul’s letters (esp. Romans)’ (2000: 37). The way this statement is phrased, it would seem that the Petrine author’s familiarity with the Pauline epistles is thought to have arisen through direct contact (i.e., reading/hearing the epistle). But elsewhere Elliott makes it clear that he is proposing a somewhat less direct exposure. When describing a possible situation in which the Petrine author may have accessed the material, he notes, ‘By the time 1 Peter was written, Paul’s letter to the Romans belonged to the body of teaching and traditional exhortation collected at Rome. The author of 1 Peter drew freely from this material, as did subsequent Christians writings from Rome’ (2000: 38). 4 So, according to Elliott, the familiarity of the Petrine author is with a larger body of accumulated tradition to which Paul’s epistle to the Romans contributed, but not necessarily with the epistle itself.
The position of Elliott is understandable as a response to the over-emphasis of Pauline dependence which marked the earlier discussion; nevertheless, it is ultimately an unsatisfactory explanation. The trouble lies in the postulation of a nebulous concept like ‘Pauline tradition’, which is thought to have circulated within the Roman community disconnected from any written documents. 5 It is true that oral traditions did play a part in the composition of 1 Peter. The author seems to make use of formulary materials on occasion (cf. 1 Pet. 2.21-25; 3.18-22; see Williams forthcoming). But, in this case, it is much simpler to acknowledge the literary links. 6 When two written documents share unique verbal and conceptual similarities, then it is natural to posit a literary relationship – especially if the potential borrower (i.e., Petrine author) and the proposed source (i.e., Romans) can be placed in geographic and temporal proximity. 7 One would be hard-pressed to argue that the scriptural allusions in 1 Peter were derived from a tradition which was shared between the Petrine author and the scriptural writers. Because a written text is available for comparison, it is naturally thought to be the source. Why is the same approach not taken when affinities are noted with the Pauline writings?
Closely related to this, another problem which illustrates the importance of properly defining literary dependence is the failure to acknowledge the citation of an author’s work from memory as an instance of literary borrowing. The discussion by Schreiner is illustrative of this approach. In his commentary, Schreiner argues that ‘the Petrine letter does not show evidence that he used Paul’s letters as a literary source’. What is illuminating is the reason he gives for this deduction: the author’s ‘wording is not exact enough to suggest that he wrote with any of the Pauline letters before him’ (2003: 34). Despite this claim, he concedes that the Petrine author had read some of Paul’s epistles (2003: 34 n. 52). 8 What prevents Schreiner from attributing the close similarities between 1 Peter and certain Pauline letters to literary dependence is his definition of the phenomenon. According to Schreiner, literary borrowing only occurs when the author of a hyper-text composes his work while physically examining his source-text.
But, again, a number of questions about this position could be raised. Are we to suppose that the Petrine author had a physical copy of each source-text in front of him as he wrote? Based on the expense of written copies of the Hebrew and Greek scriptures, few individuals could afford their own personal copy. 9 It is possible that the author had direct access to copies through a wealthy patron or local synagogue. 10 But even then, it would have required a large number of scrolls. By some estimates, 1 Peter cites from at least eight different scriptural books. It is highly unlikely that the author had each of these at his disposal while he wrote. More probably, the scriptural citations are drawn from the author’s memory on the basis of what he had previously heard or read. If this were the case, then the extensive list of scriptural citations/allusions which most scholars identify in 1 Peter would not qualify as literary borrowing – at least not according to the narrow definition of Schreiner.
The treatments of Elliott and Schreiner reveal the importance of precisely and accurately defining literary dependence. For our purposes, this phenomenon is best defined simply as an author’s re-use (i.e., hypertext/receptor-text) of an earlier written text (i.e., hypotext/source-text). 11 Although this definition leaves much to be explained, it would include the citation of a text from memory. 12 Further, contrary to the assertion of Shimada, 13 the hypertext need not reproduce the hypotext exactly in order for literary dependence to occur. While points of detailed correspondence aid in the process of identification, the ancient world provides plenty of examples in which an author borrowed from and freely modified a given source. 14
2.2. The Problem of Categorical Distinctions
A second issue which has impeded the discussion of intertextuality and has prevented scholars from considering the literary connections with other early Christian texts is the categorical distinction made between Old Testament citations/allusions and other source materials. Given the frequency with which 1 Peter makes use of the Old Testament, the popularity of this topic within Petrine studies is not surprising. In studying this phenomenon, scholars have created complex classification schemes which draw fine distinctions between types of intertextual references. Most notable are the treatments by Schutter (1989: 35-43) and Woan (2004: 213-15), who not only distinguish between types of citations, but also between levels of allusions (and beyond). 15
These taxonomic classifications serve as a useful heuristic tool for understanding textual references in 1 Peter. They are an especially valuable means of disclosing how the Petrine author used his source material. Nevertheless, there may be some (unintended) negative repercussions. The problem is that classification schemes like these tend to treat Old Testament references in isolation from other source materials, which only fuels the view that other literary connections in 1 Peter are either secondary or altogether absent. Furthermore, the narrow focus of these categories means that they are not (easily) transferable to different forms of source material. 16 If we are seeking a more holistic understanding of the author’s use of source materials, then we must broaden our classification scheme.
Although it would require limiting the number of categories, a more productive approach might be to utilize labels that could extend across different source materials. In this vein, I would suggest that references to written sources in 1 Peter be confined to two basic categories: (a) citation/quotation, which is defined as the re-use of one or more word(s) or concept(s) from a source-text as indicated by a citation formula (or some equivalent); and (b) allusion, which involves the re-use of one or more word(s) or concept(s) from a source-text without any explicit indication. 17 By their very nature, the only distinguishing characteristic between these categories is the (lack of a) formulary introduction. The benefit of using such structural definitions for all intertextual references is that the similarities between 1 Peter and early Christian writings would be treated on equal grounds with the literary connections found in Old Testament texts (rather than implying that the former represent a secondary order).
2.3. The Problem of Criteria
Most who treat the topic of intertextuality in 1 Peter conclude that the similarities between the epistle and other New Testament writings are insufficient to posit literary borrowing. In doing so, they regularly neglect to specify (a) the level of correspondence necessary to postulate a literary relationship and/or (b) the method(s) which might be used to assess the parallels. What is lacking, in other words, are criteria by which to judge the existence and direction of literary dependence. 18
But even when a clearly articulated method is employed, it is often insufficient to combat a problem which marks various discussions, viz. the inconsistent treatment of different source materials. In recent discussions of the similarities with the Pauline corpus, the literary influence of Paul on 1 Peter is treated differently from the scriptural sources. Not only are the Pauline parallels approached with much greater skepticism, the basis for judging literary dependence becomes much more restrictive. This can be seen most readily in the studies of Shimada, who is the only Petrine interpreter (to my knowledge) who has delineated a set of criteria for making source-critical judgments.
Drawing attention to the fact that most who advocate 1 Peter’s dependence on the Pauline epistles fail to explain (methodologically) how their conclusion was reached, Shimada proposes the use of four guidelines to confirm literary dependence (1998b: 105):
A passage should be quoted explicitly and extensively (and the author and writings, from which he allegedly quotes, should be identified, if possible).
From a context-analytical point of view, both the original and the reproduced passages or phrases should be exactly the same, or at least very similar.
The phrases (if possible, with the same word order) or words should be identical, or be replaced with paronyms of similar meaning.
The concepts represented should be the same or very close.
According to Shimada, literary borrowing should only be proposed when one or more of these criteria have been satisfied.
The importance of these methodological principles within Petrine scholarship derives not so much from their use by later interpreters, because few subsequent scholars have undertaken detailed examinations of the parallels using the same criteria (although, cf. Herzer 1998: 13). Instead, their significance lies in their perception among scholars. Given that Shimada was able to challenge 1 Peter’s dependence on the Pauline epistles based on a regulated method, the consensus – of which Shimada’s views are representative – is perceived to be situated on a firm methodological foundation. This was the conclusion of Elliott, for instance, who notes ‘Shimada’s extensive and meticulous analysis of the pertinent passages’ and applauds ‘his clarification of the valid criteria for comparative analysis’ (Elliott 2000: 22; cf. Achtemeier 1996: 16 n. 148; Schreiner 2003: 23 n. 6).
The major problem is that Shimada’s criteria are so rigidly constructed that literary dependence could rarely – if ever – be demonstrated, and thus the denial of a literary connection with the Pauline letters is (almost) inevitable. 19 The austere nature of these criteria is most clearly displayed in the first point. To claim, as Shimada does, that an author must provide an explicit acknowledgment of his (or her) source is to set out an almost insurmountable obstacle for identifying literary borrowing. 20 By this standard, one would have to conclude that there are no citations from the Jewish scriptures in 1 Peter, for the author only occasionally mentions that he is citing scripture and never mentions the specific author or work from which he is quoting. 21 An explicit reference is only one of many ways that an author might reflect his/her source-text, and it does not take into account the tendency of the Petrine author to employ source materials without any distinguishing notation.
Further, the remaining criteria of Shimada are equally flawed. Requiring that the source-text and citation be ‘identical’ or the ‘same’, as points 2-4 state, is too strenuous of a demand. This is not representative of how sources were cited in either the Jewish or Greco-Roman contexts in which 1 Peter was composed, nor is it consistent with the use of source materials by other early Christian authors. 22 According to this standard, 2 Peter and Jude – two texts between which scholarship widely acknowledges a literary relationship – would not qualify as a case of literary borrowing. But even more pertinent is the fact that 1 Peter itself reveals a variegated approach to the use of source material, as indicated by the letter’s citation of (and allusion to) scripture (see Jobes 2006; Müller 2007; Ådna 2011). Sometimes a large quantity of text is quoted with considerable correspondence (e.g., 1 Pet. 4.18, quoting Prov. 11.31 LXX); on other occasions, the source-text appears in a modified form with varying degrees of similarities (e.g., 1 Pet. 2.6, quoting Isa. 28.16; 1 Pet 2.8, alluding to Isa. 8.14). If the Petrine author employed such variety when referencing the scriptural text, why would we expect his borrowing from the Pauline epistles to be any different?
By way of summary, then, a close analysis of the situation reveals that the modern consensus was built upon narrowly defined terminology, unnecessary categorical distinctions and a rigid set of criteria, all of which effectively led interpreters to rule out literary dependence from the start. This consideration creates a serious problem for the popular view on intertextuality in 1 Peter. Despite this fact, however, Petrine interpreters continue to espouse the theory without any reservations. Surely, there is a need to reassess the situation afresh.
3. A New Approach toward Intertextuality in 1 Peter
One of the primary aims of this article is to expose the methodological bias which has prevented recent Petrine scholarship from entertaining the possibility that a literary connection might lie behind the similarities between 1 Peter and other early Christian writings. The objections raised in the previous section were meant to contribute toward this goal. To the extent that my efforts are successful, scholars would be enabled to take up the topic afresh, unencumbered by the methodological shortcomings which have long obstructed the discussion. In what follows, I would like to suggest how this might be done.
The key, I will argue, is that the question must be approached through an entirely different methodological framework. This will require a new set of criteria which can demonstrate, with a reasonable degree of probability, whether literary borrowing has occurred. Intersecting this need, there has been a flurry of recent attempts in both Hebrew Bible/Old Testament 23 and New Testament studies 24 to identify this intertextual phenomenon. Since these discussions have settled on many of the same criteria, they can provide a blueprint from which we can work.
Echoing many of the criteria that have been discussed in previous studies, I propose that there are five general guidelines – each serving different purposes – which could be used to gauge the strength of the evidence related to literary dependence in 1 Peter. They include:
Explicit Reference. The most certain indicator of literary dependence occurs when an author makes a direct reference to a source, either by explicitly naming the source or by using a quotation formula to introduce a citation.
External Plausibility. The relationships of time and space must allow for the possibility that an author borrowed from a particular source. That is, ‘[d]ependence can be invoked only if external factors make such dependence plausible’. 25 In order to propose a literary connection between two documents, the hypotext must have been composed prior to the hypertext, and it must have been accessible (either in written or audible form) to the author of the hypertext.
Authorial Tendency. The chances of literary borrowing increase when a given author has demonstrated a tendency to use source materials in other written works, and/or in the text in question.
Recurrence. If a source was widely known and commonly quoted by other authors during a given time period, this can increase the chances that it was cited by the author in question. 26
Verbal Agreement. 27 If the two documents share a certain level, number and type of verbal affinities, then it is plausible to posit a literary connection between them:
Level of Similarities. Verbatim agreement across multiple words or phrases provides the clearest indication of literary borrowing; however, a difference in the form and order of words need not rule out a literary connection between two texts. In some cases, a receptor-text might depart from its source-text in significant ways (cf. 2 Peter’s use of Jude). Therefore, the similarities between two texts should be the determining factor for evaluating literary dependence, not their differences. 28
Number of Similarities. As the number of similarities are multiplied, the likelihood of a literary connection also increases. Nevertheless, the numbers in and of themselves can be misleading: ‘parallels between two texts may be numerous but trivial … On the other hand, as few as two or three weighty similarities may suffice’ (MacDonald 2000: 8).
Type of Similarities. If two texts contain similar words, phrases or ideas which distinguish them from other texts (i.e., if they both employ terms or phrases that are rare, or if they both use language in a peculiar way), this increases the likelihood that a literary connection exists. 29 This can be one of the strongest arguments in favor of dependence. 30
The first criterion (Explicit Reference) is not a necessary prerequisite for positing a literary relationship between texts; but if it does occur, such a connection cannot be denied. This criterion, in many ways, goes without saying. It is nevertheless included as an initial point of departure. Moving beyond the explicit indication of literary borrowing, the second criterion (External Plausibility) serves as an essential condition which must be met in order for a literary connection to be plausible. The next two criteria (Authorial Tendency and Recurrence) mark another category of evidence, providing indirect support for literary borrowing. When these factors are present, they serve to increase the probability that borrowing has occurred, but on their own they cannot be used to demonstrate literary dependence. In this way, they are supplementary considerations which add cumulative weight to existing evidence. The final criterion (Verbal Agreement) provides ‘the most objective and verifiable’ basis for making judgments related to the use of source-texts. 31 Therefore, when linguistic and thematic similarities occur, they deserve sustained attention.
Some will no doubt view such criteria with skepticism, questioning modern scholars’ ability to draw firm conclusions regarding literary relationships between two ancient texts. This perspective was trumpeted years ago by Andrew Walls,
32
who stated, [I]t is difficult to avoid the impression that this type of criticism has distinct limitations. There is a certain rigidity about it, which together with the opposing results often obtained, makes one suspect its universal accuracy in predicting the workings of a vigorous and well-stored mind, which had treasures to draw on that we know not of, and the situations to face that we can recognize only in dim outline (1959: 43-44).
But this level skepticism is neither justified nor always entirely consistent. 33 Guidelines like these do not represent mathematical formulas by which conclusively to prove literary dependence. It must be recognized that ‘no list of criteria, however sophisticated, can altogether clarify the fuzzy logic of intertextual referencing. Criteria are tests, not laws’ (MacDonald 2000: 8). 34 They are merely ways of facilitating the assessment of literary affinities between two texts with the goal of discerning the relative probability that borrowing has occurred. Furthermore, literary connections are not determined on the basis of any single criterion being met, but on the cumulative weight from multiple criteria. 35
How, then, might these criteria be used to move the discussion forward? The validity of any criterion is, of course, measured by its explanatory ability, which can only be assessed through its application to the relevant materials. Due to spatial constraints, the present study can only be suggestive in nature (hence its function as a ‘prolegomena’). A fuller examination is required before any intertextual deductions could be made. In this case, the ideal approach would be to apply the proposed criteria to each of the texts with which 1 Peter might have a literary connection. Where the evidence allows, the interpreter would seek to make plausible hypotheses about the existence of literary relationships, and, if possible, the direction of literary borrowing. 36 Only when this type of study has been carried out, and the evidence fails to support any literary connections, then and only then would the modern consensus have any grounds upon which to stand. Only then could one deny that there is a literary relationship between 1 Peter and certain other early Christian writings. This is the challenge that must be taken up in future studies.
Footnotes
1.
According to Osborne (1981: 65) the letter contains 31 references to the Old Testament, while Schutter (1989: 43) counts 41. Some find even more scriptural references in 1 Peter: 59 references (Hiršs 2003: 13-14); 64 references (
: 610-12).
2.
A sampling of those who espoused this view includes: Steiger 1832: 7-10; Wiesinger 1856: 22-27; Schott 1861: 336-40; Huther 1877: 19-26; Alford 1878: 135; Johnstone 1888: 34-35; Hort 1898: 4-5; Monnier 1900: 259-66. For a fuller defense of this view, see Seufert 1874; 1881; Foster 1913: 411-80. Not only was this theory widespread, it was held with considerable certainty. See, e.g., ‘It is one of the most solid results of criticism, that our epistle [i.e., 1 Peter] stands in direct literary dependence on the great epistles of Paul, particularly Ephesians’ (Bacon 1900: 153), and ‘The literary connection of I Peter with the later Pauline epistles is indubitable’ (: 246).
3.
In an article which challenges the existence of an early ‘Petrine circle’ to which the letter could be attributed, Horrell (2013) points out many of the strong connections between 1 Peter and certain Pauline writings (esp. Romans and Ephesians). In doing so, he successfully argues that 1 Peter was influenced, at least to some extent, by Pauline tradition. Others have likewise continued to stress the connection with certain Pauline epistles (e.g., Làconi 1967; Migliasso 1986; Mazzeo 2002: 29; ).
4.
Others have imagined a very similar situation. Cf. : 34: ‘it seems that the theology and expressions of Romans had worked their way into the life of the Roman church and were derived by our author from there; this is inherently probable if Romans was in use in the Roman church, as it surely was, and if 1 Peter was written from Rome’.
5.
It is interesting to note that Petrine interpreters still explain the similarities between certain texts (e.g., Synoptic Gospels, 2 Peter–Jude, Colossians–Ephesians) as direct literary borrowing (see, e.g., : 21); yet what they fail to explain is why the parallels between 1 Peter and other Christian writings must necessarily be explained through shared tradition.
6.
According to Brodie, one of the factors which can mislead or cause confusion within discussions of literary dependence is the postulation of alternative explanations like shared tradition. He notes that such a position ‘is possible, and cannot be directly disproved. But it is a gratuitous claim and cannot be proved. And since it bears the burden of proof – it claims … traditions for which there is no reliable evidence – it is in the weaker position’ (2004: 47; cf. : 204).
7.
Assuming the letter’s Roman provenance (1 Pet. 5.13), the grapho-literate Petrine author had both the opportunity and the capability to access Paul’s materials in written form.
8.
This concession is made on the assumption that both 1 and 2 Peter were composed by the apostle Peter (cf. Bigg 1902: 15; Holmer and de Boor 1986: 17; : 13). In 2 Pet. 3.15-16, the author acknowledges his familiarity with some Pauline letters, even admitting that ‘there are some things in them which are hard to understand’.
9.
It is difficult to estimate this cost with any precision, but on the basis of relative prices from Diocletian’s Edict on Prices in connection with that of raw materials, : 118) has calculated the price of a large manuscript such as the Great Isaiah Scroll (1QIsaa) as equivalent to several weeks of pay for a day laborer. This is not including the cost of actually writing out the text, which would have added to the price considerably.
10.
11.
Some distinguish between ‘literary’ dependence, often described as the use of the words of a source-text, and ‘conceptual’ dependence, usually understood to include the use of idea or concepts from a source-text (see, e.g., Nihan 2007: 430; Lyons 2009: 61, 65; : 66). We have incorporated the latter in the definition of the former, so that any borrowing from a hypotext is included.
12.
It is common to find scholars include citation from memory within their definition of literary dependence (see, e.g., Sanders 1966: 30 n. 9; Aejmelaeus 1985: 104, 134; : 55).
13.
: 105) defines direct literary dependence as ‘a conscious effort on the side of the alleged borrower (in our case, Peter) to make the quotation or reference look very much alike the original text (in our case, Romans)’. While literary borrowing might involve an exact replication, it could just as easily include a significant departure from the source text (see below).
14.
Examples include the imitation and adaptation of Homer’s Iliad and Odyssey in Virgil’s Aeneid (see Winn 2010: 11-33) and 2 Peter’s independent modification and restructuring of Jude (see : 25-32).
15.
Others have worked from similar methodological perspectives which seek to create very fine distinctions (e.g., Porter 2007: esp. 107-109, who suggests the use of five categories: formulaic quotation, direct quotation, paraphrase, allusion and echo). But after a certain point the multiplying of distinctions is unhelpful (cf. : 7). For a recent and helpful discussion of the use of the Old Testament in the New, see JSNT 38.1 (2015).
16.
17.
An alternative, which is becoming more and more popular within biblical studies, would be to take a literary approach and thus define the relevant terms according to the reception of the intertextual reference. From this perspective, an ‘allusion involves (1) the use of a sign or marker that (2) calls to the reader’s mind another known text (3) for a specific purpose’ (: 29). However, the question of function is one step removed from identification. Before examining why an allusory connection was made, one has to first determine that the connection was made – which is our concern here.
18.
The absence of criteria is not just a problem within Petrine studies, but a lacuna within biblical studies more generally, as noted in a recent work on intertextuality: ‘it may come as somewhat of a surprise to realize that there is no recognized self-evident method of tracing literary dependence, whether in biblical studies or elsewhere’ (Brodie, MacDonald and Porter 2006: 291).
19.
Cf. Horrell 2013: 17 n. 66. There are other problems with Shimada’s approach as well. Most notable is his dependence on form criticism as the ultimate solution to this question. He assumes (and frequently contends) that form criticism will assist scholars to convincingly answer the question of 1 Peter’s relationship with other early Christian literature. But, at the present, we stand at a point when form criticism has been all but discredited within gospels studies (see Tuckett 2009), and beyond the skepticism in which it is now held, it would still not be the best tool for resolving this particular issue. As : 220) pointed out, ‘even in the study of the Gospels, the use of the techniques of form-criticism does not obliterate the facts of literary dependence; the most ardent form critic will be the first to admit the literary dependence of Matthew and Luke on Mark, if not also on the hypothetical “Q”’ (original emphasis).
20.
Shimada himself recognizes the obstacle which this criterion creates when he later acknowledges, ‘it is needless to point out that I Peter has no mention of Paul’s name nor any remark as to which one of his epistles Peter is quoting from. This means that, if one takes ‘PP [Primopetrine] dependence’ as his [sic] working hypothesis, he [sic] does not have any solid ground to begin with’ (1998b: 106). This same argument is repeated by others as well (e.g., Elliott 2000: 37; : 12).
21.
If this criterion were used to assess literary dependence in the Greco-Roman world more generally, then one would be left to conclude that borrowing rarely occurred. Furthermore, some of the standard examples of literary dependence within biblical literature would have to be reconsidered (e.g., Synoptic Gospels; 2 Peter–Jude; Ephesians–Colossians).
22.
On the use of source material in the apostle Paul, see Koch 1986; Stanley 1992. On the use of source material in Greco-Roman materials, see Stanley 1990; : esp. 9-57.
23.
E.g., Edenburg 1998: 72-74; Carr 2001; Stackert 2007: 21-27; Leonard 2008; Lyons 2009: 59-67; Tooman 2011: 31-34; .
24.
E.g., Koch 1986: 11-23; Hays 1989: 29-32; Stanley 1992: 33-37; Allison 2000: 10-13; MacDonald 2000: 8-9; Brodie 2004: 43-49; Adamczewski 2010: 187-205; Winn 2010: 30-33; Nelligan 2015: 18-32. Cf. also .
25.
Brodie 2004: 44. Cf. also MacDonald 2000: 8; Winn 2010: 30; : 27-28.
26.
Cf. Hays 1989: 30; : 28.
27.
The value of this criterion is downplayed by : 52), who claims that ‘an assessment of verbal similarities relies too much on whether the majority readings of extant OG [Old Greek] Scripture, to which a possible citation might be compared, was really available [to] the New Testament writer in question’. It is true that the fluidity of the scriptural text might create difficulties if there were some discrepancies between a potential citation and the extant text of the OG; but in cases where there is considerable verbal and structural agreement, it can (potentially) be a strong indication of literary borrowing.
28.
Cf. : 249: ‘The presence of shared language may serve to indicate a connection between texts or traditions. More importantly, however, the fact that a text contains additional language that is idiosyncratic or not shared in no way undermines the possibility of a connection. Unique or idiosyncratic language may be a reflection of the creativity or writing style of a given author. It may even point toward an author’s use of multiple sources. It tells us very little, however, about the existence or nonexistence of allusions in the language that is shared with other texts’ (original emphasis).
29.
Cf. Hays 1989: 30; Edenburg 1998: 72; Leonard 2008: 251-52; Bergsma 2012: 66-68; : 30.
30.
This criterion is challenged, however, on the grounds that ‘[e]ven an event [or, in this case, a word or phrase] which is intrinsically highly improbably … is quite likely to occur if it is given sufficient opportunity to do so’ (Noble 2002: 250). In relation to two different stories (or texts), Noble contends, ‘The longer they are the more opportunity they afford for a resemblance to arise just by chance; and if they are sufficiently long it becomes quite likely that they will resemble each other in some way or other, even though the particular resemblances that happen to occur may indeed be improbable’ (: 250; original emphasis). The problem, however, is that within Noble’s constructed hierarchy, a remote possibility is afforded just as much weight (if not more) than something that is highly likely. Consequently, his argument fails to account for varying levels of probability that are inherent within any decision regarding literary dependence.
31.
Leonard 2008: 247; cf. also : 246-47.
32.
33.
It would be difficult, for instance, to approach the relationship between the Synoptic Gospels with the same mentality.
34.
35.
Cf. Edenburg 1998: 72; Leonard 2008: 253-55; : 32-33.
36.
For various criteria which can aid in determining the direction of literary influence, see Carr 2001; : 257-64.