1. Introduction

 

Recently I went for a walk in woods adjoining a conference centre. I was familiar with the encompassing area but not with the woods themselves and so, after wandering for a while, I was glad to come across a noticeboard with a map which, on initial encounter, seemed likely to provide useful guidance. However, there was no red spot saying 'You are here' and no indication of where the conference centre lay in relation to the wooded area. I had no idea where I was in terms of the map, or where I would emerge if I travelled north, south, east or west. The map that promised to assist me turned out to be useless.

 

When Archbishop Bryennios stumbled on a manuscript of the Didache in 1873[1] he did not at first appreciate the enormity of his discovery. However, by the time the first critical edition was published, in 1883, its potential significance had become startlingly apparent; it created a first class literary sensation (Schaff 1885: 10-12). New editions, translations and commentaries appeared almost instantly as scholars hurried to devour a text that promised to reveal otherwise unknown details of worshipping life and organisation of the very early Church.  

 

Initial excitement turned to frustration and then disinterest. In the second half of the twentieth century the Didache became a text often referred to in passing but very seldom considered in detail. The discovery that promised so much could not be made to deliver on that promise because it proved all but impossible to determine its geographical, historical and literary context: was it a mainstream document or the rule of a backwater community? Did it belong to the late first or to the mid second century? Did it follow in the Matthean tradition or was it a much earlier text than that? Without some means of establishing these types of bearings the Didache remains as useless as the noticeboard I found in the woods.

 

The aim of this volume is to present a detailed map of the relationship between Matthew's Gospel and the Didache. Widespread acceptance of some form of connection between these two texts[2] makes this a logical starting point for any attempt to locate the Didache within the wider web of early Christian life and literature. However, recognition that a relationship exists between these texts is by no means the same as agreement regarding the nature of that relationship.

 

Historically, scholars have been divided between those who explain the connection as a product of the Didache's direct dependence on, or allusion to, Matthew's Gospel,[3] and those who see this as resulting from the dependence of both texts on common tradition(s).[4] In what follows I shall argue for the further possibility that the author of Matthew's Gospel depended directly upon a version of the Didache essentially similar to that rediscovered by Bryennios, except for the absence of Did. 8.2b;[5] 11.3b; 15.3-4 and 16.7.[6] This view has not previously received detailed consideration.[7]

 

The lack of scholarly interest in the possibility that Matthew directly depended on the Didache invites, at the start of a book on the subject, some explanation. Two features of the Didache may account for a widespread assumption that it does not pre-date Matthew's Gospel, and thus cannot have been used in the creation of that text. First, the single 'full' manuscript of the Didache was rediscovered within a collection of writings from the Apostolic Fathers.[8] Consciously or otherwise this context may have led some of the earlier students of the text to see it as secondary to the witness of the gospels. A small initial assumption of this kind can, if repeated sufficiently often, lead to the establishment of a widespread scholarly consensus.[9] 

 

A second argument against a pre-Matthean dating of the Didache is the presence of four references to 'the gospel' (8.2.b; 11.3b; 15.3-4) amid, in the case of 8.2b especially, passages that relate closely to Matthean material. These references have been understood by some scholars[10] as demonstrating knowledge of a written and authoritative Gospel of Matthew.[11] This, in turn, may be taken as suggesting a post-Matthean date for the whole text.

 

The essential flaw in this latter observation lies in its assumption that what may be true for Did. 8.2b; 11.3b; 15.3-4, must also be true for the whole of the remainder of the text. This assumption rests on a highly dubious foundation given that the Didache is widely recognized as having a complex compositional history.[12] This feature allows the possibility that the four references to 'the gospel' are among one of the later contributions to the whole,[13] in which case earlier elements could theoretically have been known and used by the author of Matthew's Gospel.

 

Despite these logical flaws it is understandable that, all other things being equal, the references to 'the gospel' and the Patristic context of the Jerusalem manuscript have favoured a post-Matthean consensus. However, as is often the case, all things are not equal.

 

The combined implication of two widely accepted factors has been consistently overlooked in previous studies of the relationship between the Didache and Matthew's Gospel. These two factors are, the complex compositional history of the Didache and the large number of points of similarity between the two texts. Taken together these lead, as I shall argue in detail below, towards a relatively simple explanation of the two texts' relationship; namely, that various elements (disparate in terms of style, origin and age) were incorporated into the Didache over time and that, at a later date, Matthew drew on the resulting text in the construction of his gospel. Later still, I shall propose, referals to 'the gospel' were inserted into the Didache (at 8.2b; 11.3 and 15.3-4) to abrogate certain teachings in favour of the similar, but now more developed,[14] instructions also to be found in Matthew's Gospel.

 

Given that the composite nature of the Didache is all but universally acknowledged, it is striking that none of the scholars who argue for the Didache's use of Matthew's Gospel address the difficulties this creates for their conclusions. For example, Massaux (1993: 176) ignores the issue entirely, and so is capable of concluding that

 

All these remarks lead me to recognize and affirm that the Didache has come under a very profound literary influence from Mt., to the point of wondering sometimes whether it is a tracing of the first gospel.

 

If Massaux had considered the possibility that the Didache was composed by more than one person, then his conclusion would require these different contributors to have worked together, in some remarkable way, to achieve such a 'tracing'. However, Massaux makes no attempt to explain how this could have happened.

 

Tuckett (1996: 93), unlike Massaux, is not guilty of entirely ignoring the issue of the Didache's compositional history. He makes this very important observation:

 

Any discussion of the problem of synoptic tradition in the Didache must take note of the question of the unity of the text ... It is almost universally agreed that the present text is, in some sense at least, "composite". Didache 1-6 incorporates an earlier Two Ways tradition attested also in the Epistle of Barnabas 18-20, Doctrina Apostolorum and elsewhere; further, within this Two Ways tradition, the section 1:3-2:1 is probably a secondary, Christianising addition. Other seams within our text have been suggested: for example, chapters 8 and 15 may be secondary additions to an earlier Vorlage. The precise number of stages of redaction which one should postulate is much debated. Nevertheless it is clear that any theories about the origins of synoptic tradition in one part of the Didache will not necessarily apply to the Didache as a whole. Each part of the text must therefore be examined separately and, to a certain extent, independently.

 

However, despite noting this crucial point, Tuckett proceeds to overlook its implication. He thus argues, 'separately and, to a certain extent, independently', that the Didache alludes to Matthew's Gospel at each point where they share parallel material. However, what this conclusion requires is that the authors of the Two Ways, the section 1.3-2.1, chapters 8 and 15, etc., all behaved in a remarkably similar manner in their use of Matthew's Gospel, despite the widely differing style, origin and content of the various sections involved. This circumstance invites an explanation. Such is not offered by Tuckett.

 

Scholars who propose that the Didache and Matthew's Gospel are principally related via the use of shared common traditions[15] have also ignored the full implications of the Didache's compositional history. For example, Köster (1957: 239) believes that five separate units were incorporated into the Didache, four of which drew upon sources such as Jewish paranesis, liturgical modes of expression, Old Testament sayings, Jewish rules, meschalim and completely enclosed pieces of Jewish tradition, which were also used by the author of the Matthew's Gospel. This type of theory relies on very high levels of coincidence in that Matthew's Gospel and the Didache must both happen to draw on the same selection of Jewish paraneses, liturgical modes of expression, Old Testament sayings, etc.. This problem is exacerbated when it is noted, with John Court (1981: 111), that the majority of the material shared by the two texts is 'from Matthew's special material or from the synoptic traditions at points where Matthew's distinctive rendering is preferred'. This means that Matthew and the Didachist are required, not only to share a varied collection of common sources, but also repeatedly to take from them uniquely similar selections. The levels of coincidence rise still further when questions are asked regarding the adequacy of Köster's extremely brief, though confident, analysis of the compositional history of the Didache.[16] In the discussion that follows I aim to show that it is highly unlikely that the complex and sometimes self-contradictory text of the Didache was put together at one time by a single compiler. If the text was compiled by more than one author over a period of time, as most scholars suppose,[17] then Köster's theory requires yet another layer of coincidence. Thus, two or more contributors to the Didache are required, over a period of time, to have quarried uniquely similar material from diverse previously existing traditions also selected by Matthew.

 

My contention is not that the above pattern of events is strictly impossible, simply that it is a highly improbable explanation for the presence of twenty-six points of similarity between the two texts.[18] It is more credible to propose that Matthew used the Didache, a composite text containing all the 'shared traditions' required to explain their several points of distinctive similarity.

 

In the chapters that follow I shall explore the possibility that Matthew knew and used a version of the Didache to which the references to 'the gospel' had not yet been added. The hypothesis of Matthew's use of some parts of the Didache requires only that the Didache is composite to a degree that is already widely accepted. However, an attempt to identify precisely which parts of the Didache were used by Matthew invites a full compositional analysis. The first stage of my argument therefore involves a detailed exploration of the compositional history of the Didache.

 

Past experience of this type of exercise will persuade most readers that this endeavour is unlikely to prove either successful or especially interesting. However, this case offers more than usual cause for optimism on both counts. The Didache is a text in which the joins between various sources may, in a significant number of cases, be relatively simply identified. Further, because the Didache contains several interpretative puzzles that are bound up with compositional questions, there is interest to be had in the need to wrestle with some of the Didache's infamous riddles while engaging in the study of its history.

Home Cube Book information Concluding Chapter



[1] Bryennios discovered this manuscript within a volume belonging to the library of the Monastery of the Holy Sepulcre ("Jerusalem Monastery"). It was published by him in 1883 and was moved, in 1887, to the Greek patriarchate of Jerusalem where it remains under the reference Kw~d. patr. 54. The manuscript is variously known by the names, Bryennios, Constantinople, H. 54 or Jerusalem. Hereafter, I shall refer to it as the Jerusalem manuscript. Van de Sandt & Flusser (2002: 16-24) give a positive assessment of the quality of this manuscript. They suggest that 'Despite the late date of its origin [1056 CE], the Jerusalem codex in some respects measures up to the Alexandrinus and Sinaiticus' (p. 18), and suggest that 'it may have originated in the patristic period' (p. 23).

[2] All the scholars listed in notes 3,4 and 7, immediately below, perceive some form of link between Matthew's Gospel and the Didache.

[3] Draper (1996a: 16) 'The earliest commentators on the text usually assumed the use of Matthew's gospel, especially since the Didache itself four times refers to "the gospel" in 8:2; 11:3; 15:3,4. Representative of those who affirm such a dependence is E Massaux.' Layton (1968), Köhler (1987: 19-56, esp. 29-30), Wengst (1984: 25-30, 61-63), Luz (1989: 93), Butler (1960 and 1961), Vielhauer (1965), Giet (1970) and Tuckett (1989 reproduced 1996) also support this view. Tuckett, being the most recent and thorough exponent of the Didache's presupposition of Matthew's Gospel will be my chief dialogue partner, with respect to this view, in the following discussion.

[4] Draper (1996a: 17) 'The debate was thrown wide open with the publication of the important study of Helmut Köster in 1957, which argued that although the reference to eu0agge/lion might point to a knowledge of a written gospel by the "compilator" of the Didache, the latter did not himself use such a source. The Didache did not stand after the gospel writers but alongside them, utilizing the same traditions.' Scholars who similarly see the Didache and Matthew's Gospel as dependent on shared common traditions include: Audet (1958: 166-186), Glover (1958: 12-29 and 1985: 234-251), Rordorf and Tuilier (1978: 83-91), Rordorf (1981: 499-513), Mees (1971), Kloppenborg (1979), Draper (1996b: 72-91) and Jefford (1989: esp. 91,160-61). This understanding of the relationship between Matthew's Gospel and the Didache has been in the ascendant since Köster's 1957 volume, although severely questioned by Tuckett and others mentioned in note 3 above.  

[5] A full text of the Didache, which shows the verse sub-divisions referred to here and elsewhere, may be found at the beginning of this volume.

[6] In the following text, wherever the proposed dependence of Matthew's Gospel on the Didache is discussed, then Did. 8.2b; 11.3b; 15.3-4 and 16.7 should always be understood as excluded from the text of the Didache that (I argue) was known by Matthew.

[7] Draper (1996a: 18-19) suggests that both texts evolved in the same community so that, 'the influence, then, could run in either direction, depending on the redactional layer of the text'. Draper stops short, however, of straightforwardly proposing the direct dependence of Matthew on the Didache, especially with respect to Did. 1.3-6; 8.1-3; 15.3-4, which he sees (1996c: 227 and 1996b: 76) as belonging to the latest redactional layer of the Didache. In Draper 1996d: 347 Did. 1.2-6 and chapters 8 and 15 are seen as later insertions. (Three of the above-mentioned articles were first published prior to Draper's 1996 edited collection of articles. Thus, 1996b first appeared in 1985; 1996c in 1993; 1996d in 1991. Full details may be found in the bibliography.)

[8] Jefford (1989: 18) notes that the Didache has been largely accepted, 'into the informal canon of early Christian literature, i.e., the Apostolic Fathers.'

[9] Not all of the earliest generation of Didache scholars assumed its post-apostolic date. For example, Sabatier (1885) considered it to be extremely early, but their views were largely ignored by contemporary and later scholarship. Some advocates of a very early date for certain parts of the Didache have reappeared in the latter half of the twentieth century. For example, Audet (1958) and Mazza (1995).

[10] Streeter (1924: 507) claimed that the Didachist seemed 'not only to have read Matthew, but also, like Ignatius, to refer to it under the title of "The Gospel"'. Carrington (1957: 500) described the Didache as 'an appendix to Matthew, to which [its] readers ... are explicitly referred no less than four times'. See also note 3 above.

[11] I accept the view that Matthew's Gospel is likely to be the text behind the Didache's references to 'the gospel'. This question is discussed in section 8.A.3. 

[12] Rordorf (1991: 396) 'The Didache  cannot, of course, be considered a homogenous text. Even those who attempt to attribute it to a single author must unhesitatingly grant that older material is used in it. This is especially true in the first five chapters.' Draper (1996b: 74-5) '... the text shows signs of considerable redactional activity, which defies any theory of unity of composition, even allowing for the activity of an interpolator. The Didache is a composite work, which has evolved over a considerable period.' Tuckett (1996: 93) 'It is almost universally agreed that the present text is, in some sense at least, "composite"'. See also scholars noted in the introduction to Part I, below.

[13] There is every reason to suppose that this is the case since references to 'the gospel' are embedded in elements of the Didache that are generally, and justifiably, recognized as being among the younger strata of the text. A full analysis of the compositional history of the Didache, including the relative age of the four references to 'the gospel', is conducted through the course of Part I.

[14] The concept of 'development' is fraught with difficulties in that traditions do not always 'develop' in uniform or predictable ways. For this reason developmental arguments are used sparingly in what follows. It is nonetheless the case, however, that the standard indicators of development speak for the antiquity of the Didache's traditions relative to those in Matthew's Gospel. This has been an important factor for those scholars who see the Didache as preserving traditions that were also used, and developed, by Matthew. Thus, van de Sandt (2002: 229) is not alone in expressing the view that, for example, in the cases of: Did. 8.1 and Mt. 6.16; Did. 8.2 and Mt. 6.9-13; Did. 9.5 and Mt. 7.6; Did. 11.7 and Mt. 12.31, 'The Didache ... witnesses to an earlier stage of the Matthean tradition'.  

[15] It is significant that this view, introduced by Köster in 1957, arose to take account of the observation that, although it was thought impossible that the Didache could pre-date Matthew's Gospel the Didache nonetheless frequently appears to preserve a more ancient version of traditions shared with Matthew. 

[16] Köster devotes just two pages to this complex question (1957: 159-160). Here he states that five separate units of previously existing tradition, more or less accurately reproduced by the compiler, may be identified 'Ohne Schwierigkeiten' - without difficulty.

[17]  Kraft (1965: 1-4) describes the Didache as 'evolved literature' which is the product of a community and not an individual. Giet (1970) takes a similar view in seeing the text as responding to changing circumstances in a process of organic development. This type of view is endorsed, with some reservations, by Draper (1996a: 19-22). Jefford (1989: 20) suggests that Kraft's view is representative of a consensus amongst more recent scholarship. 

[18] A list of the points of contact discussed in the ensuing text may be found at the end of chapter fifteen.