In this article, we delve further yet into the murky depths of the Synoptic Problem, and a new actor appears on the stage: an ancient work known as the Didache, or The Teaching of the Apostles. (The name is usually pronounced “did-a-key”.) This was an ancient document that contained both ethical teachings and instructions for conducting church, and it may well date as early as the first century. Strangely, its relevance to the Synoptic Problem has been mostly overlooked until recently. I’m going to be assuming, for the sake of argument, some of the observations and conclusions I arrived at in previous articles, which I will summarize here:
- It is practically undisputed that Mark was the earliest canonical Gospel, and that both Luke and Matthew copied the majority of Mark into their own Gospels. If you are unfamiliar with the Synoptic Problem and source criticism, you should start with my article on Editorial Fatigue in the Synoptics.
- The passages shared by Luke and Matthew but not Mark (the “double tradition”) are usually explained by a hypothetical source called “Q” that both copied from. A minority of scholars propose instead that Luke copied Matthew or that Matthew copied Luke. After analyzing one such passage (the Parable of the Talents/Pounds), I tentatively concluded that Matthew copying from Luke fit the evidence well. The fit was even better if we assumed an early version of Luke (“Proto-Luke”) that resembled Marcion’s Evangelion.
- The passages in which Luke and Matthew agree against Mark are usually explained by Mark-Q overlap, but this hypothesis is not without its problems. The passage I analyzed (the Beelzebul Controversy), one of the most troublesome of these “minor agreements”, is also readily explainable if Matthew copied Luke (or Proto-Luke).
So far, my limited exploration of the Synoptic Problem has not required the existence of Q. It has, however, required me to set aside the widespread assumption (for which there is no real evidence) that Matthew was written earlier than Luke and be open to other possibilities.
If Q existed, what kind of document(s) would it be?
For decades, scholars have applied themselves to reconstructing Q as though it were a single, cohesive document — even though no such document is extant or attested by any ancient author. Furthermore, the label could easily apply to multiple, shorter documents. Although the law of parsimony implies that we don’t need a Q at all if Matthew copying Luke explains the state of our Gospel texts, if a document that fit the description were to fall into our lap, we should obviously try incorporating it into the Synoptic Problem. The Didache might be such a document.
The Didache and its relevance
The Didache was once known to us only through references by ancient Christian authors. However, a copy was discovered in a library in 1873, and scholars soon noticed that it had close correspondences with Luke and Matthew. However, because some parts of the document implied a post-Gospel dating, its significance for the Synoptic Problem was overlooked. Scholars assumed that the Didache’s textual parallels came about because it had either copied Luke and Matthew, or had copied a common source that Luke and Matthew used.
The Didache is widely understood to be a composite text that has been edited and expanded several times; but what New Testament scholar Alan Garrow determined some years ago was that much of the document appears to pre-date our Gospels (Garrow, 2004). And that means that the Didache itself could be a candidate for our missing Q document. (Or one of them, at least.)
Garrow has an excellent series of videos on his website explaining this hypothesis in broad outline, with an academic journal article forthcoming. I would like to analyze a key passage and see what we can learn about the relationship between Luke, Matthew, and the Didache.
The Sermon on the Mount (or Plain)
A large portion of the double tradition consists of the teachings that make up the Sermon on the Mount in Matthew and the Sermon on the Plain in Luke. The passages on loving your enemies (Luke 6:27-36 / Matthew 5:38-48, 7:12) have widespread agreement, including many identical words and phrases, with each other and with Didache 1:2-5. These passages are also notable for containing two of the most memorable sayings in the entire Bible.
I’ve taken the trouble of colour-coding all the direct correspondences I could find. (Any mistakes are my own.) The non-coloured material also includes numerous paraphrastic parallels.
We can clearly see, then, that there is material shared by Luke and Didache, Matthew and Didache, Luke and Matthew, and by all three of them. This seems to imply two sets of possibilities: (1) Either Luke and Matthew are copying directly from some version of the Didache, or the converse is true. If the Didache is the source, then for all intents and purposes, it should qualify as “Q”. (2) Either Luke is also using Matthew, or Matthew is also using Luke. Can closer analysis suggest answers to these questions?
The Didache as a paradigm of the “two ways” doctrine
Garrow’s videos show the application of several textual tools for identifying the direction of dependency. One is to look for fingerprints — distinctive evidence of one author’s editing that shows up in another work. Another is to analyze whether copying in one direction would require more difficult edits than copying in the other. I’ll explain these a little more by example.
One of the reasons, Garrow shows, that Didache 1.2-5 seems to be original is that it is composed with a deliberate structure that uses a common Jewish/Christian teaching framework known as the “two ways” or “two paths”. The basic structure of this teaching, which we also find in the first-century Epistle of Barnabas, is to introduce the two ways of life and death followed by positive commandments to love God and your neighbours as well as numerous, mainly negative, ethical commandments. For example, Barnabas includes the following content:
There are two paths of teaching and authority, the path of light and the path of darkness…This then is the path of light.…
Love the one who made you…
Do not entertain a wicked plot against your neighbour.
Do not make your soul impertinent.
Do not engage in sexual immorality.
Do not commit adultery.
Do not engage in pederasty. […]
Love your neighbour more than yourself.
Do not abort a fetus or kill a child that is already born. […and so on]
(Excerpts from Barnabas 18–19)
The Didache has the same kind of structure:
There are two ways: one of life, the other of death… Now, the way of life is this:
First, you shall love the God who made you;
Second, love your neighbour as yourself.
Everything you would not have done to you, do not do to another.
[Positive sayings that illustrate the Golden Rule inserted here.]
You shall not murder.
You shall not commit adultery.
You shall not corrupt boys.
You shall not fornicate. […]
You shall not murder a child by abortion or commit infanticide. […and so on]
(Excerpts from Didache 1:1–2.2)
An early editor of the Didache inserted a list of positive sayings into this two-ways structure right after the Golden Rule. (We even know that this section was missing from some versions.) What is striking about Luke is that he has reordered these sayings and put the Golden Rule in the middle. It is harder to argue the other way, that the Didache’s “two ways” structure is developed from Luke’s sermon.
Matthew has also reworked the Didache structure. He turns the two-ways material into a sequence of mini-sermons, each beginning with “You have heard that it was said…” and incorporated the positive sayings into two of these sermon blocks, while moving the Golden Rule to a later point in the text (7:12). His arrangement is also an innovation on the more traditional structure of the Didache, yet he retains much of the Didache’s exact wording.
In summary, the “fingerprint” of the Didache — its addition of positive ethical teachings to the two-ways framework — shows up in a reworked format in Luke and Matthew, implying that it is earlier.
The Golden Rule
The Golden Rule is perhaps the single most memorable quote of the New Testament, the jewel of Christian ethics. Its development in these texts is very interesting.
The Golden Rule is not an invention of Christianity; in fact, it appears on numerous occasions in the writings of earlier Greek philosophers, as well as in the writings of Jewish sage Hillel, but practically always in the negative: do not do to others what you don’t want done to you. This is the form in which it appears in the Didache.
The version found in Luke 6:31 (and Matt 7:12) is, by contrast, positive. Do to others as you do want them to do to you. Garrow argues that this modification was a natural editorial change to make, given Luke’s new setting for it — among a series of other positive commands where a negative one would seem out of place. In the process, Luke has created something entirely new: a positive formulation of the Golden Rule that has no prior attestation, but sounds so much better that it becomes the standard version from then on. According to Garrow, the positive version of this rule “does not appear anywhere in the ancient record before it appears in Luke and Matthew”.
It is unlikely that the Didache copied from Luke and Matthew, turning the Golden Rule back into a negative command.
What, if anything, suggests that Luke and not Matthew was the innovator? It seems to me that one hint is its placement in the two Gospels. What are the odds that Luke scoured Matthew, found the positive Golden Rule two chapters further on, and moved it into a passage that just happens to parallel the Didache’s negative Golden Rule? On the other hand, it is easy to see Matthew borrowing Luke’s positive Golden Rule and repositioning it within his Sermon on the Mount.
Furthermore, Matthew’s version contains some additional wording from the Didache that Luke’s doesn’t have. It is easy to imagine that Matthew, having both texts in front of him, preferred Luke’s positive Golden Rule but inserted some additional Didache text. It is harder to imagine that Luke performed the intricate editing operation of comparing Matthew to Didache and deleting the words that matched up. As you can see below, for Luke to have copied Matthew, he would have to have copied all the blue text while selectively omitting the red portions.
πάντα δὲ ὅσα ἐὰν θελήσῃς μὴ γίνεσθαί σοι, καὶ σὺ ἄλλῳ μὴ ποίει
In everything, as you would not have happen to you, do not do to another.
καὶ καθὼς θέλετε ἵνα ποιῶσιν ὑμῖν οἱ ἄνθρωποι, ποιεῖτε αὐτοῖς ὁμοίως
As you desire that men should do to you, do to them likewise.
Πάντα οὖν ὅσα ἐὰν θέλητε ἵνα ποιῶσιν ὑμῖν οἱ ἄνθρωποι, οὕτως καὶ ὑμεῖς ποιεῖτε αὐτοῖς
In everything, as you desire that men should do to you, likewise do to them.
Red = Matt + Didache, Blue = Matt + Luke, Pink = Matt + Luke + Didache
These examples, which I have probably done a poor job of explaining, make a reasonably good case for Didache as the earlier source, followed by Luke and then Matthew — at least as far as this passage is concerned. But let’s look at one more example.
Love your enemies
“Love your enemies” is also to be counted among the Bible’s most memorable sayings. This exhortation appears in Luke 6:27 and Matthew 5:44, with a similar expression in Didache 1.3. Its succinct, provocative wording is unattested before Luke and Matthew. Furthermore, it’s easy to see how it can be derived from the Didache.
Luke has arguably taken Didache’s “love those who hate you” (1:3cβ) and split it in half to create a couplet, “Love your enemies / do good to those who hate you”. (Luke’s additions are in bold.) The word “enemies” (ἐχθροὺς) is derived from Didache 1:3b, where it is used in its verb form, “hate” (ἐχθρῶν).
Luke then adds a second couplet, “Bless those you curse you / pray for those who mistreat you.” This is a verbatim quote of Didache 1:3b, except that Luke substitutes “mistreat” for “hate” (which he has already used in the first couplet). Luke does not copy Didache’s reference to persecution, a topic that he is hardly concerned with in his Gospel, unlike Matthew. (Luke’s only use of the word that implies a recent or present concern for the audience is in 11:49, a verse that did not appear in the Evangelion and might have been added later.)
Matthew simplifies Luke’s two couplets into one. He copies Luke’s “love your enemies” verbatim, followed by “pray for those who persecute you” — reinserting the word “persecute” from the Didache into the Lukan arrangement. And it should come as no surprise that ongoing persecution is a theme of Matthew’s Sermon on the Mount (appearing three times in 5:10–12) and elsewhere (10:23), hence his interest in using it here as well.
Garrow provides other examples in his videos demonstrating the apparent dependence of Luke and Matthew on the Didache.
Interestingly, Delbert Burkett (p. 90, see bibliography) has concluded independently that Luke’s arrangement of these commandments is secondary, and that the “love your enemies” quote (Luke 6:27) must have originally belonged next to its justification (Luke 6:32-33), “if you love those who love you, what credit is that to you? For even sinners love those who love them”. Sure enough — in the Didache, those lines appear side-by-side! In telling us what the Q source ought to look like, Burkett has unknowingly described the contents of Didache 1.3. (Noting, as an aside, that Didache has the more original “Gentiles” rather than Luke’s more appropriate “sinners”.)
Canonical Luke vs. Proto-Luke
There are few differences between the Evangelion and canonical Luke that can be identified with certainty. One well-attested, though minor, difference is slightly different phrasing of the beginning of 6:29. According to Harnack’s Greek reconstruction, this version started with ἐάν τίς (“if someone”), which happens to be a perfect match to the Didache. (Canonical Luke 6:29 now starts with τῷ, “to [the one who]”.) We may speculate that the former was how Luke originally read.
Matthew 5:45 has an interesting quotation that seems to derive from a source other than Luke and the Didache — though Luke’s parallel in 6:35 means more-or-less the same thing.
…so that you may be sons of your Father in [the] heavens; for he makes his sun rise on the evil and on the good, and sends rain on the righteous and on the unrighteous. (Matt 5:45)
The first part is equivalent to Luke’s “you will be sons of the Most High [or God]”, but the part about the sun and rain is interesting. It could, in fact, be an old rabbinical saying, because the expression “the sun rises on the good and the evil” in connection with God’s benevolence appears in a Jewish midrash on Canticles (Cant. Zuta 1.1; see Kister, p. 132). Perhaps Matthew saw the keyword πονηρούς (“the evil”) in Luke 6:35 and remembered a saying that expressed a similar sentiment in more poetic terms. I would suggest that Luke’s version (“for he is kind to the ungrateful and the evil”) is a little more to the point, and thus more original, since he has just described the virtues of giving to people who are ungrateful and evil. That said, it is hard to see Luke passing up such a nice expression if he had been copying from Matthew.
By all appearances, a strong case can be made that Luke and Matthew are using the Didache as a shared source in much the same way they use Mark. If these observations continue to hold up, then perhaps we should regard Didache as “Q” — or at least, one “Q” source. Furthermore, like the passages examined in earlier articles, this one too exhibits signs of Matthew copying from Luke. Alan Garrow’s forthcoming paper on the Synoptics and the Didache will undoubtedly shed more light on this topic. If this hypothesis persuades other scholars, will the field of New Testament studies be willing to reevaluate its long-held reliance on the traditional Q hypothesis?
Alan J. P. Garrow, “An Extant Instance of Q” (online video series).
Alan J. P. Garrow, The Gospel of Matthew’s Dependence on the Didache (JSNT Supplement Series 254), 2004.
Delbert Burkett, Rethinking the gospel sources: Volume 2, the unity and plurality of Q, 2009.
Menahem Kister, “Words and Formulae in the Gospels”, The Sermon on the Mount and its Jewish Setting, 2005.
Garrow’s videos were also recently discussed in the blogosphere at Exploring Our Matrix.