The Development of the Lord’s Prayer

It’s time for another look at the Synoptic Problem. This time I’m interested in a very well-known piece of text called the Lord’s Prayer or Pater Noster, which is how it begins in the liturgical Latin version.

Recitation of the Lord’s Prayer from the King James Bible was a daily ritual at my childhood school, so it is something I know by heart like perhaps the majority of Christians over the past two thousand years. It is precisely this familiarity that makes the textual development of the prayer difficult to analyze, since scribes had a strong tendency to correct the version in front of them with the version they knew from memory. Thus, the version we find in Luke 11 of the late Greek Textus Receptus — and in English versions that are based on this text, like the King James — is nearly identical to that found in Matthew 6. But it was not always so.

The prayer is absent from Mark and is often regarded as originating with the Q source — the hypothetical lost document of sayings that underlies all the shared Luke-Matthew material. It also appears in the Didache, a document which British scholar Alan Garrow has shown (to my satisfaction) actually was a source for Luke and Matthew.

Now, I primarily write these articles as a teaching tool for myself, and I am always surprised by what I discover. Here’s a summary of what I’ve covered so far.

  • I began with a post on some examples of editorial fatigue that show quite conclusively the reliance of Luke and Matthew on Mark (which scholars have long believed anyway).
  • I continued with an analysis of the Parable of the Talents/Pounds, a story found in Luke and Matthew but not Mark, and therefore part of the “double tradition”. Several explanations for the origin of this story were examined: the Q hypothesis, Goodacre’s hypothesis that Luke copied Matthew, and more recent hypotheses by Garrow and Klinghardt that Matthew copied Luke (possibly the early version used by the Marcionites). The last option seemed to offer the simplest solution that fit all the evidence.
  • I looked at an example of “minor agreement” wherein Luke and Matthew appear to make the same changes to a story found in Mark: the Beelzebul Controversy. Once again, Matthean Posteriority (Garrow’s position) provided the best explanation.
  • I examined Garrow’s proposition that the Didache’s “two ways” teaching was the source behind certain teachings in the Sermon on the Mount/Plain, notably the Golden Rule. In a way, this makes the Didache a “Q” source.

What if we analyze the Lord’s Prayer in the same way? Can we explain the differences by Matthew’s use of Luke? And how does the Didache fit in?

The Lord’s Prayer in canonical Luke

Our oldest manuscripts of Luke (e.g. P75 and Codex Vaticanus) have a significantly shorter version of the Lord’s Prayer than Matthew and later manuscripts of Luke do. In them, the prayer reads as follows:

I. Father, hallowed be your name.
II. Your kingdom come.
III. Give us each day our daily bread.
IV. And forgive us our sins, for we ourselves forgive everyone indebted to us.
V. And do not lead us into temptation.

By the principle of lectio brevior — the shorter reading is likely more original — we should favour this version Luke’s prayer as older than Matthew’s. Scribes are more likely to embellish a formulaic text like the Lord’s Prayer than to pare it down.

Michael Goulder, one of the few scholars firmly in the camp that Luke copied Matthew, admitted the difficulties with this passage. “It is not so easy to be confident of [Luke’s] motives for omission, in the nature of the case,” he wrote in his commentary on Luke (1989, p. 497). Goodacre, the other main proponent of Lucan posteriority posits that Luke has rewritten the prayer “in line with the version more familiar to him” (2001, p. 139). This strikes me as a plausible but somewhat ad hoc explanation, framed to support an existing conclusion rather than to explain the available evidence.

The case Luke’s prayer was copied and embellished by Matthew (whether he got it from Luke or from Q) is strengthened by the distinctly “Matthean” nature of many of his additions.

Another, possibly older, Lucan version

Several ancient witnesses attest to a different beginning to the Lord’s Prayer in Luke. According to church fathers Gregory of Nyssa and Maximus the Confessor, as well as two Greek manuscripts (mss 162 and 700), the Lord’s Prayer did not say “Your kingdom come”. Instead, it read “Your holy spirit come upon us and cleanse us”. This petition was also part of Marcion’s Evangelion (an early and possibly more original form of Luke) according to Tertullian, who does not suggest it should be otherwise. And Codex Bezae gives partial support for this reading, retaining the phrase “upon us”, which is explicable only as a remainder of a previous reading (Streeter, p. 277).

I. Father, hallowed be your name.
II. Your holy spirit come upon us and cleanse us.
III. Give us each day our daily bread.
IV. And forgive us our sins, for we ourselves forgive everyone indebted to us.
V. And do not lead us into temptation.

There are several lines of evidence suggesting that this is in fact the original reading of the prayer — at least in written form. One indicator occurs in the prayer’s immediate context. It is followed by a teaching on prayer that includes reassurance about receiving the holy spirit — something that fits perfectly if asking for the holy spirit was part of the prayer, but is somewhat out of place otherwise.

If you then, who are evil, know how to give good gifts to your children, how much more will the heavenly Father give the holy spirit to those who ask him!” (Luke 11.13)

In fact, we could consider this an example of editorial fatigue. An ancient scribe changed the prayer to contain Matthew’s Jewish sentiment “your kingdom come”, but failed to alter the following passage that refers back to asking for the holy spirit.

Furthermore, the cleansing of the holy spirit is a theme that seems to have been part of the liturgy in the earliest Christian communities. The connection between baptism and receiving the holy spirit is made explicit on numerous occasions in the New Testament.

“Repent and be baptized every one of you in the name of Jesus Christ so that your sins may be forgiven; and you will receive the gift of the Holy Spirit.” (Acts 2.38)

He saved us, not because of any works of righteousness that we had done, but according to his mercy, through the water of rebirth and renewal by the Holy Spirit. (Titus 3.5)

This would strengthen the proposal put forward by Leaney in particular (1956) that Luke adapted the prayer from a liturgical formula used during baptism and the eucharist. See footnote 1 below for a bit more detail on this version.

The Accursed Fig Tree by James Tissot
The Accursed Fig Tree by James Tissot

Markan material in Luke’s prayer

The Lord’s Prayer might not be independent from Mark after all. If we look at one of the very few passages in Mark where Jesus gives instruction on praying, we find the following:

“Whenever you stand praying, forgive, if you have anything against anyone; so that your Father in the heavens may also forgive you your trespasses.” (Mark 11.25)

In Mark, this teaching is given in the context of Jesus cursing the fig tree—a pericope that Luke omits from his own Gospel, though he surely knows about it. Instead, he has taken this very brief teaching that associates prayer with forgiveness—both ours and God’s—and has paraphrased it in the fourth petition of the Lord’s Prayer.

Another reference to prayer in Mark can be found in the Garden of Gethsemane story:

Keep awake and pray that you may not enter into temptation; the spirit indeed is willing, but the flesh is weak. (Mark 14.38)

Here, Jesus has instructed Peter on another use of prayer: to avoid being tempted or tested—εἰς πειρασμόν in Greek, a phrase somewhat ambiguous in meaning. Luke has made this the final petition of his prayer, an entreaty not to be led into temptation (using the same wording as Mark, εἰς πειρασμόν).

Daily bread each day

The request to receive “daily” bread could indicate the prayer’s use at eucharistic meals (so Leaney speculates). This petition is troublesome, since the word translated “daily”, ἐπιούσιον, is actually uncertain in meaning and attested in no Greek source aside from this prayer and its parallels in Matthew and the Didache (Betz 398-399). Perhaps the author of Luke himself wasn’t even sure what it meant, and added “each day” (καθ’ ἡμέραν) to clarify. Then again, the same phrase (καθ’ ἡμέραν) appears in LXX Exodus 16.5 concerning the manna God gave to the Israelites each day, and Luke might intend a direct analogy with the provision of manna through this prayer.

Gethsemane by Wassilij Grigorjewitsch Perow, 1878
Gethsemane by Wassilij Grigorjewitsch Perow, 1878

The Lord’s Prayer in Matthew

The Lord’s Prayer in Matthew (11.9–13) has been somewhat expanded from the version found in Luke.

I. Our father in the heavens, hallowed be your name.
II. Your kingdom come.
III. Your will be done, as in heaven, also on earth.
IV. Give us today our daily bread.
V. And forgive us our debts, as we also have forgiven our debtors.
VI. And do not lead us into temptation, but rescue us from evil.

First, we can see that Matthew has more fully quoted Mark 11.25 by including “in the heavens”.

Additionally, he has recognized the Gethsemane allusion in the prayer and added a new one. It would only be fitting that the one example of Jesus praying in Mark should be used as a model for Christian prayer, so he paraphrases Mark 14.16, “not my will, but yours be done,” as the third petition in his prayer, “your will be done”. (See Goulder 1963 for a thorough analysis.) The comparison of heaven and earth is a common theme in Matthew and can be found elsewhere in the Greek Septuagint.

The clarification added to the last petition is probably intended to resolve a theological issue. The word used for temptation, peirasmos (πειρασμός), can mean testing by God as well as the lure of the devil, and the fact that this is a request made to God suggests the former. The idea that God tempts and tests human beings—the righteous in particular—is fairly common in Jewish theology and can found in the Old Testament as well as in 1 Cor. 10:13. (See Betz, p. 406, 408) Matthew, however, finds it necessary to counter this by clarifying God’s positive role of “rescuing” us from evil (or “the evil one”), effecting a reinterpretation of 12a.

Betz (p. 407) notes that James 1:12–15 seems to stand in direct contradiction with the Lord’s Prayer (and Paul) on the subject of temptation, and there is no indication that the author of James even knew of the Lord’s Prayer.

No one, when tempted, should say, “I am being tempted by God”; for God cannot be tempted by evil and he himself tempts no one. (James 1:13)

James Tissot, Le “22Pater Noster”
James Tissot, Le “Pater Noster” (The Lord’s Prayer)

The doxology and further revisions

Matthew has generally been the church’s favourite Gospel, and its version of the Lord’s Prayer became the one people were most familiar with. It is not surprising, then, that most of Matthew’s additions ended up being copied back into Luke beginning in the fourth and fifth centuries.

Another major change was the addition of the final doxology, which all those familiar with the prayer from church will recognize:

For thine is the kingdom, and the power, and the glory, for ever. Amen. (Matt. 6:13c KJV)

This ending is not found in any of our oldest and best manuscripts, and new Bible translations generally omit it. It was probably spoken in church liturgy fairly early on—perhaps by the congregation in response to the reading of the prayer by the liturgist—and eventually inserted into manuscripts of Matthew because of its familiarity to scribes. (Betz p. 414) One possible origin for this distinctly Jewish doxology can be found in 1 Chronicles 29:10–11:

“Blessed are you, O Yahweh…forever and ever. Yours, O Yahweh, are the greatness, the power, the glory, the victory, and the majesty; for all that is in the heavens and on the earth is yours; yours is the kingdom, O Yahweh, and you are exalted as head above all.

The Lord’s Prayer in the Didache

Building on Garrow’s keen observations that the “two-ways” teaching in the Didache is the textual source for the “love your enemies” sermons in Luke and Matthew, with Matthew further using Luke as a source, it would sure keep things neat and tidy if we found that the same was the case for the Lord’s Prayer. Instead, in Didache 8.1-3, we find a version of the prayer that is almost identical to Matthew’s, and indeed may have been copied from Matthew.²

This turns out not to be such a problem, though, because the Didache is recognized as having undergone multiple revisions (Garrow 2004), and Did. 8.1-3 in particular is recognized for unrelated reasons as being a later addition. Draper (1996 p. 85, cited by Garrow 2004 p. 124) observes:

Did. 8 appears to be a later addition to the earliest text of the Didache. It is inserted after the reference to the baptismal fast in 7:4, but it has quite a different reference to ‘stationary fasts’ and daily prayer. It breaks up the natural flow in the catechetical manual from baptism to the eucharist. Moreover, it is not introduced by the formula which characterizes the liturgical sections of the Didache…and in the Ethiopian version it is set after 11:3-13.

Additional evidence lies in the fact that the passage with the Lord’s Prayer (like some other late additions to the Didache) cites the “gospel” as its source and paraphrases Matt. 6.5:

And do not pray as the hypocrites, but as the Lord commanded in his gospel. (Did. 8.2a-b)

This suggests to me the possibility that while the core of the Didache predated the four New Testament gospels, the Lord’s Prayer was added by a scribe who was familiar with Matthew — perhaps as his community’s only gospel text, and at a time before it had acquired its canonical title. (It should be remembered that our titles for the New Testament gospels are likely the result of second-century guesswork.)

Interestingly, though, the Didache does include a doxology that is simpler than one that was added to Matthew in later centuries.

For the power and glory are yours forever. (Did. 8.2c)


1. The possibility that the true text of Luke 11.2 read “Your holy spirit come upon us and cleanse us,” is often dismissed for lack of direct early manuscript evidence. Nonetheless, it is put forward by, among others, Streeter (1930), Leaney, and BeDuhn (2013), the latter citing a reference to eighteen other modern scholars who favour this view. The presence of this reading in diverse witnesses spanning the second through twelfth centuries suggests that it should be taken seriously. In particular, Greek ms 700, though late (eleventh century), contains a variety of unusual readings generally considered original despite their rarity, including the absence of “…who is in heaven” in Luke 11.2 and the name “Jesus Barabbas” in Matthew.

2. Garrow, at least as of 2004, believed that the minor discrepancies between Matthew and the Didache — particularly replacing “heavens” with the singular “heavens” — posed a problem for the copying of Matthew by the Didache (2004 pp. 134–137). However, I think the notion that the author of the Didache embellished the prayer from Luke is more problematic, and Garrow’s arguments seem weaker here than elsewhere. I don’t know Garrow’s current position in light of his more recent work on Matthew’s use of Luke. At any rate, some sort of literary relationship is universally acknowledged.


Alan Garrow, “Didache and Matthew”, viewed Oct. 21, 2015.

Alan Garrow, The Gospel of Matthew’s Dependence on the Didache, 2004.

Robert Leaney, “The Lucan Text of the Lord’s Prayer (Lk xi 2-4), Novum Testamentum 1.2, 1956

Hans Dieter Betz, The Sermon on the Mount (Hermeneia), 1995.

Jason BeDuhn, The First New Testament: Marcion’s Scriptural Canon, 2013.

Michael Goulder, “The Composition of the Lord’s Prayer”, J Theol Studies 14.1, 1963.

B. H. Streeter, The Four Gospels: A Study of Origins, 1930.

J. A. Draper (ed.), The Didache in Modern Research, 1996.

17 thoughts on “The Development of the Lord’s Prayer

    • Sounds like you’re being facetious. 🙂 I welcome any rebuttal to the specific points I’ve examined so far. To be honest, I expected the standard Q hypothesis to provide the best explanation before I started comparing Synoptic passages line-by-line for myself.


      • Well, I sat in Alan’s early presentation of his hypothesis at a British New Testament conference around 10 or 15 years ago, and well remember the stunned (and incredulous) silence at the end. A friend heard him last week in Cambridge, and I think there was the same reaction. To say this is a minority view in NT studies is an understatement.


      • I am generally sceptical that you can use criteria like lectio brevior to say anything about relative dating. I don’t think it would be hard to find places where Matthew shortens Luke, Luke shortens Matthew, and Mark shortens Luke. In that case, all are earlier than all the others.

        Besides, you would need to add evidence that this is an habitual scribal or literary practice—which it clearly isn’t. I think the structural/poetic arguments about the prayer as we find it in Matthew offer compelling evidence that this is an early form.


  1. I would agree that these text-critical principles are not a substitute for definitive evidence. However, there seems to be widespread support (certainly among all the books and papers I consulted) for the idea that Luke’s version is more original — though some version of the Q hypothesis is usually marshalled to explain this.

    As for the minority Garrow/Klinghardt hypothesis, I am interested to see if it holds up when I investigate more double tradition and minor agreement material.


  2. I see the “prayer” as a compilation of headlines or identifiers from Rabbinical prayers, which you will still fnd in the (rarely read) long preamble to daily prayers, right next to the bit about thanking God for not having made me a woman. I and II survive i the Kaddish, and one passage reads “and do not lead us, either into temptation or into shaming, and remove the Satan from before us and behind us”.

    I do not see how any serious scholarly discussion of the “prayer” (or on my reading, list of prayers) could ignore this Pharisaic liturgy; Jesus in the disputations sounds, of course, like a Pharisee (hence resurrection of the flesh) of the school of Shammai (hard line on divorce).


    • Thanks for the comment, Paul. I purposely avoided going into too much background on the Kaddish and other rabbinical prayers, since my interest here is focused on the synoptic problem and the textual relationship between the Christian forms of the prayer. I would agree that much of it probably has the background you suggest.


  3. I have a recently published paper advocating Goulder’s theory that Luke abbreviated Matthew’s version of the Lord’s Prayer [Ken Olson, “Luke 11.2-4: The Lord’s Prayer (Abridged Edition)”, in Marcan Priority Without Q: Explorations in the Farrer Hypothesis, edited by John C. Poirier and Jeffrey Peterson (London: Bloomsbury T&T Clark, 2015) 101-118]. I argue that the principle that liturgical texts become longer in transmission cannot be applied to the Lord’s Prayer, both because there are notable counterexamples (the Hebrew and Syriac versions of Apocryphal Psalm 151.1-2 is perhaps the clearest case) and also because Luke is not a copyist of a liturgical text, he is an author composing a narrative comprised of episodes in a sequence. Luke’s abbreviations can readily be explained on the basis of his redactional tendency to abbreviate his sources by eliminating repetition. He eliminates the further identification of the Father as the one in heaven, having already identified him as “Lord of Heaven and Earth” in Luke 10.21 (every subsequent address to the Father in Luke has simply “Father”). He omits “thy will be done on earth as it is in heaven” because it restates the content of “thy kingdom come” which immediately precedes it, and “deliver us from evil” because it restates positively what “lead us not into temptation” has said in the negative. Luke has omitted unnecessary repetitions but has retained each separate thought from Matthew’s prayer.


    • Thank you very much for the comment, Dr. Olson. I did not come across your article during my research, probably because of how recent it is. I will certainly read it if I get the opportunity.

      I am, of course, familiar with Goulder’s arguments, and there is certainly a plausible case to be made. I will have to think about them a bit more, but I get the sense that these are secondary arguments that have more weight if one has already concluded from other evidence (as Goulder had) that Matthew is using Luke. They have less force if one is not persuaded of that position.

      Generally speaking, I have found that most arguments for Lucan posteriority are less compelling the more closely I examine the text itself. For example, editorial fatigue in the Parable of the Minas is one of the planks of Goodacre’s case, and (as you can see from my analysis), I arrived at the opposite conclusion despite having a sympathetic view of his position beforehand. However, I have many more passages to analyze. (I am manually producing my own colour-coded synopsis of the Gospels as I work through each pericope.)


  4. Great article and discussion. It looks like the prayer developed first from a Pharisaical Kaddish background, short and pithy lines about temptation, as employed in Mark.

    Next we see Matthew following the breadcrumbs Mark left to create a grander prayer. Certainly Matthew reproduces more of Mark overall than Luke does.

    Then Luke abbreviated the prayer.

    While John had little use for it at all since he was so obsessed with Jesus speaking about himself rather than talking about the kingdom of God in parables. Instead, Jesus is constantly speaking parabolically about himself in John.

    There are other possible interpretations as pointed out in the article. Though some sort of trajectory of growth does seem to have occurred.


  5. I think an issue is that one needs to accept certain scholarly presuppositions about the Synoptic problem before any reasoned debate can be had on additional more detailed passages.

    The biggie is Luke. There is a growing scholarly cohort that recognizes the existence of a “proto-Luke” that predates our canonical Luke and which more resembles Marcion’s version. While not yet a majority opinion, it has certainly moved out of fringe territory and I personally think the arguments for it are extremely strong.

    This makes all the more sense since there IS a scholarly consensus that much of the sayings material in Luke predates what is in Matthew, and when put side by side, the more primitive Lukan version is the older. While I disagree with Paul that proto-Luke=Q, this becomes a minor disagreement because Q scholars have long agreed most of Q is found in Luke anyway.

    I’m not sure whether the Lord’s Prayer goes back to the historical Jesus or not; it’s a fairly straightforward Jewish prayer and it’s lack of overt Christianizing (and wide attestation) suggests to me it’s probably similar to something Jesus prayed with his disciples.


    • Thanks for the comment, Andrew. I too am surprised at the way some New Testament scholars will ignore recent research on the Evangelion and its implications for the Synoptic Problem. The implications would certainly be uncomfortable for certain views of the sanctity of scripture, but I couldn’t say how much that influences academic trends.

      Whether the Lord’s Prayer (or any other particular saying) can be traced back to Jesus is a topic I’ve avoided, since strong opinions on such matters can easily cloud the waters. Our inquiry must start with the texts themselves, and the Synoptic Problem is essentially a literary one regardless of the source of the material.

      That said, Goulder’s paper mentioned above gives several arguments against the possibility that the Lord’s Prayer originated with the historical Jesus.


  6. The Lord’s Prayer in Matthew (11.9–13) has been somewhat expanded from the version found in Luke.

    Matthew’s version of the Lord’s Prayer is in 6:9-13. Since Luke’s version is in chapter 11, and one Marcan passage cited above is from chapter 11, I infer that we have a case of Pauline editorial fatigue. 😉


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s