Almost every Christian who says the Lord’s Prayer knows that the New Testament records it twice, once in Matthew and once in Luke, and most have heard that the two versions are not identical. Very few know how much the manuscripts complicate the picture, or how early the complications begin. The prayer as it is printed in English today is one reading among many: a Matthean text with a doxology tacked on the end that is absent from the earliest Greek witnesses, a verse about debts rather than trespasses, and a bread that is daily rather than ordinary. Before the gospels there is a witness older than Luke’s Gospel in some ways and stranger than it: a first-century letter of instruction, the Didache, which quotes a form of the prayer that shares with Matthew a reading which the Greek manuscripts otherwise reserve for a different textual family. And behind all of this lies the ordinary philological fact that a prayer does not stay still: it is copied, shortened, expanded, translated, sung and argued over, and every copy is a decision. What follows is an attempt to look at the Lord’s Prayer as a manuscript phenomenon, tracing it from the Greek gospels through the first-century evidence outside them and into the English versions that most readers will have grown up with, and asking what each stage of the tradition was doing and why.<\/p>\n
The two gospel versions are not two copies of one text. They are two texts, and the differences between them run deeper than a missing verse. Set the Matthean form alongside the Lukan and the shape of the divergence becomes clear. The standard Greek text of Matthew, as printed in the current Nestle-Aland and United Bible Societies editions, runs: Pater hemon ho en tois ouranois, hagiastheto to onoma sou, eltheto he basileia sou, genetheto to thelema sou hos en ourano kai epi ges, ton arton hemon ton epiousion dos hemin semeron, kai aphes hemin ta opheilemata hemon hos kai hemeis aphe(k)amen tois opheiletais hemon, kai me eisenegkes emas eis peirasmon alla rhysai emas apo tou ponhrou.<\/em> The Lukan version has the same petitions, in a shorter order, in fewer words, and without the Matthean expansion of the third petition. The Matthean form occupies eleven verses; the Lukan form, seven.<\/p>\n Three differences in particular have carried most of the scholarly weight. The first is at the end of the third petition, where Matthew has genetheto to thelema sou hos en ourano kai epi tes ges<\/em>, “let your will be done on earth as it is in heaven”, where Luke has simply “as in heaven”. Whether the Matthean addition is an expansion of the shorter reading or the Lukan form a compression of a longer one is precisely the question at issue. The second is in the fifth petition, over the word for bread: ton arton hemon ton epiousion dos hemin semeron<\/em>, “give us this day our daily bread”. The adjective epiousion<\/em>, which the King James Version renders “daily”, is famously difficult: it may mean what is needed for the day, or what is necessary, or simply what comes with the day, and the interpretation adopted colours the whole prayer. The third is the sixth petition, where Matthew has a plural ta opheilemata<\/em>, debts, and Luke has the singular of the same noun, a debt, with a different verb, apheis<\/em> rather than aphes<\/em>, and a different preposition for the people who are owed.<\/p>\n That third difference is the one to watch, because it is the pivot of the whole question. Codex Vaticanus reads at Matthew 6:12 kai aphes hemin ta opheilemata hemon hos kai hemeis aphe(k)amen tois opheiletais hemon<\/em>, with the plural debts and the perfect apekamen<\/em>, “we have forgiven”; the singular, aorist reading is the one preserved in the oldest non-biblical witness, the Didache. Which of the two came first? If the singular is original, then Luke preserves it and Matthew has pluralised it; if the plural is original, then the singular is a Western simplification. The neatness of the problem is precisely what has kept it alive for a century and a half of argument.<\/p>\n On the question of direction, the field has two broad positions. Joachim Jeremias argued for the priority of Luke, on the ground that in the early period, before wordings were fixed, liturgical texts were elaborated, expanded and enriched; on that view the Matthean form is the expanded one, elaborated in the liturgical setting, and Luke preserves the earlier skeleton. Michael Goulder, Thomas Mosbo and Ken Olson have argued the opposite, seeing the shorter Lucan version as a reworking of the Matthean text, removing unnecessary verbiage and repetition. Craig Keener’s survey of the field notes that most scholars accept the prayer in its pre-Synoptic form as authentic, while a minority, Crossan among them, deny the authenticity of the prayer altogether on the ground that a prayer of Jesus should have surfaced in more than two gospels and in a more uniform version. In practice the question has been settled by habit rather than by demonstration: the Matthean form is the one in general Christian use, as Marcus Leaney observed in 1956, and it is the Matthean form that the churches sing and the lectionaries print. Almost nobody in the pews is aware of the Lukan version at all, which is itself a fact worth pausing on, since a prayer transmitted in two forms for two thousand years has spent almost all of that time in one.<\/p>\n One feature of the Matthean Greek is often overlooked, and it bears on the history of the prayer as a spoken text. The first three petitions use the third person passive imperative: let your name be sanctified, let your kingdom come, let your will be done. From the fourth petition onwards the address switches to the second person active imperative: give us, forgive us, lead us not. The change of grammatical mood marks a change of addressee. The prayer begins by speaking about God in the third person and ends by speaking to him directly, and it is in the second half, where the speaker asks for bread, forgiveness and deliverance, that the prayer becomes a description of need rather than a declaration of praise. That structure is common to both gospel versions and to the Didache, and it is the structure that the Latin, Syriac and English translators preserved, whether they knew they were doing so or not.<\/p>\n It is also worth registering how densely scriptural the prayer is, because that density is easy to lose in English. “Let your kingdom come” is not a New Testament coinage; the noun is the noun of the Hebrew Bible’s central political and theological hope, and the third petition of the Lord’s Prayer condenses into three words a theme that runs from the prophets through the psalms into the apocalyptic literature. “Lead us not into temptation” is not a New Testament coinage either, and the noun behind it, peirasmos<\/em>, carries the sense of testing rather than of entrapment: the prayer asks not to be tested, but to be rescued from the one who tests. And the sixth petition’s word for what is forgiven is a word from the world of obligations and accounts. The prayer is a compression of scripture in a way that is unusual even in a gospel, and this is one reason that the question of which form came first is not a question a translator can dodge by preferring one manuscript. Every variant is a choice about how much scripture to let into the mouth of a congregation.<\/p>\n The doxology that closes the prayer in most printed English Bibles and in most liturgical use – “For thine is the kingdom, and the power, and the glory, for ever and ever. Amen” – is not in the oldest Greek manuscripts. That is the plain fact, and it is stated without fuss in the apparatus of the standard editions: most older translations are based on a Byzantine-type text with epi tes ges<\/em> at verse 10 instead of epi ges<\/em>, and with aphiemen<\/em> at verse 12 instead of apekamen<\/em>, and they add at the end of verse 13 the doxology hoti sou estin he basileia kai he dynamis kai he doxa eis tous aionas. amen.<\/em> The doxology is absent from the earliest Greek witnesses of the passage. It is also absent from the Didache, which has a shorter closing formula of its own, and from the Lord’s Prayer as it stands in the oldest Latin and Syriac versions at their core. Its language, however, is not invented: the wording “the kingdom, and the power, and the glory” is drawn from a familiar place in the Hebrew scriptures, the opening of the prayer of David at 1 Chronicles 29:11-13, and it is on that biblical foundation that the early church is generally understood to have built the addition. Bruce Metzger, in his standard commentary on the Greek New Testament, held that the early church added it to the Lord’s Prayer; the reasoning is straightforward. A community that had inherited the formula of David’s prayer, and that recited the Lord’s Prayer daily, would naturally have ended it with the words that ended David’s.<\/p>\n The Byzantine scribes who transmitted the prayer with the doxology were not making an error; they were making a decision, and the decision was liturgical. To sing the prayer in a congregation is to say it with the assembly, and the assembly wants a conclusion. A prayer that ends with “deliver us from evil” ends in the wrong place for a doxology, and the solution adopted was to append one. The result is a text that is later than the gospels and older than most of its readers, and it is worth holding both halves of that statement together. The doxology is not in the manuscripts; the doxology is in the liturgy; and the liturgy has a way of becoming the manuscript.<\/p>\n At the same time, the Greek manuscripts preserve the shorter, or “Western”, reading of the sixth petition, and this is the point at which the manuscript evidence and the patristic evidence converge most strikingly. The table below sets out the three principal forms side by side – the liturgical Greek text, the text of Codex Vaticanus, and the text of the Didache – and the differences are not merely orthographic. At the third petition, the Didache reads “and on the earth”, with the article, where the liturgical text has “and on the earth” without it and the Vaticanus has neither article nor the preposition compounded. At the sixth petition, the Didache agrees with the Vaticanus in reading a single debt rather than debts, and the aorist of forgiveness rather than the perfect. At the seventh petition all three are identical: “and lead us not into temptation, but deliver us from the evil one”.<\/p>\n That last clause deserves a note, because it is a case where the Greek is ambiguous and the English translators have had to choose. The word rendered “the evil one” is the definite article followed by a noun meaning “evil” or “wickedness”; the same phrase can be read as “the evil” in the abstract. The manuscripts and the versions divide. The Didache has “from the evil one”. The Prayer Book of 1662 and most modern worship books have “from evil” without the article, which nudges the reader towards the moral meaning; the more literal reading, with an article, suggests a personal enemy, and some English versions have followed it. The difference is not trivial in theology, and it has not gone unnoticed by the churches, which is why so many modern printed forms of the prayer manage to print both: the current lectionary text reads “Forgive us our sins, for we also forgive everyone who sins against us”, with a note explaining that the Greek at that point actually says “everyone who is indebted to us”, and adds the clause “but deliver us from evil” in the same way, noting the alternative “from the evil one”. A footnote is the modern analogue of the doxology: a way of carrying two readings at once.<\/p>\n It is worth pausing over what the Western reading at verse 12 would mean if it were original. “Forgive us our debt, as we also forgive our debtors” is a sharper sentence than “forgive us our debts”. It is singular because a debt is what one person owes another, and the prayer’s logic is exact reciprocity: the one thing asked for is the one thing asked in return. The plural form allows a looser sense, debts in general, and it fits more comfortably into a liturgy. The manuscripts, in other words, preserve two stages of a development that is also a moral argument: a prayer about a specific relationship, becoming a prayer about a general condition.<\/p>\n It is worth being concrete about the mechanism, because nothing about it requires the sort of doctrinal deliberation that later readers have imagined. A scribe copying a codex in the fourth or fifth century comes to the end of the prayer and finds that the line runs straight on into the next section of the gospel. In such a setting, a reader who knows the psalm of David and the habit of ending prayers with praise has every reason to supply the missing words, and a community that has been reciting the prayer daily for a century has every reason to have settled on one form of the conclusion. The supplied words then travel, because they are written down, and the next copyist copies them because they stand in his exemplar. That is all. No council decided anything; no committee added a doxology. A congregation’s need for a conclusion became a line on a page, and the line became the tradition. It is the ordinary history of a text, and it is worth remembering that the same mechanism, working on a shorter scale, could produce the singular reading at verse 12. A copyist who has heard the prayer said with one debt might write one debt; a copyist who has heard it said with debts might write debts. The variations are not errors in the modern sense. They are the fossil record of a prayer being used.<\/p>\n The Didache, or the Teaching of the Twelve Apostles, is one of the strangest documents of the early church. It is usually dated to the first century, or at any rate to within a generation or two of the writing of Matthew’s Gospel, and it is generally thought to come from a Syrian or Palestinian setting. It contains two teachings, the Way of Life and the Way of Death, instructions on baptism and fasting and prayer, and a set of rules about the eucharist. It has no narrative and no narrative about Jesus, and its silence about the life of Christ is part of what makes it so instructive for the early Christian period. And it quotes a form of the Lord’s Prayer that is not identical with the Matthean text and not identical with the Lukan text either.<\/p>\n One feature of the Didache is so striking that it is easy to overlook. The document records no saying of Jesus. Not one. There is no parable, no logion, no teaching of the Sermon on the Mount, and no narrative of any kind. And yet the prayer is there, quoted and quoted in a form that differs from the gospels, and it is presented as an instruction to be handed on: teach this to the catechumens, and pray it three times a day. The implication is considerable. By the time the Didache was written, the prayer was evidently being treated as authoritative material in its own right, detached from the teaching of Jesus and transmitted as a fixed liturgical form with its own authority and its own slightly varying wording. The Didache is thus evidence not of the origin of the prayer but of its early status: by the first century or the second, a short sequence of sentences had come loose from the mouth of a teacher and become a thing in its own right, carried from one community to another in a form that could be quoted, compared and criticised.<\/p>\n Here is the passage in the translation of J. B. Lightfoot, which follows the division of the text into chapters that the English reader will recognise: “Give us this day our daily bread; and forgive us our debt, as we also forgive our debtors; and lead us not into temptation, but deliver us from the evil one; for Thine is the power and the glory for ever and ever.” The same text appears in the Greek and in the older patristic translations. Two details about this quotation are worth noticing before anything else. The first is that the prayer ends with a doxology, but not the Byzantine one: it has the power and the glory, without the kingdom. The second is a matter of numbering, and it is the sort of detail that reminds us how much of what we call the manuscript evidence is in fact editorial. In Lightfoot’s text, and in many older editions, this prayer sits in chapter eight; in the common modern numbering it sits in chapter nine, with the eucharistic prayers that follow in chapter ten. A chapter and a verse number, in other words, is itself a piece of editorial history, and anyone citing “the Lord’s Prayer at Didache 9:2” is citing a modern editor’s arrangement as much as an ancient manuscript.<\/p>\n The interest of the Didache for the history of the prayer lies in two features. The first is that its text is closest to Matthew’s of the two gospel forms, which suggests a relationship with the Matthean tradition, whether direct or mediated. The second is that at the one point where the Greek manuscripts divide most sharply, the Didache sides with the reading that scholars call Western, reading a single debt and using the aorist of forgiveness. If the Didache were simply a copy of Matthew, that agreement would be surprising; if it is an independent witness to a form of the prayer that circulated alongside the Matthean one in the first century, it is exactly what we would expect. This is the foundation of the argument, developed most fully by C. Clifton Black, that the Didache preserves the earliest trace we can document of an additional ending or doxology, and that the prayer existed in more than one form in the first century, in a state of flux rather than a state of fixity.<\/p>\n The context in which the Didache gives the prayer is worth a moment’s attention, because it shows the prayer doing liturgical work rather than merely being preserved. The instruction immediately before it is that the leaders are to teach the catechumens these things, so that they may be careful to keep the teaching of Christ. Then comes the prayer, and then the instruction that “three times in the day pray ye so”. Then, at chapter nine, come the eucharistic thanksgivings, and they are built out of the same building blocks: “We give Thee thanks, O our Father, for the holy vine of David”, and a blessing over the cup, and then a blessing over the bread, and then the instruction that no one may eat or drink of these until they have been reconciled with their brothers, and that if it should happen that some are prevented from assembling, they are to remember that the Lord will come and receive them, and that the Lord “who spread out his hands for you and gave you food and drink, has given you life”. The eucharistic prayers are the Lord’s Prayer taken apart and rebuilt. A community that prayed the Pater three times a day and then broke bread with prayers that echo it was living inside a single small piece of text, using it as liturgy rather than as literature.<\/p>\n Two pieces of external evidence show how widely the Didache travelled. Epiphanius of Salamis, writing in the fourth century, quotes the Didache at length in his Panarion, where he deploys the eucharistic prayers against the Arians: the text was still, by then, a weapon in an argument about the nature of the bread and the cup. And Tertullian appears to allude to the Didache’s regulations on fasting in his treatise on fasting, describing the practice of those who fast on the fourth and fifth days of the week, which is the Didache’s practice. These are small indications, but they are the kind of evidence that turns a document from a curiosity into a witness with a history.<\/p>\n And yet the question of dependence remains open, and it is worth stating the alternatives honestly rather than choosing one by atmosphere. The Didache’s prayer may be dependent on Matthew, reproducing a form it found in the Matthean gospel text. It may be independent, preserving a tradition of Jesus’s teaching that Matthew also preserves, in a different recension. Or it may be a liturgical form, a prayer for use that developed alongside the gospel text and was itself a source for it. Each possibility has been argued, and the manuscript evidence does not settle the question; it establishes only that the prayer was circulating in more than one form, in more than one setting, and in more than one language, in the first century and the second.<\/p>\n The English history of the prayer is a history of two words, and the arguments they have generated are as old as the Reformation. The first is the word for bread. Matthew’s epiousion<\/em> was rendered “daily” in the Authorised Version and has been rendered “daily” by most English translators since, which fixes the petition in a daily rhythm and quietly resolves a genuine ambiguity in the Greek in favour of time. The second word is the one that matters most for the character of the prayer: the word for what is forgiven. The Greek at the sixth petition gives debts. Tyndale rendered it as debts in 1526, and the Authorised Version of 1611 followed him: “And forgive us our debts, as we also forgive our debtors.” The Prayer Book went the other way. The 1549 Book of Common Prayer, as printed, puts the petition into the mouths of English Christians in this form: “OUR father, whiche arte in heaven, hallowed by thy name. Thy kyngdom come. Thy wyll be done in earth as it is in heaven. Geve us this daye oure dayly bread. And forgeve us oure trespasses, as we forgeve them that trespasse agaynst us. And leade us not into temptacion. But deliver us from evill. Amen.” Note what has changed and what has not. The bread is still daily, and the petition is still a petition for forgiveness and for deliverance from evil; but the object of forgiveness is now trespasses, and the people from whom forgiveness is asked are those who trespass. The word is a psalm word. Trespass belongs to the vocabulary of the Hebrew psalms, of David, of the confessions that run through the Psalter, and by choosing it the compilers of the Prayer Book made the Lord’s Prayer a psalm in the mouth: a confession of transgression rather than a note of indebtedness. The King James Version’s “debts” belongs to a commercial vocabulary, and it makes the prayer a transaction between two parties who owe one another money. The difference is not ornamental. One tradition makes the prayer an act of penitence, the other an act of accounting.<\/p>\n The Geneva Bible of 1560, which carried the Prayer Book’s reforming sympathies into a translation with its own preface and its own annotations, shows how thoroughly the two habits were mixed. Its Matthew 6 reads: “Give us this day our daily bread. And forgive us our debts, as we also forgive our debtors. And lead us not into temptation, but deliver us from evil: for thine is the kingdom, and the power, and the glory forever, Amen.” The debts of Tyndale are kept in the sixth petition. But two verses later the same book has: “For if ye do forgive men their trespasses, your heavenly Father will also forgive you.” The word has migrated by four verses. And the Geneva also carries a doxology at the end of the prayer, in the shorter form without “the power” – a small reminder that the closing clause was not fixed even within a single translation, and that translators were making the same choices about it that the Greek copyists had made.<\/p>\n By the time the Prayer Book was revised in 1662, the two features that would define the English tradition for three centuries were both in place: the trespasses, and a doxology that expanded the Matthean petition into a doxological conclusion – “And lead us not into temptation, But deliver us from evil; For thine is the kingdom, the power, and the glory, For ever and ever. Amen.” The BCP form is the ancestor of the “trespasses” line in modern Anglican worship and, through the English Prayer Book, of much of the wider tradition; and it is the ancestor, indirectly, of the modern habit of printing two alternatives in a bracket, since the twentieth-century ecumenical movement produced liturgical forms that tried to hold the Prayer Book’s “trespasses” and the Authorised Version’s “debts” and the modern “sins” in a single page, and could only do it by printing notes.<\/p>\n It is worth noticing what the modern bracket form actually does. The contemporary printed prayer reads, in effect: “Forgive us our sins, for we also forgive everyone who sins against us”, with a note that the Greek says “everyone who is indebted to us”; and “lead us not into temptation, but deliver us from evil”, with a note that the Greek may mean “from the evil one”. The note has become part of the text. The reader who says the prayer in church is now, whether they know it or not, saying a form that reports its own textual uncertainty, and the parenthesis in the worship book has quietly taken the place that the apparatus criticus occupied in the scholar’s hand. That is the end of the story of this particular line of transmission. A prayer that entered the manuscripts of Matthew and Luke in the first century was expanded in the fourth, preserved in the earliest non-biblical witness, translated into Latin and Syriac, debated by sixteenth-century English printers, fixed in the Prayer Book, and is now printed with its own footnotes in a parish church. It has never once stopped being a manuscript problem, and it has never once stopped being a community’s prayer.<\/p>\n There is a third variant in the story of the English prayer that is usually mentioned only as a curiosity, and it is worth mentioning because it shows how much latitude a translator has always had with the petition’s central noun. A rendering with “dough” for bread – “our daily dough” – has a place in the history of English versions, and the word made this possible because in older English it could mean bread, or food in general, rather than the unbaked paste that it now denotes. It should be said plainly that the Prayer Book as printed in 1549 reads “our dayly bread”, and that the Geneva Bible of 1560 likewise reads “our daily bread”; the rustic rendering belongs to the margins of the tradition rather than to its settled text. But its existence is instructive. The noun in the Greek is artos<\/em>, ordinary bread, and the adjective epiousion<\/em> is a puzzle. Any translator who took the puzzle seriously might try to make the petition mean more than bread, and “dough”, with its sense of food in general, is exactly the kind of solution a sixteenth-century reader with a taste for literalism might reach for.<\/p>\n What all of this makes visible, in the end, is that the Lord’s Prayer is one of the best-documented cases we have of a text that never stopped being a text. The gospels give two forms. The Didache, probably of the first century, gives a third that shares the oldest reading at the decisive point and adds a doxology of its own. The Greek manuscripts add the Byzantine ending and preserve the shorter Western reading of the sixth petition. The Latin and Syriac translators, the copyists, the commentators, the liturgists, Tyndale and the Genevans and Cranmer and the King James translators and the twentieth-century lectionary committees all made decisions, and each decision is visible in the shape of the prayer as it is now spoken. To say that the prayer “is” one thing or another is to decide, silently, a question that the evidence answers only by showing that there was never a single answer to give. That is not a defect of the tradition. It is what a living text looks like from the inside.<\/p>\n Almost every Christian who says the Lord’s Prayer knows that the New Testament records it twice, once in Matthew and once in Luke, and most have heard that the two versions are…<\/p>\n","protected":false},"author":1,"featured_media":855,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[9],"tags":[784,785,312,783,341,749],"class_list":["post-856","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-text-transmission","tag-didache","tag-english-bible","tag-gospels","tag-lords-prayer","tag-new-testament","tag-textual-variants"],"_links":{"self":[{"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/posts\/856","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/comments?post=856"}],"version-history":[{"count":1,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/posts\/856\/revisions"}],"predecessor-version":[{"id":857,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/posts\/856\/revisions\/857"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/media\/855"}],"wp:attachment":[{"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/media?parent=856"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/categories?post=856"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/masoretes.online\/wordpress\/wp-json\/wp\/v2\/tags?post=856"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}The Longer Ending and the Shorter Reading<\/h2>\n
Before the Gospels: The Didache<\/h2>\n
Dough, Debts and Trespasses: The Prayer in English<\/h2>\n
References<\/h2>\n
\n