Luke 22:17-20 describe the events of the Last Supper. However, there are many variants of these verses in Greek and other languages (e.g. mss without some of all of Luke 22:19b-20), and the most common variant is very different to the parallels in Mark and Matthew. Why do these variants exist, and which one was the original form of these verses? This page discusses these issues on the assumption of Markan priority, i.e. that the gospel of Mark was written before either Matthew or Luke, and that aLuke (the author of Luke) had access to Mark (See The Synoptic Problem for more details of this issue). English text is taken from the King James Version (KJV) unless specified otherwise, not for any doctrinal reason, but purely because it is the one with which this author is most familiar.
The Merriam-Webster dictionary has this definition of Jesus' Words of Institution at the Last Supper:
The portion of a Christian Communion service based on the words of Mk 14:22–24 and used as the warrant from Jesus Christ for the continued celebration of the Eucharist.
According to Wikipedia, the Words of Institution:
… are words echoing those of Jesus himself at his Last Supper that, when consecrating bread and wine, Christian Eucharistic liturgies include in a narrative of that event…
No formula of Words of Institution in any liturgy is claimed to be an exact reproduction of words that Jesus used, presumably in the Aramaic language, at his Last Supper. The formulas generally combine words from the Gospels of Mark, Matthew and Luke and the Pauline account in First Corinthians (1 Cor) 11:24-25.
The main places in which the Words of Institution are presented in the synoptic gospels (Mark, Matthew, Luke) are in the parallel passages describing the events of the last supper, at Mk 14:22-25, Mt 26:26-29, and Lk 22:17-20 respectively. The passages in Mark and Matthew are virtually the same, and both refer to the bread and the cup (of wine) in that order. In contrast, the parallel in the majority of manuscripts (mss) of Luke has some significant differences, chief of which are an additional reference to the cup before the bread, and text in Lk 22:19b-20a that has no parallel in either Mark or Matthew.
In the following discussion synoptic implications (mainly regarding the order in which Mark, Matthew and Luke were written) are generally left until after the various possible sources have been discussed, with the exception that Markan priority (that Mark preceded both Matthew and Luke) is assumed. In particular, no assumption is made regarding the order in which Matthew and Luke were written, nor whether the author of one knew the other. Additionally, for reasons given below this analysis allows the possibility that Marcion’s Gospel of the Lord (The Evangelion, or Ev) preceded both Matthew and Luke. In places a ‘Mark/Matthew’ version of the text is referred to, and in these instances it is because in these instances the text in Matthew is the same (or essentially the same) as the text in Mark. It is assumed that aLuke (the author of Luke) knew the version in Mark, but may or may not have known the version in Matthew. Because some of the key points of discussion involve textual variants (particularly in Luke), different paths via which the variants could have been created are considered, with variants possibly resulting from the use of different source materials.
In the majority of mss of Luke (mainly Greek, but also some in other languages) the passage relating to the Words of Institution reads as follows in English, with the verses here shown sub-divided in a way that highlights various issues discussed later:
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [22:17]
For I say unto you, I will not drink of the fruit of the vine, [22:18a]
until the kingdom of God shall come. [22:18b]
And he took bread, and gave thanks, and brake it, and gave unto them, saying, This is my body [22:19a]
which is given for you: this do in remembrance of me. [22:19b]
Likewise also the cup after supper, saying, [22:20a]
This cup is the new testament in my blood, [22:20b]
which is shed for you. [22:20c]
This ‘Majority Text,’ as found in manuscripts and versions P75 A B C L Tvid W Q f1 f13 (33 is defective here) 565 579 700 892 1241 aur c f q r1 vg hark pal so bo arm geo eth slav, contains elements seen in the Markan parallel at Mk 14:22-25, although not in the same order. However, it also contains text not seen in Mark, and instead is very similar to that in 1 Cor 11:23b-25. On first impression the Majority Text appears to be a hybrid of these two sources, although with various significant differences from both, which read as follows:
Mk 14:22-25
And as they did eat, [14:22a]
[Jesus][He] took bread, and blessed, and brake it, and gave to them, and said, Take[[, eat:]] this is my body. [14:22b]
And he took the cup, and when he had given thanks, he gave it to them: and they all drank of it. [14:23]
And he said unto them, This is my blood of the [[new]] testament, [14:24a]
which is shed for many. [14:24b]
Verily I say unto you, I will drink no more of the fruit of the vine, [14:25a]
until that day that I drink it new in the kingdom of God. [14:25b]
1 Cor 11:23b-25
… the Lord Jesus the same night in which he was betrayed took bread: [11:23b]
And when he had given thanks, he brake it, and said, [[Take, eat:]] this is my body, [11:24a]
which is [[broken]] for you: this do in remembrance of me. [11:24b]
After the same manner also he took the cup, when he had supped, saying, [11:25a]
this cup is the new testament in my blood; [11:25b]
this do ye, as oft as ye drink it, in remembrance of me. [11:25c]
The square brackets enclosing words shown above identify variants in these verses, with double square brackets marking those that are not considered to be original, as discussed below. The significance of these variants is that if they were original, and the Majority Text is indeed derived from these parallels, then we might expect to see them in Luke, but if they are not original, then we might not.
Take, eat
The mss evidence indicates that neither Mk 14:22b nor 1 Cor 11:24a originally included the word “eat,” with 1 Cor 11:24a also omitting “take.” Aland & Aland write:
The textual variants in 1 Cor 11:24 bear on only a detail of Paul’s account of the Last Supper. Here the Majority text together with the traditional versions add at the beginning the words of institution labete phagete. These words are taken from Matt. 26:26, and they represent nothing more than parallel assimilation, clearly opposed by the evidence of the textual tradition.
Similarly, the textual evidence is heavily in favor of ‘phagete’ (eat) in Mk 14:22b also being an assimilation to Matthew, and so not present in the original text of Mark.
The ‘New’ Testament
In the KJV Mk 14:24a, Mt 26:28a, Lk 22:20b, and 1 Cor 11:25b all refer to the “new testament.” However, “new” is not found in the early mss of both Mark and Matthew, and the NET has this comment on Mk 14:24:
Most mss (A Ë1,13 Ï lat sy) have καινῆς (kainh", “new”) before διαθήκης (diaqhkh", “covenant”), a reading that is almost surely influenced by the parallel passage in Luke 22:20. Further, the construction τὸ τῆς καινῆς διαθήκης (to th" kainh" diaqhkh"), in which the resumptive article τό (referring back to τὸ αἷμα [to |aima, “the blood”]) is immediately followed by the genitive article, is nowhere else used in Mark except for constructions involving a genitive of relationship (cf. Mark 2:14; 3:17, 18; 16:1). Thus, on both transcriptional and intrinsic grounds, this reading looks to be a later addition (which may have derived from τὸ τῆς διαθήκης of D* W 2427). The most reliable mss, along with several others (א B C Dc L Θ Ψ 565), lack καινῆς. This reading is strongly preferred.
The NET also comments on Mt 26:28:
Although most witnesses read καινῆς (kainhs, “new”) here, this is evidently motivated by the parallel in Luke 22:20. Apart from the possibility of homoioteleuton, there is no good reason for the shorter reading to have arisen later on. But since it is found in such good and diverse witnesses (e.g., Ì37,45vid א B L Z Θ 0298vid 33 pc mae), the likelihood of homoioteleuton becomes rather remote.
According to the NET, not only is "new" not original in either Mark or Matthew, but both are interpolations resulting from the inclusion of "new" in Lk 22:20. It should also be noted that, as seen in the above comment on Mk 14:24, “testament” in the KJV is translated in other bibles (and in other places in the KJV) as “covenant.” For example, we see “the blood of the covenant” in Heb 9:20a, and in this instance it is possible that Mark, Matthew, and Hebrews here all ultimately derive from Ex 24:8:
And Moses took the blood, and sprinkled it on the people, and said, Behold the blood of the covenant, which the LORD hath made with you concerning all these words.
The Atonement Doctrine: Given and Broken For You
In 1 Cor 11:24 the KJV has: “This is my body which is broken for you,” as is read in the majority of (mainly later) Greek mss, three old Latin mss, the Syriac Peshitta, and several other versions. However, p46 S* A B C* 33 1739* all omit “broken,” reading instead: “This is my body which is for you." Most of the old Latins, vg, and cop have: “This is my body which is given for you,” and D* has: “This is my body which is shattered for you.” The mss evidence is strongly in favor of the shortest reading in 1 Cor 11:24 being original, with “given” being added (possibly as an assimilation to Lk 22:19) to avoid the problem of not having a verb, while “broken” and “shattered” may refer back to the bread having been broken.
Because of the variant readings just noted, in the discussion below the words shown above in double square brackets will be omitted, and “testament” will be replaced by “covenant.” However, there are other variants where it is far from clear which reading is original, in particular those in Lk 22:19-20.
Some of the Majority Text variants of these verses (e.g. in Lk 22:18) have a close parallel in Mark but not in 1 Corinthians 11, while some (e.g. in Lk 22:19b) have a close parallel in 1 Corinthians 11 but not in Mark. However, perhaps the most significant of the differences between the Majority Text and the parallels is the order of the references to the cup and the bread: Mark and 1 Corinthians 11 have bread-cup, while in Luke the Majority Text has cup-bread-cup.
Why and how these differences arose has been the subject of much discussion, with no really satisfactory solution being proposed. However, the Majority Text of Luke given above is not the only known variant of these verses, and there are a number of mss (mainly Old Latin and Syriac) that depart from it. In comparison with the Majority Text, all omit text and/or have a different verse order:
D (the only known Greek ms in which this is the case), and the Old Latin mss a, b, d, e, ff2, i all omit Lk 22:19b-20. Apart from b and e, these mss read only Lk 22:17, 18, 19a.
In mss b and e Lk 22:19a is swapped with Lk 22:17-18, so reading Lk 22:19a, 17, 18.
There are other variants in the Syriac and in a small number of other mss:
The Curetonian Syriac (Sy-C) reads Lk 22:19, 17, 18, except that “Given” is omitted from Lk 22:19b, so reading the same as 1 Cor 11.24b here.
The Sinaitic Syriac (Sy-S) is similar to Sy-C, but with additional text either side of Lk 22:17. (Some authorities state that Sy-C reads as Lk 22:19, 20a, 17, 20b, 18, but there is no mention of “the cup” in the parallels of Lk 22:20a and 20b.)
The Peshitta (Sy-P), L32, two Sahidic mss, and one Bohairic ms, read Lk 22:19-20.
It is generally considered that one of the above variants must represent the original text of Luke, and that the others are all derived from it. There are two predominant views: either Luke originally contained the long form as seen in the Majority Text, and this was then edited to create all the shorter forms; or it originally contained the short form seen in the Old Latin, which was then expanded to create all the longer forms. Various arguments have been advanced in support of both views, for example:
In favor of the originality of the longer text:
(a) The vast majority of the extant mss (including all known Greek mss except D) support it;
(b) It is easier to suppose that someone seeing the cup-bread-cup sequence would eliminate one of the cups, than that someone would ‘correct’ the cup-bread sequence by adding the second cup from 1 Cor 11, without at the same time removing the first one.
(c) One or more copies of Luke had text removed in order to protect the Eucharist (disciplina arcana).
In favor of the originality of the shorter text:
(a) The shorter reading is generally preferred (Lectio brevio);
(b) The more difficult reading is generally preferred (Lectio difficilior potior);
(c) Lk 22:19b-20 contain a number of non-Lukan linguistic features;
(d) Lk 22:19b-20 are very similar to 1 Cor 11:24b-25, suggesting that the latter is the source of the additional words in the longer text.
Although both of these positions have their supporters, both also have their problems, and neither is conclusive. Chadwick writes:
The Lucan account with its notorious difficulties of text and interpretation, the one being inextricably bound up with the other, has been the Waterloo of many investigators. Both the shorter and the longer texts present the student with their individual mass of problems. Since Hort the arguments against the originality of the longer text have had so wide an influence, especially perhaps on English speaking writers, that if these thorny questions could be settled by taking a vote, it is possible that even now the shorter text would have it. Unfortunately the weight of argument is rather more evenly divided.
Given that both expanding the shortest form and shortening the longest form of this Lukan text are problematic, it is perhaps surprising that suggesting that one of the ’intermediate’ forms may have been the starting point, and that it was shortened by some and lengthened by others, appears to have not been seriously considered. For this we need to look in more detail at the possible sources of the text, but before doing so, there are four additional sources to be taken into consideration: the first because it may reflect an early version of this passage; the second because it sheds some light on possible changing practices relating to the Last Supper; and the third and fourth because they almost certainly have influenced the text of the Syriac variants.
Most analyses of this issue ignore what may have been in the ‘Gospel of the Lord’ (Ev), which around 140 A.D. was championed by Marcion as the true gospel, and is essentially a shorter version of what we see as Luke, for the simple reason that the majority view is that Luke preceded Ev. However, the great majority (232) of the 297 verses of Luke that are not present in Ev are unique to Luke, as shown in The Synoptic Gospels and Marcion, which concludes:
Overall, 78% of the verses in Luke that are not in Ev are unique to Luke. Even when Luke 1 and 2 are excluded (as per Mark) nearly 61% of the Lukan verses not in Ev are unique to Luke. It should be clear from this alone that if Ev is based on Luke then aEv [the author of Ev] appears to have gone to a great deal of trouble to exclude mainly verses that aLuke had previously added. Of course, this also implies that aEv was able to identify those verses unique to Luke from his knowledge of both Mark and Matthew, because without that knowledge it would appear almost impossible for him to have achieved this observed result.
Unfortunately, there are no extant mss of Ev, and so we have to rely on reconstructions created by using information from people who wrote about him. Chief of these is Tertullian, who around 207-208 A.D. wrote Adv. Marcion, Book IV, in which he compares the text of this gospel with that of Luke, and notes or comments on differences between the two.
Even where Ev is mentioned in analyses of Lk 22:17-20, it is generally reported that here it followed the Majority Text. However, a detailed examination of Tertullian’s comments on these verses (Epiphanius does not mention them) paints a different picture. In the second half of Adv. Marcion IV, chapter 40, Tertullian mentions (in the order given below) various elements of these verses, without giving any indication that the text he saw in Ev differed from that in his copy of Luke. He states (translated from the Latin original) that “He so earnestly expressed His desire to eat the Passover” (Lk 22:15), and then writes:
Then, having taken the bread and given it to His disciples, He made it His own body, by saying, "This is my body," that is, the figure of my body…
He likewise, when mentioning the cup and making the covenant to be sealed "in His blood," affirms the reality of His body.
He then refers to wine several times, for example:
In order, however, that you may discover how anciently wine is used as a figure for blood, turn to Isaiah…
… Thus did He now consecrate His blood in wine, who then (by the patriarch) used the figure of wine to describe His blood.
At first sight Tertullian appears to refer to all the important elements that we see in the Majority Text variant. He refers to bread, the cup, and wine, quotes “This is my body” (Lk 22:19a), and mentions the covenant to be sealed in His blood (Lk 22:20b), and it is these references that have led to the view that Ev contained all of Lk 22:17-20, i.e. as in the Majority Text. However, it is important to note the order in which Tertullian mentions these elements and also what he does not say, and a detailed examination shows that he departs from the Majority Text in several ways. For example:
He has the order: bread/body (Lk 22:19a); cup (either Lk 22:17 or 20a); blood (Lk 22:20b); and wine (Lk 22:18); instead of cup – wine – bread/body – cup - blood.
In his only reference to “the cup,” he does not mention it as being “after supper.” This singular, undifferentiated, reference to the cup suggests that he saw only the first mention of it, in Lk 22:17.
He does not mention any of Jesus’ atoning words in Lk 22:19b and 20c (“This is … for you”),
He does not mention any of Lk 22:20a, and in Lk 22:20b refers to “the covenant,” rather than “the new covenant.”
Although Tertullian is of course not required to mention every part of every verse of Ev, nor to maintain the same text order in his discussion, where there are apparent differences we should not just assume that they are insignificant without further examination. As Lietzmann & Richardson: remark:
… in any inquiry into the genuineness of a disputed text, and especially in the case of the earliest witnesses, the evidence cited on its behalf cannot be assumed to testify to more than the actual words quoted.
Lietzmann & Richardson also make various points regarding Tertullian’s account:
Tertullian’s only definite quote is that Christ “made it his own body by saying ‘This is my body.’”
Tertullian uses the word ‘distributum’ regarding the bread instead of (in Greek) “eklasen kai edōken” (broke it and gave it), leaving doubt as to whether he saw “eklasen,” or whether it was just “an example of Tertullian’s free method of quotation.”
Tertullian does not mention that the bread is “given for you.”
From Tertullian's words: “Sic et in calicis mentione testamentum constituens sanguine suo obsignatum” it is likely that he saw Mark’s parallel to Lk 22:20.
“It is moreover unlikely that Tertullian, who has so much to say of the old and new covenants, himself omitted the ‘kainē’ [new] inadvertently.”
From these points they “conclude that the maximum reconstruction of the text of Ev that can be reached with any degree of probability is:”
Having taken the bread and … it to his disciples, (he said) [c.f. Lk 22:19a]
This is my body [? Which is given for you]. [c.f. Lk 22:19b]
And in like manner the cup, (saying) [c.f. Lk 22:17]
This cup is the covenant in my blood [or This is my blood of the covenant]. [c.f. Lk 22:20b]
(We cannot infer any mention of “Do this in remembrance of me.) [c.f. Lk 22:19c]
As mentioned in the discussion of the version of First Corinthians attributed to Marcion, Tertullian also does not refer to remembrance in either 1 Cor 11:24b or 11:25c: "this do in remembrance of me." and "this do ye, as oft as ye drink it, in remembrance of me," instead stating only:
In like manner, when treating of the gospel, we have proved from the sacrament of the bread and the cup the verity of the Lord's body and blood in opposition to Marcion's phantom...
However, Tertullian’s later mention of wine being “used as a figure for blood” would be very unusual if he had not seen wine mentioned in Mcg. Therefore, the above all suggests that Tertullian did not see Lk 22:19c, 20a, and 20c in Ev gospel, and that what he did see read in the order Lk 22:19a (possibly omitting the second half), 17, 20b, 18. Not only do Tertullian’s words not support the view that Ev had what we see in the Majority Text, but because of the reference to both blood and the covenant, neither can they support either of the shortest variants as seen in the Old Latin. Therefore, we need to determine whether what Tertullian actually reports (both as to wording and order) possibly:
Matches any of the other known Lukan variants of these verses;
Matches any of the potential sources or parallels of the verses; or
Could represent an otherwise unknown stage in the development of the text.
The Didache
The Didache (or Teaching of the Twelve Apostles) was first discovered in 1883, as part of a larger Greek codex. Fragments have since been found at Oxyrhyncus: P. Oxy 1782 from the 4th century; and P. Lond. Or. 9271 (Coptic) from 3/4th century. In addition, many 2nd and 3rd century references to it exist, especially in Syria and Egypt. Although it is undated, the history of the text, and the primitive Church that it suggests (for example referring in Ch 11 to apostles and prophets), indicates that it was likely written before the middle of the 2nd century, and possibly even some time in the 1st. The importance of the Didache for this discussion is that in Ch 9 it describes the Eucharist, and in particular does so in the order cup-bread, with no mention of the atoning sacrifice of Jesus that we find in the synoptic gospels:
You shall give thanks as follows: First, with respect to the cup: “We give you thanks, our Father, for the holy vine of David, your child, which you made known to us through Jesus your child. To you be the glory forever.” And with respect to the fragments of bread: “We give you thanks our Father, for the life and knowledge that you made known to us through Jesus your child. To you be the glory forever”
Mack notes:
The prayer of thanksgiving (eucharist) for the community meal in chapters 9 and 10 are also significant. That is because they do not contain any reference to the death of Jesus. Accustomed as we are to the memorial supper of the Christ cult and the stories of the last supper in the synoptic gospels, it has been very difficult to imagine early Christians taking meals together for any reason other than to celebrate the death of Jesus according to the Christ myth. But here in the Didache a very formalistic set of prayers is assigned to the cup and the breaking of bread without the slightest association with the death and resurrection of Jesus.
McGowan points out that there are other very early Christian examples of the cup-bread order:
A different ritual pattern is also attested, and by implication a more definite contrast with the Greco-Roman symposium as usually celebrated. The Eucharistic meal of the Didache, the earliest surviving account of any “Eucharist” and prayers associated with it, famously prescribes a cup-bread pattern of blessing and consumption instead of the expected synoptic meal-drink pattern.
The Didache meal has sometimes been interpreted as an “Agape” rather than Eucharist, because of the difference from the later normative order of the sacramental meal and the lack of reference to the Last Supper or death of Jesus. That logic, however, is circular. As we have seen, the cup-bread order differs as much from typical Greco-Roman expectations for communal meals as it does from the “institution narratives” in particular.
There are other early Christian examples of the cup-bread pattern. Although the difference between this order and the Greco-Roman symposium in general (and the Last Supper in particular) is curious, there are notable parallels for the cup-bread pattern with Jewish meals. The Dead Sea Scrolls and later Rabbinic literature both witness to the opening of formal meals with the blessing of a cup. Although opening toasts or libations were not unknown in other Greco-Roman settings, these Jewish and Christian meals also seem to omit altogether the typical mixing and drinking of wine after the food.
Although the exact relationship of the meals described in the Dead Sea Scrolls and the Eucharist is not clear, they do appear to be related, as Shanks attests:
But however the meal of the Qumran convenanters is interpreted, its messianic character, the prominence of bread and wine, the fact that it was repeated regularly, and the explicit eschatological associations do in fact remind one of elements found in the New Testament words about the Lord's Last Supper.
Whatever the exact relationship between the Dead Sea Scrolls and the Eucharist, the Didache does seen to be evidence of an early Christian meal ritual that not only placed the cup before the bread, but also had no mention of the Atonement Doctrine seen in the synoptic gospels and 1 Corinthians 11. Crossan goes even further, seeing in the Didache evidence that such ritual elements did not appear until later:
… even late in the first century C.E., at least some (southern?) Syrian Christians could celebrate a Eucharist of bread and wine with absolutely no hint of Passover meal, Last Supper or passion symbolism built into its origins or development. I cannot believe that they knew about those elements and studiously avoided them. I can only presume that they were not there for everyone from the beginning, that is, from solemn formal and final institution by Jesus himself.
The Diatessaron is a harmony of Matthew, Mark, Luke and John, written around 160-174 A.D by Tatian, an Assyrian. The harmony merges the text of the synoptic gospels, (mostly) eliminates duplications, and in order appears closest to Matthew. Neither of the genealogies from Matthew and Luke are present, as is also true of the Pericope Adulterae. The Diatessaron was originally written in either Greek or Syriac and became widely used especially in the Syriac church. In the 5th century it was largely replaced by the Syriac version of the four gospels (the Evangelion da-Mepharreshe [Evangel of the Separated Ones]), that had been originally written around the end of the 2nd century:
He (Tatian) composed the Gospel which is called Diatessaron, cutting out the genealogies and such other passages as show the Lord to have been born of the seed of David after the flesh. This work was in use not only among persons belonging to his sect, but also among those who follow the apostolic doctrine, as they did not perceive the mischief of the composition, but used the book in all simplicity on account of its brevity. And I myself found more than two hundred such copies held in respect in the churches in our parts. All these I collected and put away, and I replaced them by the Gospels of the Four Evangelists. (Theodoret, Bishop of Cyrrhus, 453 A.D.)
Although there are no known extant Syriac copies of the Diatessaron, some portions of the text are known from the commentary written in the 4th century by Ephraem Syrus. For the rest only the general order and form of the text is known, from late translations into Arabic (with text close to that of the Peshitta) and Latin (conformed to the Vulgate in many places). Nevertheless, and although the Diatessaron is not strictly a parallel of any gospel, it is a witness to at least the general arrangement and content of the text of the gospels around 170 A.D. It is also an early source from which the Syriac gospels were derived, and therefore must at least be considered when investigating the original wording of Lk 22:17-20. In “The Diatessaron of Tatian,” Hemphill provides the following information regarding the equivalent of Lk 22:17-20 in the Diatessaron, stating that the text basically follows these gospel verses (in this order):
Mk 14:22a, Mt 26:26b, Mk 14:23a, Mt 26:27b, Mk 14:23b-24a, Mt 26:28b-29, Lk 22:19b
He also provides the following two quotes from Ephraem Syrus:
“The Lord blessed and brake”
“Hereafter I will not drink of this fruit of the vine, until the Kingdom of my Father”
In his commentary on early Syriac gospel evidence (Evangelion da-Mepharreshe), Burkitt writes:
The narrative of the supper in the Diatessaron, apart from these extracts [the quotes from Ephraem], is very hard to ascertain. The Arabic (xlv 12-16) gives Matt xxvi 26-29 (= Mk xiv 22-25) followed by Lk xxii 19b ‘and so be doing for my memory.’ But we cannot trust the details where they agree in text with the Peshitta, and possibly the true text of the Diatessaron contained more elements derived from S. Luke. But obviously S. Luke’s account, in which the Cup comes before the Bread, was not made the basis of the narrative.
Hemphill and Burkitt agree that the general order of this passage in the Diatessaron follows Mk 14:22-25 or Mt 26:26-29 (equivalent to Lk 22:19a, 17, 20b-c, 18), followed by the second half of Lk 22:19b (i.e. omitting “which is given for you”), and without Lk 22:20a. However, given the closeness of the text of Mark, Matthew, and Luke in this passage it cannot be certain which gospel text was followed more closely. Indeed, in some places the Diatessaron (as we see it today) clearly merges text, in places from all three synoptic gospels:
And he took a cup, and gave thanks, and blessed, and gave them, and said, Take and drink of it, all of you. [Mk 14:23, Mt 26:27, Lk 22:17]
We also have the anomaly of Lk 22:19 being split, with Lk 22:19a (or the Mark/Matthew equivalents) at the beginning of the passage, and Lk 22:19b at the end. This situation is unique, with all other variants either containing just Lk 22:19a, or Lk 22:19 complete and in contiguous form. In addition, none have any part of Lk 22:19 located at the end of the passage, and in all the other Syriac variants Lk 22:19 is located at the beginning. There therefore appears to be no reason why Tatian would have split Lk 22:19 this way, and also omitted “which is given for you.”
Although Burkitt states that Lk 22:19b is present, what he is in reality saying is that in the late Arabic and Latin translations this passage ends with “and so be doing for my memory” (or perhaps “and thus do ye in remembrance of me”), which he interprets as being the last clause of Lk 22:19b. However, in this position (at the end of the passage) 1 Cor 11:25c seems to be a better match for this text. The difficulty here, of course, is that the Diatessaron is a gospel harmony, and as such it would be very hard to explain the inclusion of just this small fragment of 1 Corinthians 11.
Instead, as both Sy-S and Sy-C (dated to the 4th and 5th centuries respectively) include Lk 22:19b, and both pre-date the extant Arabic and Latin mss of the Diatessaron by several centuries, this text could have been interpolated into the Diatessaron from the Syriac version of Luke during the intervening period. The issue then becomes one of location. Why insert the text at the end, rather than in the ‘correct’ place? The problem is likely to have been uncertainty over what the ‘correct’ place was. In the Syriac (as witnessed by Sy-C and Sy-S) the text is in a different order to that of the Majority Text: It omits some or all of Lk 22:20, and Sy-C has the same words as in 1 Cor 11:24b (omitting ‘given’) in place of Lk 22:19b. These differences, combined with the fact that 1 Corinthians 11 repeats the second half of Lk 22:19b in 1 Cor 11:25c, led to the placing of the phrase at the end of the passage in the Diatessaron, where it could be made to apply to both the bread and the cup.
Around 155 A.D. Justin Martyr wrote the first of two Apologies, in which he referred to and quoted from at least some of the gospels, but as what he writes does not always match what we see in the canonical gospels there is some dispute as to exactly which sources he refers to. For example, in Chapter 66 he writes (re-formatted, and annotated with possible canonical verse references):
For the apostles, in the memoirs composed by them, which are called Gospels, have thus delivered unto us what was enjoined upon them; that
Jesus took bread, and when He had given thanks, [Lk 22:19a, b]
said, “This do in remembrance of Me, this is My body; [Lk 22:19d, c]
and that,
after the same manner, [1 Cor 11:25a / Lk 22:20a]
having taken the cup and given thanks, [Mk 14:23a / Mt 26:27a / Lk 22:17a]
He said, “This is My blood;” [Mk 14:24a / Mt 26:28a]
and gave it to them alone. [Mk 13:23b / Mt 26:27b / Lk 22:19b]
From the tenses he uses it is not always clear whether Justin is quoting from his sources, or paraphrasing. Where it appears he is quoting Jesus’ words then he seems using both Luke and Matthew as sources, and because he uses ‘eucharistēsas’ when referring to Jesus giving thanks it is unlikely that he is using Mark at this point. Some of what he says appears to follow 1 Cor 11:24-25, but is perhaps more likely to have Lk 22:20 as its source. However, as multiple variants of Lk 22:17-20 do exist, it is possible that Justin had one or more of the less well-known variants of these verses as a source, or that perhaps he saw an otherwise unknown variant. There is no mention of either Lk 22:18 or its Mark/Matthew parallels, but it is possible that one or other followed the above text in Justin’s source(s), and he simply did not mention it as it was not germane to the point he was making regarding the Eucharist.
Whether in the text above Justin reliably quotes from his sources or not, it is clear that he follows the Bread-Cup order, and does not have “which is [given] for you.” So, like the Diatessaron, Justin does not support either the Majority Text or the Western text seen in D, a, d, ff2, i, l. However, he does quote “This do in remembrance of Me” from Lk 22:19b, and uses “Likewise” (or “after the same manner”) from Lk 22:20a, so being either directly or indirectly influenced by 1 Cor 11:24-25, and similar to, and possibly having influenced, the Syriac variants of Luke, in particular Sy-S.
As seen above, both 1 Corinthians 11 and Mark 14 contain close parallels to the words we see in Luke, and therefore are possible sources of these words. It is also possible that liturgical practices or different, unknown, sources may be behind the words in Luke but, as Billings states:
If the longer text is a harmonization/conflation in the manner suggested by supporters of the shorter reading then evidence of this should be detectable in the vocabulary, grammar and syntax of Lk 22:19b-20.
Then, after a detailed discussion of the “vocabulary, grammar and syntax,” he concludes that:
It can be … sustained that the words of the longer reading are demonstrably non-Lukan and do not conform to the usual vocabulary, style, and syntax of the writer of the Third Gospel. A source critical analysis reveals that there are enough very close similarities between the text of the longer Lukan reading [Lk 22:17-20] and the parallels in Mark and Paul to at least suggest, and very possibly sustain, literary dependence.
In other words, given the closeness of the text in the parallels, a direct literary relationship between the parallels and Luke is most likely. To highlight these relationships all the parallels are shown in the table below, aligned according to the order of the Lukan Majority Text, together with the comments from Tertullian and the corresponding text of Tatian’s Diatessaron (from the Arabic translation). Lk 22:15-16 are also shown, as Tertullian refers to Lk 22:15, and both these verses have parallels in the Diatessaron.
None of the parallels contain all the elements found in the Majority Text variant of Lk 22:17-20, and what text they do have is not identical to that in Luke, but perhaps the most significant differences are with regard to order:
1 Cor 11:23b-25 refers to bread/body and cup/blood, in that order, with no mention of wine;
Mark, Matthew, Ev, and the Diatessaron have the sequence bread/body - cup/blood - wine;
Luke (Majority Text) has cup – wine - bread/body - cup/blood.
Tertullian has bread/body – cup/blood – wine.
Only the sequence ‘bread/body – cup/blood’ is common to all, while only the Majority Text has the initial cup. Tertullian, the Diatessaron, Mark, and Matthew refer to wine (”the fruit of the vine”) last, while in the Majority Text it follows the initial cup, and is not mentioned in 1 Corinthians 11. Tertullian does not directly connect “the cup” with the “covenant” (as is done in 1 Cor 11:25), but instead connects it with “His blood.” On this basis it appears that Tertullian’s words more closely match the order and content of the text found in Mark than that in the Majority Text of Luke:
Ev [Per Tertullian]:
Then, having taken the bread and given it to His disciples, He made it His own body, by saying, "This is my body," that is, the figure of my body…
He likewise, when mentioning the cup and making the covenant to be sealed in His blood, affirms the reality of His body…
In order, however, that you may discover how anciently wine is used as a figure for blood, turn to Isaiah…
Mk 14:22-25
And as they did eat, Jesus took bread, and blessed, and brake it, and gave to them, and said, Take it: this is my body. [c.f. Lk 22:19a]
And he took the cup, and when he had given thanks, he gave it to them: and they all drank of it. [c.f. Lk 22:17]
And he said unto them, This is my blood of the covenant, [c.f. Lk 22:20b]
which is shed for many. [c.f. v. 20c]
Verily I say unto you, I will drink no more of the fruit of the vine, [c.f. Lk 22:18a]
until that day that I drink it new in the kingdom of God. [c.f. Lk 22:18b]
It is conceivable that the words Tertullian saw in Ev were derived from Matthew, but on balance this seems unlikely. For example, Mk 14:25a has a variant that may bear on the text of Luke. Although this is not shown in the KJV, it can be seen by using the Lexham English Bible text instead:
Neither Mt 26:29a nor Lk 22:18a have ‘ouketi’ (any longer). This is considered to be a minor agreement of Matthew and Luke against Mark, although it (or its Latin equivalent) is not present in Mark in 01, C, D, L, W, Ψ, 0103, 892, 1342, pc34, a, c, d, f, bo. However, D, a, d, f, instead have the Greek or Latin word meaning ‘proceed, go ahead, continue, do again,’ before ‘drink,’ so that d (for example) reads “Quia non adponam bibere,” while two Greek mss have both.
The variants seen in Mark are also seen in Luke. As mentioned above, Lk 22:18a does not contain ‘ouketi,’ and this is also the case in Lk 22:16a in NA27, following a number of important mss including P75, 01, A, and B. Then, although ‘hoti’ is present in Lk 22:16a, it is not present in Lk 22:18a in P75vid, B, C, (D, G), L, (f1, 157), d, and NA25. In contrast to both Mark and Luke Matthew has no variants in Mk 26:29a, with ‘hoti ouketi’ not present in any extant mss. The uncertainty regarding Mark and Luke, combined with the lack of variants in the parallel in Matthew, make it more likely that Luke is here derived from Mark than from Matthew.
Because Tertullian does not mention any differences between Ev and Luke in this passage we can reasonably assume that his copy of Luke had essentially the same text as he saw in Ev. Although Tertullian does not quote the exact words we see in Mark, some differences may simply be translation issues, because it is likely that he had an Old Latin copy of Luke. However, there may be a more specific explanation for the change from the text of Mk 14:25b to that of Lk 22:18b:
… until that day that I drink it new in the kingdom of God. [Mk 14:25b]
… until the kingdom of God shall come. [Lk 22:18b]
It is possible that the difference can be explained as a desire to remove the ‘personal’ connection between Jesus and the kingdom of God implied by the text in Mk (and even more so in Mt 26:29b), in favor of a ‘generic’ reference to the coming of the kingdom of God, as seen in many other places in Luke.
Tertullian does not refer to any equivalent to Mk 14:22a “And as they did eat,” and it is likely that he did not see it in either Ev or his copy of Luke. Unlike both Mark and Matthew, in this early variant of Luke (with the text order as reported by Tertullian) the phrase would have been immediately preceded by references to eating in Lk 22:15-16, and so the author may have felt that this ‘scene setting’ was not necessary here. It is significant that this phrase does not appear in any of the Lukan variants, even in the Majority Text, where it might have been expected to have been added between Lk 22:18 and 19. Allowing for the typical variations we see elsewhere between Mark and Luke, and given that all the variants contain language closer to Luke than to Mark, it is therefore possible that what Tertullian saw in Ev read approximately as follows:
And he took bread, and blessed, and brake it, and gave unto them, saying, This is my body. [cf. Lk 22:19a]
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17]
And he said unto them, This is my blood of the covenant. [c.f. Lk 22:20b]
For I say unto you, I will not drink of the fruit of the vine, [Lk 22:18a]
until the kingdom of God shall come. [Lk 22:18b]
As seen in the comparison table above, the text of the Diatessaron appears to support this conclusion, omitting both “which is given for you” from Lk 22:19b, and also Lk 22:20a. As the preceding text in the Diatessaron relating to the last supper (i.e. Lk 22:15-16) can only have come from Luke (these verses have no parallels in Mark or Matthew), we can be certain that Tatian was not simply ignoring Luke at this point in his harmony. From the above we can see that the Diatessaron does not follow the Majority Text, and indeed appears to not know it. Instead, it follows the Matthew/Mark order, and arguably contains nothing that derives from 1 Cor 11:23b-25, as “and so be doing for my memory” may be a later assimilation. This suggests that the text of Luke used by Tatian did not include Lk 22:19b-20a, and that it also followed the Matthew/Mark order. Although the Arabic version of the Diatessaron does include “And while they were eating,” it is likely that this text was included because it is in both Mark and Matthew, although not Luke.
In other words, an early form of the Lukan text (as known to both Tertullian and Tatian) was similar to what we see today as Lk 22:19a, 17, 20b, 18, with the most significant difference being that neither Tertullian nor Tatian mention the atoning sacrifice that we see in Lk 22:20c (Tatian has the Matthean form), which Tertullian therefore may not have seen in Ev. The above suggests that this was an early (Mid-second century at the latest) Western form of the text in Luke, but for this to possibly be the earliest variant we must be able to satisfactorily account for the creation of the other variants from it. In particular, if this is an early form of the Lukan text, how and why did text apparently taken from 1 Cor 11:24-25 make its way into later variants of Luke, including the Majority Text?
As stated above, there are six different variants of Lk 22:17-20 in the extant mss of Luke, with Ev and the parallels in Mark, Matthew, and 1 Cor 11:24-25 providing four more accounts. Although not strictly a parallel, the Diatessaron includes text from Luke and can therefore also be considered to be a different account, and these eleven different accounts share several patterns (where verses in Luke are given read non-Lukan sources as having the corresponding parallels):
Lk 22:19a is present in all Lukan variants, with Mark, Matthew, 1 Corinthians 11, and the Diatessaron all containing close parallels;
Nine accounts begin with Lk 22:19a, giving the sequence ‘bread-cup,’ while two have Lk 22:17-18 before Lk 22:19a, giving the sequences ‘cup-bread’ and ‘cup-bread-cup’ respectively;
Some of the accounts do not contain all of Lk 22:19-20, but those parts that are present are always in the same order as in the Majority Text, i.e. 19a, 19b, 20a, 20b, 20c, although they are not always contiguous (As indicated above, it is assumed here that what appears to be part of Lk 22:19b in the Diatessaron is a late interpolation);
All accounts except 1 Corinthians 11 and the Peshitta contain both Lk 22:17 and 18, in that order, although not always in contiguous form;
Where Lk 22:17-18 precede Lk 22:19a they always exist as contiguous verses, but where they follow Lk 22:19a they may be separated by Lk 22:20b and c;
Parallels of Lk 22:17-18 exist in both Mark and Matthew, and although the wording in Luke is not exactly the same as in these parallels, the essential details are present and in the same order.
The accounts can be split into two groups, depending on whether they omit both Lk 22:19b and 20a (or their parallels), or include either one or both. It is convenient to consider the two groups separately, and for ease of visualization the two tables below show the order of the text (from left to right) in each account, using the Lukan verse numbers, or those of the parallels in Mark, Matthew, 1 Corinthians 11, or Diatessaron, as appropriate. The Diatessaron is included in the first group on the assumption that the fragment of Lk 22:19b after Lk 22:18 is a late interpolation, as already stated.
None of the variants of Luke shown in the above table (Ev, D, and the Old Latins) require knowledge of the contents of 1 Corinthians 11, and so could have been created just by editing the text of Mark. Indeed, because of the closeness of the text, it is reasonable to believe that Mark is the sole source of the text of these variants. With the exception of D the Lukan variants shown here are written in Old Latin, although D does seem to preserve a very early form of the Greek text, perhaps even pre-dating the Old Latin. Billings writes:
The antiquity of the Western text type, embodied by D, was of course a key tenet of Westcott and Hort’s textual theory who thought it to go back at least until the second century and even went so far as to identify the origin of the ‘Western Non-Interpolations’ with the Lukan autograph itself. The antiquity of the Western textual tradition, embodied by D, has been supported also by A. F. J. Klijn, who, in an extensive investigation, finds evidence of a Western text type predating the Old Latin versions in Marcion. Kirsopp Lake went further, even suggesting that Wescott and Hort had in fact probably underestimated the antiquity of the Western text type as it is represented by D, insisting it is ‘everywhere found where we have any evidence for the text of the second century in patristic references’… Perhaps the most that can be decisively claimed is the conclusion of Dom Gregory Dix that ‘the oldest form of the Western Text’ as represented by D and its allies ‘must certainly have existed in the early second century’.
Whether Tertullian was commenting on a Greek text of Ev, or a Latin translation from Greek, the above lends support to the likelihood that he was commenting on text that was, or was very close to, the earliest Greek form of this passage.
In addition to the omission of Lk 22:19b-20 (one of the major Western Non-Interpolations identified by Hort), these accounts are all very similar (all being classed as Western). It is therefore unlikely that they are descended from several different independent translations of a number of Greek originals. Instead, it is more likely that they have a single (non-extant) Western Greek ancestor in common, and some other Old Latin variants (in which a number of mss have the same unusual reading, e.g. at Lk 6:42 and 13:8) provide support for this view.
As shown above, each variant of Luke in this group could have been created by editing text found in Mark only. However, if the initial Lukan text was based on the Markan version of what we see today as Lk 22:19a, 17, 20b-c, and 18, then why is Lk 22:20c absent from all these accounts, and Lk 22:20b absent from all the Latins? The important point is that Lk 22:20b-c refer to the atonement doctrine:
In Christian theology the atonement refers to the forgiving or pardoning of sin through the death of Jesus Christ by crucifixion, which made possible the reconciliation between God and creation. (Wikipedia)
In Luke today Lk 22:20b-c reads:
This cup is the new covenant in my blood, which is shed for you.
However, the atonement doctrine itself is non-Lukan, as Ehrman emphasizes:
When Jesus says in Luke 22:19b-20 that his body is given “for you” (hyper hymōn) and that his blood is shed “for you” (hyper hymōn), he is stating what Luke says nowhere else in his long narrative. Neither in his Gospel nor in Acts does he portray Jesus’ death as an atonement for sins. Even more significantly, Luke has actually gone out of his way to eliminate just such a theology from the narrative he inherited from his predecessor, Mark.
Ehrman also notes that Luke has no parallel to Mk 10:45, and that the force of the centurion’s words in Mk 15:39 is completely changed in the parallel at Lk 23:47, from Jesus being “the Son of God,” to just “a righteous man.” Also, it is likely that the change from the words in Mk 14:25b (Jesus participating in the kingdom of God) to what we see in Lk 22:18b (just the coming of the kingdom of God) is part of this elimination process. Billings portrays the issue in this way:
The omission of Mk 10:45 and of any explicit reference to the death of Jesus as a vicarious atonement for sins is made all the more striking by the fact that Luke, who alone among the Gospel writers has opportunity to look back and theologically reflect on the events of the passion in his second volume, nowhere summarizes the death of Jesus as an atoning sacrifice in any of the speeches in the Acts but seems to consistently portray the death of Jesus as that of an innocent martyr whom God vindicated and glorified by the resurrection.
The lack of any mention of the atonement in Luke - Acts except in Lk 22:19b-20 is, Ehrman contends:
… a key factor in recognizing the secondary character of the longer text, a factor that has been surprisingly underplayed by most previous studies of the problem. … only one of our two textual variants, the shorter reading, coincides with the Lukan understanding of Jesus’ death; the other attests precisely the theology that Luke has otherwise taken pains to suppress.
The conclusion to be reached from the above is that the original variant of Lk 22:17-20, like the Didache, did not contain any reference to Jesus’ atoning sacrifice, and, beginning with a ‘base’ text from Mark 14, the simplest way for the author of Luke to have achieved this would be to remove mention of Jesus’ blood being shed for anyone from Mk 14:24b, leaving a close match to what is reported by Tertullian:
And he took bread, and blessed, and brake it, and gave unto them, saying, This is my body. [Lk 22:19a]
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17]
This is my blood of the covenant, [Mk 14:24a]
For I say unto you, I will not drink of the fruit of the vine, until the kingdom of God shall come. [Lk 22:18]
However, Billings reports that in the second century:
The Christian meal practices seem to have attracted particular suspicion among their social peers” and the “Christian meals were widely believed to include also the drinking of human blood, a ritual practice associated in Greco-Roman texts with sorcery, magic, and the making of oaths ("blood covenants," or sacramentum) that carried with it the further implication of the presence of a sacrificial victim, usually assumed to be an infant.
This issue was of importance to Justin Martyr, as in Dialogue with Trypho (Chapter 10) (written around the mid-2nd century) he says:
Is there any other matter, my friends, in which we are blamed, than this, that we live not after the law, and are not circumcised in the flesh as your forefathers were, and do not observe sabbaths as you do? Are our lives and customs also slandered among you? And I ask this: have you also believed concerning us, that we eat men; and that after the feast, having extinguished the lights, we engage in promiscuous concubinage?
For this reason it is likely that some Christian groups would want to disassociate themselves from any suggestion of drinking blood, and an example of this can be found in The Homilies of Aphraates, who wrote in the period 337-345 A.D., in or around Seleucia. This account of the last supper (Aphraates 221) appears to mainly follow the Diatessaron (although omitting some text quoted by Ephraim), and is close to the Nestorian ‘Liturgy of Theodore:’
He took bread and blessed and gave to His disciples and said to them: “This is my body, [c.f. 1 Cor 11:23b-24a]
the new covenant that for many is shed to forgiveness of sins: [c.f. Mt 26:28]
thus be ye doing for my memory [c.f. 1 Cor 11:24b]
whenever ye be gathered together.”
The most important point here is that in place of text either from 1 Cor 11:25 or Lk 22:20, we have an edited version of what we see as Mt 26:28, in which only the references to the cup and the blood have been omitted. This text provides support for the possibility that the reference to blood could have been omitted in an early Old Latin variant of this passage, leaving just the text seen in b and e:
And he took bread, and blessed, and brake it, and gave unto them, saying, This is my body. [Lk 22:19a]
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17]
For I say unto you, I will not drink of the fruit of the vine, until the kingdom of God shall come. [Lk 22:18]
The above all suggests that the development of these shorter Lukan variants took place in two stages:
The initial Greek version of Lk 22:17-20 (as seen by Tertullian in Ev) included text from Mark, using the same order, except that the reference to Jesus’ blood being shed was removed.
The Old Latin text seen in b and e was based on translations of a single Greek ms from the first stage, but with all references to blood being removed.
Although this also explains the content of D, a, d, ff2, i, and l (the same as b and e), it does not explain the verse order, in which Lk 22:19a follows vv. 17-18 (so changing the bread-cup sequence into cup-bread). It has been suggested that this was due to different liturgical practices in different areas, and given what we see in the Didache this may be the case, but it is also likely that the order was changed to be closer to what was by then becoming the Greek ‘Majority Text,’ as described below.
The Lukan variants in this table all include text seen in Lk 22:19b (and also in Lk 22:20a in all but Sy-C) that does not exist in Mark, Matthew or in any of the accounts in the first group, and instead has a close parallel in 1 Cor 11:24b-25a. However, this does not on its own require that 1 Cor 11:24b-25 must be the source of Lk 22:19b-20, as commented on by Willker:
The proponents of the shorter form normally argue that the longer version is an adaption of 1 .Co 11:24-25. This is possible, but it is equally probable that the words are so similar, because they are used liturgically from early on.
However, as shown below, when the parallels in Mark and Matthew are also taken into account, the evidence does in fact suggest that 1 Cor 11:24-25 is likely to be the source.
which is given for you: this do in remembrance of me. [22:19b]
Likewise also the cup after supper, saying, [22:20a]
This cup is the new testament in my blood, [22:20b]
The closeness of the parallels between Lk 22:19b-20 and 1 Cor 11:24b-25 in the Majority text is indisputable, as shown below (using the Lexham English Bible text), with Mk 14:24a, Mt 26:28a and Heb 9:20a shown for comparison:
Lk 22:19b which is given for you. Do this in remembrance of me. to hyper hymōn didomenon touto poieite eis tēn emēn anamnēsin
1 Cor 11:24b which is for you. Do this in remembrance of me. to hyper hymōn touto poieite eis tēn emēn anamnēsin
Lk 22:20a And in the same way the cup after they had eaten, saying, kai to potērion hōsautōs meta to deipnēsai legōn
1 Cor 11:25a Likewise also the cup, after they had eaten, saying, hōsautōs kai to potērion meta to deipnēsai legōn
Lk 22:20b This cup is the new covenant in my blood, Touto to potērion hē kainē diathēkē en tō haimati mou
1 Cor 11:25b This cup is the new covenant in my blood. Touto to potērion hē kainē diathēkē estin en tō emō haimati
Mk 14:24a And he said to them, This is my blood of the covenant kai eipen autois Touto estin to haima mou tēs diathēkēs
Mt 26:28a for this is my blood of the covenant touto gar estin to haima mou tēs diathēkēs
Heb 9:20a saying, "This is the blood of the covenant Legon Touto to haima tēs diathēkēs
Although Lk 22:19b adds “given,” to 1 Cor 11:24b, and Lk 22:20a has a slightly different word order to 1 Cor 11:25a, they are essentially the same, while in contrast there are no parallels to Lk 22:19b-20a in either Mark or Matthew. Utley has this to say regarding the phrase “do this in remembrance of me” in v. 19b:
The phrase is unique to Luke’s Gospel. [i.e. it does not appear in either Mark or Matthew] The word anamnēsis occurs twice in Paul’s account of the Lord’s Supper in 1 Cor 11:24, 25. Luke may have gotten his terminology from Paul’s churches. This is probably why there are several non-Lukan forms and words in vv. 19b-20.
The text of Lk 22:20b is much closer to that of 1 Cor 11:25b than to either of its Mark or Matthew parallels, in particular mentioning “this cup,” and also the “new covenant,” instead of just the “covenant”’ as is found in the early Mark/Matthew mss, Heb 9:20a, and as mentioned by Tertullian. As indicated, it is highly unlikely that either Mark or Matthew is the source of these words in Luke. Therefore, because (as Billings states) “the words of the longer reading are demonstrably non-Lukan” and there is “literary dependence,” there is a very high probability that the words of Lk 19b-20b came from 1 Cor 11:24b-25.
which is shed for you. [22:20c]
In the Majority Text the words in Lk 22:20c are similar to that of Mk 14:24b / Mt 26:28b (all three beginning “which is shed [or ‘poured out’] for…”), but are very different to those in 1 Cor 11:25c and Heb 9:20b. Although in Mark/Matthew the blood of Jesus is shed for “many,” while in Lk 22:20c it is shed for “you,” this appears to refer back to Ex 24:8 (in which the covenant is also made with “you”) and not 1 Cor 11:25c:
Because of this similarity between Lk 22:20c and the Mark/Matthew parallels, it would be reasonable to expect that Lk 22:20b would also be similar to the words in Mark/Matthew, but as previously noted the source of Lk 22:20b is instead most likely 1 Cor 11:25b. In other words, Lk 22:20b and Lk 22:20c appear to have different sources: Either the variants in the second group that contain both Lk 22:20b and 20c derive from an earlier variant that contained the Mark/Matthew form of Lk 22:20b, which was then replaced by 1 Cor 11:25b, or they arose independently of the (shorter) variants in the first group. However, the fact that (when present) Lk 22:20c is always similar to the Mark/Matthew parallels, and not to 1 Cor 11:25c, suggests that the second group of variants developed from one in which the Mark/Matthew-based parallel was already present.
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [22:17]
For I say unto you, I will not drink of the fruit of the vine, [22:18a]
until the kingdom of God shall come. [22:18b]
And he took bread, and gave thanks, and brake it, and gave unto them, saying, This is my body [22:19a]
In addition to Lk 22:19b-20a, in the second group all accounts but the Peshitta include Lk 22:17-18, which have no parallels in 1 Corinthians 11. These variants therefore also required knowledge of Mark/Matthew, whatever was the source for Mark/Matthew, or an earlier variant of Luke that already contained these verses. As the text common to most of the Lukan variants in both tables (e.g. Lk 22:17-19a) is basically what we see in Luke instead of that in the Mark/Matthew parallels, the second group of variants is most likely based on an existing variant of Luke that already contained this text, and not an independent translation from either Mark/Matthew or an earlier source.
On this hypothesis each Lukan account in the second group reflects a merging (or interpolation) of the text in 1 Corinthians 11 with a Lukan variant from the first group, and this can be tested by seeing whether it can satisfactorily explain the existence of all the accounts. However, before investigating any textual development within the second group, we first have to determine whether the accounts in the first group could have given rise to all those in the second group, or not.
One important point to consider is the implication of the simple fact that there is a Majority Text variant, and its sheer numerical superiority is a strong (perhaps the strongest) point in favor of it being the original form. Holding writes:
The evidence in favor of the longer text, however, appears to be very strong. First, the external evidence vastly favors it. Metzger writes, "the longer, or traditional, text of cup-bread-cup is read by all Greek manuscripts except D and by most of the ancient versions and Fathers;" and "The external evidence supporting the shorter reading represents only part of the Western type of text, whereas the other representatives of the Western text join with witnesses belonging to all the other ancient text-types in support of the longer reading."… Metzger concludes on the matter regarding the opinions of the textual committee:
A minority preferred the short text as a Western non-interpolation… The majority, on the other hand, impressed by the overwhelming preponderance of external evidence supporting the longer form, explained the origin of the shorter form as due to some scribal accident or misunderstanding…
However, arguments that rely on an un-explained “scribal accident or misunderstanding” to support them are resting on thin ice, as without an extant exemplar from which the shorter form descends it is just conjecture. So, what is the strength of the numerical argument in favor of the longer text? If the Majority Text does represent the initial form then its numerical superiority is not in any way remarkable, as many copies could have already been written before the shorter variants were created. However, if the Majority Text is not the initial form then how could the shorter variant that WAS the initial form not become numerically superior (at least in the first few centuries) instead? If the variants that do not include text seen in 1 Corinthians 11 were written first then we might reasonably expect to see many more extant copies of these shorter forms today. However, there are valid reasons why this might not be the case:
The shorter variants may have existed in only a very few copies, known only to a particular local group, or perhaps had a limited geographic range. The fact that the non-Majority Text variants are primarily Old Latin and Syriac lends weight to this possibility;
Even if many copies of the shorter variants were created, their sheer age may have meant that only a few copies survived to the present day;
Whatever became the view of the established church is more likely to have been copied (and thus survived) than early beliefs, traditions, or practices that fell out of favor. For example, all but one of the non-Majority text variants omit the crucial part of the atonement doctrine text that occurs in both Mark and Matthew: “which is shed for you” (Lk 22:20c), the addition of which to Luke may have become very important to the church leadership.
The fact that the Vulgate supplanted the Old Latin, and, as mentioned earlier, the Diatessaron was replaced by the separate Syriac gospels, shows clearly that the numerical superiority of a variant today says nothing about its age. The same can be said to apply to all the ‘Western’ textual variants, which, while very old, are clearly in the minority in the manuscripts.
A further point to consider is why it was apparently felt necessary to interpolate text from 1 Corinthians 11 into Lukan text that originated in Mark (thus creating the cup-bread-cup sequence in Luke), but not into the parallel text in Mark (or Matthew) itself. This suggests that Luke was somehow viewed differently than Mark and Matthew. One possibly is that Luke became ‘tainted’ due to the association with Marcion, and it was felt necessary to make changes to the text of Luke to make it more distinct from Ev, for example by creating a different ‘last supper’ narrative. Lietzmann & Richardson put it this way:
St. Paul’s Epistles make frequent use of such words as ‘given for us,’ but not in connection with the Supper-rite; and the eucharistic use of this phrase requires separate investigation. Marcion may have known it; but, even so, the proper conclusion to be drawn from the hazards of his evidence, as from the far clearer witness of Justin Martyr, is that the words of institution had as yet no finally fixed form. Marcion’s importance in the sphere of eucharistic liturgy, as in that of doctrine and in the history of the formation of the New Testament canon, lies chiefly in his having hastened into fuller definiteness processes that were already maturing.
It may well be no coincidence that despite the Western text type being “everywhere found where we have any evidence for the text of the second century in patristic references” (Kirsopp Lake), we have the evidence of P75 that the Majority Text variant existed in the early third century, and, because it seems highly likely that P75 and B had a common ancestor, probably even earlier. In fact it was the existence of P75 that in large part resulted in the re-instatement of the text of the Western Non-Interpolations (including Lk 22:19b-20):
The discovery of P75 has had a profound effect on New Testament criticism. The demonstration that the B text is older than B seems to have encouraged a much stronger belief in its originality. The UBS committee, for instance, placed the Western Non-Interpolations back in their text based largely on the evidence of P75. (Waltz)
However, as Colwell points out, the existence of P75 should not have had this effect:
Professor Aland … reverses Westcott and Hort on the Western non-interpolations because P75 disagrees with them in agreeing with Codex Vaticanus. But there is nothing in that agreement that is novel to Hort’s theory. Hort did not possess P75, but he imagined it. He insisted that there was a very early ancestor of his Neutral text, that the common ancestor of Vaticanus and Sinaiticus was a remote ancestor, not a close ancestor. P75 validates Hort’s reconstruction of the history, but P75 does not add a new argument for or against that theory.
In essence, P75 cannot be used to support the existence of the ‘B text’ any earlier than one generation prior to itself. However, it can support the view that there was a shift away from the Western text to the ‘B text’ (including the addition of Lk 22:19b-20) shortly after Marcion became labeled as a heretic in AD 144. If this was the case then it would be natural for the Majority Text variant of Lk 22:17-20 to be promoted and become the dominant form, and the shorter variants (associated with Marcion) to be preserved only outside the ‘mainstream’ Greek-speaking Christian community.
The above not only provides a good reason for the creation of an expanded variant of Lk 22:17-20, but also why this variant could become dominant even though it was not the original form of these verses, and this adds a third possible stage to the development of the various Lukan variants:
The initial Greek version of Lk 22:17-20 (as seen by Tertullian in Ev) included text from Mark (using the same order), except that the reference to Jesus’ blood being shed was removed.
The Old Latin text seen in a, b, d, e, ff2, i, l, was based on translations of a single Greek ms from stage 1, but with all references to blood being removed.
The Greek text from stage 1 (or, less likely, a translation of one of the old Latin variants from stage 2) was merged with text from First Corinthians 11, and the atonement doctrine text was re-instated, leading to the Majority Text and the Syriac variants.
The above provides an explanation for how these variants of Lk 22:17-20 came to include text from 1 Corinthians 11, but does not explain why the original Mark/Matthew verse order (19a, 17, 20b-c, 18) became changed, with Lk 22:17-18 coming first in the Majority Text variant.
As noted above, the text most obviously ‘missing’ from the shorter variants is some or all of the atonement doctrine, and if the shorter variants are early then it is a reasonable supposition that someone would want to insert it at some point. One obvious possibility is that words from Mk 14:24 could have simply been ‘re-instated’ more or less as they were:
… This is my blood of the testament, [14:24a] which is shed for many. [14:24b]
However, while an edited version of Mk 14:24b is present in the Majority Text (as v. 20c), it appears that the opportunity was taken to include the more ‘complete’ text from 1 Cor 11:24-25 instead of Mk 14:24a. Then, when this text was inserted (interpolated) into the original Lukan version of the Mark/Matthew text the most straightforward change would appear to have been to replace “This is my blood of the covenant” by 1 Cor 11:24b-25b. However, this change on its own would create problems, because the result would then read:
And he took bread, and blessed, and brake it, and gave unto them, saying, This is my body. [Lk 22:19a, from Mk 14:22]
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17, from Mk 14:23]
which is for you: this do in remembrance of me. [1 Cor 11:24b]
After the same manner also he took the cup, when he had supped, saying, [1 Cor 11:25a]
this cup is the new covenant in my blood; [1 Cor 11:25b]
which is shed for you. [Lk 22:20c, from Mk 14:24b]
For I say unto you, I will not drink of the fruit of the vine, until the kingdom of God shall come. [Lk 22:18, from Mk 14:25]
Here Lk 22:17 obviously interrupts the text from Lk 22:19a and 1 Cor 11:24b, and there are three simple ways to resolve this:
Remove Lk 22:17;
Swap Lk 22:17 and 19a;
As 2), but also merge the text from 1 Cor 11:23b-24a with the original version of Lk 22:19a from Mk 14:22.
As Lk 22:17 is still in the Majority Text the first alternative was obviously not used, but as either of the other two solutions places what we see as Lk 22:17 first, the issue revolves around how similar the text of Lk 22:19a is to either Mk 14:22 or 1 Cor 11:23b-24a. For example, with what we see as Lk 22:19a no longer placed immediately after Lk 22:15-16, it might be expected that “while they were eating” (from Mk 14:22) might have been re-instated in order to provide more context here, but as this phrase does not appear in any Lukan variant there is no doubt that this did not happen.
Instead, evidence that at least some text from 1 Cor 11:23b-24a was used can be seen by noting that in Mk 14:22 Jesus is blessing (‘eulogēsas’), whereas in what we see today as Lk 22:19a he is giving thanks (eucharistēsas), as is also seen in 1 Cor 11:24a. The choice of Greek words here appears to have no theological significance, and was instead a simple authorial choice, as indicated by the following from Pope Benedict XVI in 2009:
Jesus' action is described in the words: ‘gratias agens benedixit -- he gave you thanks and praise’. In this expression, the Roman liturgy has made two words out of the one Hebrew word berakha, which is rendered in Greek with the two terms eucharistía and eulogía.
As Lk 9:16 and 24:30 both have ‘eulogēsen’ (showing that aLuke used this term in other places), it would be natural for aLuke to use it in Lk 22:19a as well, so the fact that Lk 22:19a has ‘eucharistēsas’ instead suggests that this word originated in text from a different author, e.g. from 1 Cor 11:24a.
Interestingly, in Lk 22:19a the Old Latin ms d has ‘blessed’ (benedixit), while the Vulgate and most of the other Old Latins (including b and e) refer to Jesus giving thanks (gratias egit). This might appear to suggest that the Majority Text (which includes text from 1 Cor 11:24a) came before the Old Latins, except that it seems very unlikely that ‘eucharistēsas’ would be translated as ‘benedixit’ in even one of the extant Old Latin mss if this was the case. Instead, it is more likely that only the earliest Old Latin originally had ‘benedixit’ (as seen in d), while the text of the Vulgate, and then the later mss, were influenced by what had by then become the Greek Majority Text. Also, although the Syriac versions also refer to Jesus giving thanks (w'wdy), this is to be expected, as they all include text from 1 Corinthians 11. The remaining small differences can be easily explained:
Jesus is not referred to by name in Lk 22:19a as his name is not needed here (the first mention of Jesus in Lk 22 is at Lk 22:47).
The superflous “Take it” in Mark is omitted in Lk 22:19a, following text from 1 Cor 11:24a.
This inclusion of the additional text from 1 Cor 11:23b-24a as just described results in the following:
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17, from Mk 14:23]
And he took bread, and gave thanks, and brake it, and gave unto them, saying, This is my body [Lk 22:19a, from Mk 14:22b/1 Cor 11:24a]
which is for you: this do in remembrance of me. [from 1 Cor 11:24b]
After the same manner also he took the cup, when he had supped, saying, [from 1 Cor 11:25a]
this cup is the new covenant in my blood; [from 1 Cor 11:25b]
which is shed for you. [Lk 22:20c, from Mk 14:24b]
For I say unto you, I will not drink of the fruit of the vine, until the kingdom of God shall come. [Lk 22:18, from Mk 14:25]
This simple change – the insertion of text from 1 Corinthians 11 into the original Mark/Matthew text and the consequent removal of any duplication – naturally gives rise to the creation of the cup-bread-cup sequence, leaving what is essentially the Majority Text variant of Luke, except for one final issue: The position of Lk 22:18. The most simple explanation for what we see today would seem to be a desire to contrast Jesus’ comments regarding not drinking wine with the action of asking the disciples to drink, while at the same time making sure that nothing followed the important atonement message at the end of Lk 22:20.
D and the old Latin mss a, d, ff2, i, and l all contain the same short text as b and e (Lk 22:19a, 17-18), but in the order vv. 17-18, 19a, so changing the bread-cup order seen in Mark to cup-bread instead:
And he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17]
For I say unto you, I will not drink of the fruit of the vine, until the kingdom of God shall come. [Lk 22:18]
And he took bread, and blessed, and brake it, and gave unto them, saying, This is my body. [Lk 22:19a]
This variant contains nothing derived from 1 Corinthians 11, and consequently there is no obvious reason for the text order to have been changed from that seen in Mark/Matthew and the other short variants, except perhaps to reflect the order seen in the Didache. It is also possible that the order was changed simply to be closer to that in what was by then becoming the Greek ‘Majority Text,’ as described above, but (like b and e) without adding any of the atonement doctrine text (vv. 20b-c), or any references to blood. The only Greek ms containing this variant, D, almost certainly simply reflects the order seen in d.
The Majority Text seems to have presented some people with a problem, which, as many have noted, is the duplicated references to ‘the cup’ (giving the familiar cup-bread-cup sequence). In the Syriac variants Lk 22:19a is placed first (as in Mark), so referencing the bread at the beginning, but what would otherwise be duplicated references to the cup in the following verses are then avoided in different ways.
Sy-P (Peshitta) omits the whole of Lk 22:17-18, leaving just Lk 22:19-20.
And he took bread, and gave thanks, and brake it, and gave unto them, saying, This is my body [Lk 22:19a]
which is given for you: this do in remembrance of me. [Lk 22:19b]
Likewise also the cup after supper, saying, [Lk 22:20a]
This cup is the new testament in my blood, [Lk 22:20b]
which is shed for you. [Lk 22:20c]
The fact that Lk 22:19-20 are both adjacent and present in their entirety shows that the Peshitta is derived from the Majority Text. This can also be seen by the fact that if it had been derived from any variant in which Lk 22:18 followed Lk 22:20 then there would have been no need to remove Lk 22:18.
Sy-S and Sy-C are representative of the Syriac version of the four gospels. Both include text that may have originated in the Diatessaron, while Sy-C also appears to have been revised towards the Greek, although the text of Sy-C appears, in general, to be older than that of Sy-S. Wright provides the following example:
But it is very difficult to believe that ‘Father, forgive them,’ was inserted in C just before Lk xxiii 34b, exactly at the place where the words are inserted in Greek mss, if we are to suppose that the interpolator took them from the Diatessaron. For according to the Arabic Harmony (lii 6, 7), apparently supported by S. Ephraim’s Commentary, the saying ‘Father, forgive them,’ was assigned by Tatian to our Lords last moments, whereas in the mss of the Gospels the saying is always placed at the moment of crucifixion.
He later quotes Hort on the character of Sy-C:
He [Hort] said … ‘The character of the fundamental text confirms the great antiquity in its original form; while many readings suggest that, like the Latin version, it degenerated by transcription and perhaps also by irregular revision.’
And also:
Where S and C agree we may be confident that we have the original text of the Evangelion da-Mepharreshe, the earliest translation of the Four Separate Gospels into Syriac; where S and C agree together with the Diatessaron, we must believe that the translator of the Evangelion da-Mepharreshe was content to adopt the rendering of the Diatessaron into his Syriac text of the Four Gospels.
In contrast to the Pehsitta, in both Sy-S and Sy-C Lk 22:19a, 17 and 18 are present (in that order, so matching the order of the parallel in Mark), but either some or all of Lk 22:20 is omitted. Both Sy-S and Sy-C group the references to the ‘cup’ and wine together at the end, and from the order of the text it is clear that neither is derived from the Majority Text, although the inclusion of text similar to that found in Lk 22:19b and 1 Cor 11:24b-25 following Lk 22:19a suggests that whoever wrote these texts knew of it.
As previously mentioned, what Justin Martyr reports in his first Apology regarding the Eucharist has a lot in common with the Syriac variants. Not only does it begin with text very similar to Lk 22:19 (albeit with some change of order), it then follows with some of Lk 22:17, which (as in Sy-S) is preceded with part of Lk 22:20a and followed by a reference to blood similar to that in Mk 14:24a. Although Justin does not mention Lk 22:18, there is the possibility that in his source(s) it followed the text he does report, but that he had no reason to refer to it. Whether or not this was the case, there is enough similarity between what Justin reports and the Syriac variants to suggest that this is their origin.
Sy-C is the shorter of the two, omitting all of Lk 22:20. Also, although Sy-C appears to include the whole of Lk 22:19, the word ‘given’ is omitted, suggesting that these words are either taken directly from 1 Cor 11:24b, or an early variant of Lk 22:19b:
And he took bread, and when He had given thanks, He brake it, and gave to them, saying, This is My body, [Lk 22:19a]
Which (is) for you: this do in remembrance of me. [c.f. Lk 22:19b, 1 Cor 11:24b]
And he took a cup, and when He had given thanks, He said, Take this and divide it among yourselves: [Lk 22:17]
I say unto you, I will not drink of the produce of the vine, until the kingdom of God shall come. [Lk 22:18]
By not including Lk 22:20 this variant avoids the second mention of the cup, and also the references to supper, the covenant, and blood being shed for anyone. Sy-S is similar to Sy-C, including the whole of Lk 22:19 (including ‘given’), but surrounding Lk 22:17 with text possibly from Lk 22:20a or 1 Cor 11:25:
And he took bread, and gave thanks, and brake it, and gave unto them, saying, This is my body [Lk 22:19a]
which is given for you: this do in remembrance of me. [c.f. Lk 22:19b, 1 Cor 11:24b]
And after they had supped [c.f. Lk 22:20a, 1 Cor 11:25a]
he took the cup, and gave thanks, and said, Take this, and divide it among yourselves: [Lk 22:17]
this is my blood, the new covenant. [c.f. Lk 22:20b, 1 Cor 11:25b, Mk 14:24a]
For I say unto you, I will not drink of this fruit, until the kingdom of God shall come. [Lk 22:18]
Although ‘given’ is included (suggesting that Lk 22:19b was the source of this part of the text), it is possible that this word is a later interpolation. It is also not possible to tell whether the words “And after they had supped” were taken from Lk 22:20a or 1 Cor 11:25a, as this part of these two verses is identical. In contrast, “this is my blood, the new covenant” appears much closer to Mk 14:24a than to either Lk 22:20b or 1 Cor 11:25b. However, because Lk 22:17 precedes these words the reference to the cup in Lk 22:20b/1 Cor 11:25b is unnecessary, and the inclusion of the word ‘new’ suggests that, as mentioned above, these words have been revised, so this is not conclusive.
Based on the words alone, it is not possible to tell whether 1 Cor 11:24b-25 or a variant of Luke already containing Lk 22:19b-20 was used as the source of these additions, but instead the order of the text provides this information. The basic structure of Sy-S follows that of Mark (Lk 22:19a, 17, 20b, 18), into which edited text from (or very similar to) 1 Cor 11:24b-25a has been interpolated, without moving any of the existing text.
Sy-C and Sy-S both testify to an original text order the same as that in Mark, the Diatessaron, and also as seen by Tertullian, but both also including some text from 1 Corinthians 11. The words used in both mss show that neither are derived directly from Mark, but instead are based on a Lukan text that in the main matched that reported by Tertullian. However, unlike in the Majority Text, the subsequent inclusion of words from 1 Corinthians 11 did not result in a re-arrangement of the text in either mss, thus maintaining the bread-cup order.
In contrast to the Majority text, in which the text from 1 Cor 11 23b-25b was apparently given priority, and the original Lukan text re-ordered to fit around it, in Sy-C and Sy-S the Lukan text has priority, and smaller portions of the text from 1 Cor 11:24b-25b were interpolated directly into it. Both Sy-S and Sy-C group the references to the ‘cup’ and wine together at the end, and neither have any equivalent to the atonement doctrine seen in Lk 22:20c.
Sy-C and Sy-S both include Lk 22:19b (from 1 Cor 11:24b), although Sy-C omits “given.” Sy-C contains nothing else from 1 Corinthians 11, suggesting that whoever wrote it wanted to avoid any connection between wine and Jesus’ blood. Sy-S adds this connection, but still avoids any mention of the blood being shed for anyone. Sy-C and Sy-S are clearly related, although it is very unlikely that one is a direct descendant of the other. Kenyon stated:
The general result (so far as first impressions go) would seem to be that the Curetonian and Sinaitic texts represent two closely allied branches of a common stock, each of them having been somewhat considerably altered in the course of transmission, but altered in different directions.
Sy-C and Sy-S appear to be related both to what Justin Martyr reports and to the Diatessaron, while neither ms shows any specific signs of knowledge of the Majority Text. Of the two the text of Sy-C is earlier, while that of the Peshitta is later than both, and is clearly a variant of the Majority text.
Although the passages in Mark and Matthew are so similar that they cannot be independently created versions of the one in Luke, it is possible (and indeed more likely) that the version in Luke was derived from the Mark/Matthew version, with other sources possibly providing material not in either Mark or Matthew. The question then arises, how and why did the differences occur? Perhaps the most important point is that Mk 14:18-25 and Mt 26:21-29 are so similar and that Lk 22:15-22 is significantly different, although clearly relating the same event (the Last Supper). For example, Mk 14:21 and Mt 26:24 are almost identical, and while Lk 22:22 is similar to Mk 14:21a and Mt 26:24a, it has nothing corresponding to Mk 14:21b and Mt 26:24b. Additionally, as discussed above, Mark and Matthew have nothing corresponding to Lk 22:19b-20a.
Does this mean that for this passage Mark and Matthew used the same source or that one used the text of the other, while aLuke ‘re-wrote’ their version (adding Lk 22:19b-20a), or alternatively that aLuke simply used a source other than Mark or Matthew that included text not in either? Even if aLuke was using a different source here it is highly unlikely to have been Q (a hypothetical gospel that was the source of the Double Tradition text, i.e. that common to Matthew and Luke but not in Mark), as according to the IQP (International Q Project - see The Sayings Gospel Q), in Q the only parallels to Luke 22 were to Lk 22:28, 30.
However, this does not rule out a different source (i.e. with text different to that in Q), whether seen by aMatthew or aLuke, or by both, and one such possible source is Marcion’s gospel (Ev) – on the assumption that it was earlier than Luke. Even if as originally written Luke contained Lk 22:19b-20a, why would these verses then be omitted in the Western text and the Diatesseron, and why would some mss either removed the second ‘cup,’ or alternatively reverse the ‘Bread-cup’ order seen in the /Mark/Matthew version? These questions all make it far from sure how the text of the Last Supper developed, and what sources were used in the creation of the various forms we see today.
Perhaps the biggest barrier to understanding how all the variants of Lk 22:17-20 arose is to view the process as essentially linear, either beginning with the shortest (the Old Latin), with successive interpolations leading to the Majority Text, or beginning with the Majority Text, with successive omissions leading to the Old Latin.
One of the problems with both of these theories is that there is no attempt to suggest why, given the prior existence of Mk 14:22-25 (and possibly Mt 26:26-29), either the shortest (the Old Latin) or the longest (Majority Text) variants would have been the first to appear in Luke. In addition, such is the influence of the extant variants that it is hard for most people to seriously consider the possibility that all of the known variants might have arisen from another non-extant one. The impact of the Majority Text has meant that evidence that provides clues that such a variant did exist, and also suggests what the content of that variant might have been, has, it seems, been consistently overlooked. For example, Willker comments as follows:
That Marcion supports the longer text is a strong argument for its originality. Schürmann makes it probable that also Justin and Tatian attest the longer form.
Herrick makes a similar point
The longer reading is attested by the following: 1) all the Greek manuscripts, including p75 (AD 175/225); 2) all the versions with the exception of the Old Syriac and part of the itala and 3) by all early Christian writers beginning with Marcion, Justin and Tatian.
Unfortunately, these comments relating to Marcion, Justin, and Tatian are based on a misinterpretation of the evidence. While it is true that Tertullian, Justin, and the Diatessaron testify to the existence of variants of Lk 22:17-20 that contain text not present in the short Old Latin variants, there is no basis on which to say that these must therefore have contained all of the Majority Text. As seen above, Tertullian’s comments can only be taken to refer to Lk 22:19a, 17, 20b, 18 at most, and Justin appears to refer to Lk 22:19a,b,d,c, 20a, 17, while what we know of the original text of the Diatessaron covers only Lk 22:19a, 17, 20b-c, 18.
Not only was Lk 22:19b-20a not present in any of these three variants (the second half of Lk 22:19b that follows Lk 22:18 in the Arabic and Latin translations of the Diatessaron is a later interpolation), but their text order does not match that of the Majority Text (Lk 17, 18, 19, 20). However, it is a close match to what we see in Mark, the Lukan equivalent of which is Lk 22:19a, 17, 20b-c, 18, and this is therefore the most probable basic form of the earliest variant of this passage. Once it is realized that the earliest variant of Lk 22:17-20 was not much more than an edited copy of the text we see today in Mk 14:22-25 and Mt 26:26-29 (i.e. in the order Lk 22:19a, 17, 20b-c, 18, and without what we know as Lk 22:19b-20a), then all the other known variants can be seen to have resulted from just two basic changes to this original text:
Either partially or completely removing the references to blood and Jesus’ atoning sacrifice in Lk 22:20b-c;
Combining the initial Mark/Matthew text with 1 Cor 11:24b-25 (in two different ways), and removing any duplication.
These changes are summarized below:
The initial Greek variant of this passage was based on the text seen in Mark and had the bread-cup sequence. It was seen by Tertullian in Marcion’s gospel [Ev], and based on his lack of comment it was also very similar to what Tertullian saw in his own Old Latin bible. He does not mention the text of Lk 22:20c (or it’s Mark/Matthew equivalent), and so his copy of Luke may have omitted this reference to Jesus’ blood being shed. This variant was also used by Tatian when creating the Diatessaron.
Old Latin mss b and e go further than Tertullian, not having the whole of Lk 22:20b-c (originally Mk 14:24/Mt 26:28), thus not having any mention of Jesus’ blood.
Text from 1 Cor 11:24-25 was interpolated into the initial Greek variant. To avoid problems with duplications and the order of the text, what we now know as Lk 22:17-18 were placed before the rest of the text, thus creating the Majority Text variant with its cup-bread-cup sequence.
Sy-P (Peshitta) is based on the Majority Text, but avoids the first ‘cup’ by simply omitting Lk 22:17-18.
Justin’s apology has the bread-cup order, has nothing from Lk 22:18, appears to rely on Mk 14:23-24 rather than Lk 22:17 (so not following the Majority Text) and does not appear to have used 1 Cor 11.
Sy-S and Sy-C have their origins in the initial Greek variant and also include portions of text from First Corinthians 11, but avoid the first cup without re-ordering any of the initial text. Both omit Lk 22:20c, and in addition:
Sy-S surrounds Lk 22:17 with shortened versions of 1 Cor 11:25a and 25b, with Lk 22:18 following;
Sy-C has the same placing of Lk 22:17-18 as Sy-S, but omits 1 Cor 11:25 entirely.
Later old Latin mss (a, d, ff2, i, and l), and also D, are influenced by the order of the Majority Text, and swap the early versions (based on the Mark/Matthew text) of Lk 22:19a with Lk 22:17-18, giving the cup-bread sequence, while not including any text from First Corinthians 11.
Aland, Kurt and Aland, Barbara: The Text Of The New Testament: An Introduction to the Critical Editions and to the Theory and Practice of Modern Textual Criticism, p. 311
Billings, Bradley S: Do this in Remembrance of Me: The Disputed Works in the Lukan Institution Narrative (Luke 22.19b-20): An Historical–Exegetical, Theological and Socialogical Analysis, 2006
Burkitt, F. Crawford: Evangelion da-Mepharreshe - the Curetonian Version of the four gospels, with the readings of the Sinai palimpsest and the early Syriac patristic evidence edited, collected and arranged by F. Crawford Burkitt, 1904. p 301.
Chadwick, Henry: The Shorter Text of Luke XXII. 15–20. Harvard Theological Review, 50, 1957, pp 249-258
Colwell, Ernest C: Hort Redivivus: A Plea and a Program: Studies in Methodology in Textual Criticism of the New Testament, XI, 1967, pp 148-172
Crossan, John Dominic: The Historical Jesus: The Life of a Mediterranean Jewish Peasant, 1992, p. 364
Ehrman, Bart D: The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament, 1993, pp198-205
Hemphill, the Rev. Samuel, B.D: The Diatessaron of Tatian: A Harmony of the Four Holy Gospels Compiled in the Third Quarter of the Second Century, 1888
Herrick, Greg: The Atonement in Lucan Theology in Recent Discussion. A Study
Hobson, Alphonzo Augustus: The Diatessaron of Tatian and the Synoptic Problem: 1904
Holding, James Patrick. Ed: Defending the Resurrection, Did Jesus Rise from the Dead?, 2010,
Justin Martyr: The First Apology, Dialogue with Trypho (Chapter 10)
Kannaday, Wayne Campbell: Apologetic Discourse and the Scribal Tradition: Evidence of the Influence of Apologetic Interests on the Text of the Canonical Gospels, 2004, pp155-162
Kenyon, Sir Frederic George: Our Bible and the ancient manuscripts, 1897, p 156
Lietzmann, Hans & Richardson, Robert Douglas: Mass and Lord's Supper: A Study in the History of the Liturgy
Mack, Burton: Who Wrote the New Testament?, pp. 240-241
McDaniel, Thomas F, PhD: Miscellaneous Biblical Studies, Chapter Ten, Recovering Jesus’ Words By Which He Initiated The Eucharist, 2009
McGowan, Andrew: Rethinking Eucharistic Origins
Metzger, Bruce M: A Textual Commentary on the Greek New Testament. 2nd ed. United Bible Societies, 1994.
Rice, George Edward: The Alteration of Luke's Tradition by the Textual Variants in Codex Bezae, 1974, pp 223-230
Scrivener, F. H. A.: A plain introduction to the criticism of the New Testament, for the use of biblical students, 1874, pp 519-521
Shanks, Hershel: Understanding the Dead Sea Scrolls, p195
Tabor, James D: Eat My Body, Drink My Blood–Did Jesus Ever Really Say This?, TaborBlog, 2015
Utley, Bob: Luke 22 Study
Waltz, Robert: The Encyclopedia of New Testament Textual Criticism
Willker, Wieland: A Textual Commentary on the Greek Gospels Vol. 3 Luke
Wright, W, PH.D., LL.D: The Homilies of Aphraates, the Persian Sage Edited From Syriac Manuscripts of the Fifth and Sixth Centuries, with an English Translation, 1869, pp 300-302