Are the Canonical Gospels Cento Compositions Developed from Secret Mark and Other Sources?

As we noted in a previous post, one of the most interesting parts of Morton Smith's 1973 book Clement of Alexandria and a Secret Gospel of Mark is where it grapples with Robert Grant's assertion that 'Secret Mark' is a cento. A cento is a literary composition which basically looks like a 'patchwork' (= Lat. cento) of lines from other authors. Morton Smith spends a number of pages denying Grant's thesis (which must have been reached Smith's ears through personal contact with Grant). Yet Smith's arguments, as FF Bruce notes, are ultimately unconvincing.

Indeed it might be worth citing from Bruce's 1974 EM Wood Lecture (FF Bruce, 'The 'Secret' Gospel of Mark', London 1974) because many of the same arguments would ultimately be used by Francis Watson to argue that the text was a modern forgery (a view F F Bruce did not subscribe to). Instead Bruce, like Grant, saw the work as an ancient 'pastiche' - or a cento - which Bruce argues wasn't written by Mark himself but some later figure and ultimately accepted by Clement of Alexandria:

The fact that the expansion is such an obvious pastiche, with its internal contradiction and confusion, indicates that it is a thoroughly artificial composition, quite out of keeping with Mark’s quality as a story-teller. Morton Smith indeed argues that it is no mere pastiche or cento, but I find his arguments unconvincing. That the letter-writer was disposed to acknowledge it as part of a fuller edition of Mark’s Gospel, written by the evangelist himself, is quite in line with evidence which we have of Clement’s credulity in face of apocryphal material. He treats the work entitled the Preaching of Peter as a genuine composition of the apostle Peter (Strom. ii. 15.68; vi. 5.39ff., etc.) and he similarly accepted the authenticity of the Apocalypse of Peter (Eusebius, Hist. Eccl. vi. 14.1). And we shall see how readily he acknowledged as dominical sayings ascribed to Jesus in the Gospel of the Hebrews and the Gospel of the Egyptians, explaining them in terms of his own philosophy

Of course I do not agree with Bruce's assessment but he is quite correct in pointing out that Smith's rebuttal of the parallels between the canonical gospels and 'Secret Mark' is surprisingly weak.  I think he is correct that Secret Mark is not a pastiche, yet the obvious parallels necessitates that the canonical gospels are the centos and the Alexandrian text at least one of their literary sources.  Instead of tackling this issue Smith cites from Grant's 1960 study and some correspondences between the two men and by failing to address Grant's objections convincingly opens the door for lazy scholars to go along with Watson's claims that because Secret Mark is a pastiche it must be a forgery (!)

I will cite at length from Smith's discussion of Grant's correspondences and his 1960 book and then the pertinent section from the Secret Sayings of Jesus will follow.  First, from Smith's 1973 book:

The high frequency of parallels in the longer text affords support for a special theory of imitation which has been suggested independently by P. Benoit and R. Grant, viz.: The longer text is a cento produced from the text of the canonical Gospels. Grant supports this theory by a reference to Irenaeus (Harvey, I.1.15 - 20 = Stieren I.8.1 - 9.5). Irenaeus is there attacking the Valentinians. He says that, since the they have a theory which neither the prophets proclaimed nor the Lord taught nor the apostles handed down, but which they read out of agrapha (that is, uncanonical works, Harvey), they try to twist the dominical, prophetic, or apostolic sayings to fit their teachings, so as to have some evidence for what they say, and to this end they neglect the order and context of the scriptural passages they use and also distort them. He compares their treatment of Scripture to the breaking up of a mosaic in order to make a different picture with the same tessarae. The examples he gives to illustrate this, however are examples of allegorical or esoteric exegesis of individual sayings or passages of the canonical Scriptures and afford no evidence for the composition of new, pseudo-Scriptural centos. However, he goes on to say (Harvey, I.1.20, middle = Stieren, I.9.4): "Then, collecting scattered expressions and terms, they transfer them, as we said, from the in reality to an unreal much as do the Homeric poems, so that less experienced readers might think Homer had composed the verses, since both are attempting a similar and, indeed, identical feat." And he concludes that, as the man acquainted with Homer will recognize the verses, but not the theme, and by referring to their proper contexts will show the theme to be spurious, thus the true Christians "will recognize the terms from the scriptures and the expressions and the parables, but will not recognize the blasphemous theme." He will acknowledge the tessarae, but not the picture which has been made of them, "and, referring each of the things said to its proper place and fitting it into the body of the truth, he will expose their fiction and show it to be unsubstantial."

On the strength of this passage, Grant has suggested that the longer text may be a gnostic work of the sort attacked by Irenaeus.  However the longer text has no connection with the Valentinians, and though it was used by the Carpocratians it was also used by Clement's church, which is commonly supposed to have been orthodox.  Clement expressly asserts that the Carpocratians got it from the orthodox (that is, from his church), and nothing in the text is clearly gnostic. Therefore there is no reason to associate the text with the Valentinian centos, unless it can be shown to be a cento, which is the point in question. Further, the text of Irenaeus does not precisely say that the Valentinians made centos. Irenaeus may have intended to give that impression. [But in the opinion of EB he actually had in mind compositions like the Qumran hymns, which are full of OT echoes but are not true centos. He introduced the bit of cento merely to give his Greek readers the best example he could of the sort of thing he had in mind.] At all events he does not explicitly state that the Valentinians made centos, and — what is most important - he does not produce and demolish any Valentinian cento. This suggests that either he had no such document and was merely using the comparison as a reductio ad absurdum of their neglect of context in exegesis, or he had a Valentinian Gospel which paralleled the canonical Gospels in many places and which he wished to discredit, so he charged that it was a cento but did not give an example from it for fear of discrediting his charge.

On the other hand Irenaeus (Harvey 1. 20. 2 = Stieren, 1. 25. 4) quotes a Carpocratian version of the counsel to be reconciled quickly with one's adversary, which alternately parallels Mt. 5.25 and Lk. 12.58 in a way that can be interpreted as deliberate choice of elements suited for Carpocratian exegesis (Grant-Freedman, 95). And Dodd, New Gospel 24ff, has practically proved that the text on fragment 1 verso of P. Egerton 2 is a cento of Jn. 5.39,45 and 9.29. The cento form goes back in Greek tradition at least to Aristophanes, Pax 1090- 1094, and appears in the OT with the psalms in Chronicles and Jonah. So the possibility that the longer text was produced as a cento is undeniable. And there is no necessity of connecting the cento form with the gnostics: Paul used it in Rom. 3.11-18; Tatian's Diatessaron, the most famous example of the form, was not a gnostic work; and even Mt. and Lk. could be considered, loosely as centos compiled from Mark, Q, and other sources. So the question of form and method of composition need not be confused by introducing the question of doctrinal affiliation.

In favor of the cento theory is the high frequency of parallelism and particularly the frequency of the long parallels discussed above. Against it, however, are the following facts: (1) Some elements of the longer text are not paralleled from the canonical Gospels, which would be impossible were it a true cento of the canonical texts. (2) The great majority of the parallels are brief formulas, most of them many times in the canonical Gospels and more likely to have been put together freely by an imitator than to have been picked out laboriously from here and there by the compiler of a cento. (3) The text cannot be made up by drawing elements from only two or three stories; to suppose it a cento, one must also suppose that the author derived his scraps from practically every chapter of Mk., to say nothing of the other Gospels - not a likely procedure, especially in antiquity, when most writers, even in citing explicitly, cited from memory. (4) Many details in the text do not look as if they had been produced by the compiler of a cento; see, for examples, the notes on [various words follow] (5) The text is too well constructed and economical to be a cento: there are no irrelevant details, every word comes naturally in its place, the narration moves without delays or jumps. Possibly some leisured litterateur might have might have succeeded in piecing together such a text from the phrases available in the canonical Gospels, but the easier explanation is to suppose it a free composition. (6) The hypothesis that the text is a cento requires the supposition that someone went to great pains to imitate the style of Mk. as closely as possible, since making a cento is the most laborious, but the closest, kind of imitation. But the longer text is datable, by external evidence, before 1 25 (see above, section I of this chapter) and at this date Mk.'s prestige was not high enough to motivate this sort of imitation. Matthew and Luke, in their expansions of Mk. made no attempts to imitate his style. Jn. and the earlier apocryphal gospels (Hebrews, Thomas, Egyptians, Peter) show no considerable effort to imitate synoptic style. [EB remarks that almost none of the early apocryphal Gospels are even attributed to canonical authors.] So to suppose the longer text a cento would be to suppose it a work unparalleled and unlikely in its time. (Morton Smith, Clement of Alexandria and a Secret Gospel of Mark 1973, p. 141 - 143)

There is of course more to Smith's argument against the pastiche in the 1973 book but it is my contention again that they too prove to be ultimately unconvincing. We have emboldened the argument that I think Smith should have went with - namely that it was the canonical gospels which were arranged as artificial centos by the orthodox. I will continue to talk about this in future posts but I have laid out at least some argument to at least demonstrate that Irenaeus had the skills to pull off such a feat.

I will close this post with a citation from Grant's 1960 work referenced by Smith above. I think there are a number of problems with Grant's logic. He completely ignores the possibility that the orthodox gospels could well be centos. Nevertheless he does give a fascinating overview of the cento-parallels with the canonical gospels which I think will be useful for my readership:

If the Gospel of Thomas reflects the atmosphere of the Gnostic sects of the second and third centuries, we may well ask what use these Gnostic sects made of the gospels recognized by the Church. Is there any correlation between the ways in which Gnostics used them and the ways in which gospel-type sayings are transmitted in Thomas? It is fairly certain that the methods employed are approximately the same. Sometimes the Gnostics combined materials derived from various gospels simply for the sake of completeness; sometimes they combined them and rearranged them, in order to make their special theological points. Since they believed that they possessed the true understanding of Jesus's teaching, they felt free to restate what he had taught so that its Gnostic significance would become clear.

In the middle of the second century a sect known as the Carpocratians held that "Jesus spoke in a mystery [or "secret"] privately to his disciples." What he spoke about was presumably the Carpocratian system; but the Carpocratians were probably justifying their idea by appealing to the Church's gospels. The mention of "mystery" is found in Mark 4:11; and "privately" occurs in Mark 4:34. Irenaeus gives us one extensive example of their exegesis, and of the text they used. First the text.

When you are with your adversary on the way, [Matthew 5:25]

take pains to be freed from him, [Luke 12:58]

lest he give you to the judge and the judge to the servant [Matthew 5:35]

and he cast you in prison [Luke 12:58]

verily I say to you, you will not come out from there until you pay the last quadrant [Matthew 5:26]

This oscillation between the two gospels may be based on either a primitive harmony of the gospels or on a manuscript of either one in which readings from the other were to be found. The Carpocratians' exegesis, however, seems to show that they were consciously choosing from each gospel the items which they found useful. The "adversary" is the devil, one of the angels in this world, who leads souls astray to the "prince," the first of the creator-angels. The "prince" is not mentioned in the verses Irenaeus has quoted, but he is to be found in Luke, not in Matthew. Moreover, the idea of being freed from the adversary (Luke) is very different from that of being reconciled to him (Matthew). The "servant" (Matthew; Luke has the official term proctor) is the servant of the creator-angel, and his service is that of transferring souls from one body to another— the "prison" is the body. "Come out from there" means to escape from the body to the God who is above the creator-angels. "Pay the last quadrant" means to undergo all the humiliation of which the body is capable. (We do not know why the Carpocratians preferred Matthew's "quadrant" to Luke's "lepton"; both are very small coins, but perhaps the quadrant was better known in the Roman world.

This exegesis seems to show (1) that the text the Carpocratians used is not fully reproduced by Irenaeus; but also (2) that it was based on careful selection of items from each of the two gospels they were explaining at this point. Similarly, we know that Marcion, who taught at Rome about the same time, used only the Gospel of Luke and tried to free it from interpolations. In his opinion it was more Jewish than the authentic gospel of Jesus, for the apostles— with the exception of Paul— tried to present the gospel in such a way as to appeal to Jews. They thus corrupted it. Unfortunately, Marcion's critical work seems to have been entirely subjective. There is no evidence that he used any other gospel or traditional source for sayings of Jesus, even though some of his omissions could have arisen as a result of comparing Luke with John. The contemporary Valentinians had their own Gospel of Truth, mentioned by Irenaeus,2 and found among the Nag Hammadi documents; as Irenaeus says, its form is quite different from that of the Church's gospels. In addition, they used the four canonical gospels, not altering the text but changing the meaning, especially in regard to the parables, by explaining that everything Jesus said really referred to Valentinian doctrine. Impressed by the frequency with which groups of three occur in the parables, they usually said that such groups reflected their own view that there were three classes of men: those with spirit, those with soul, and those who were merely material. One disciple of Valentinus, named Ptolemaeus, produced the earliest bit of exegesis we possess for the opening verses of the Gospel of John; another, Heracleon, wrote notes on at least the first eight chapters of the same book. These facts indicate that the Valentinians generally relied on exegesis of the Church's books, rather than on revision of them, in order to make them their own.4 To be sure, die Valentinians claimed that they possessed secret traditions; but they do not seem to have produced new gospels like the canonical ones. In this regard they were like the Essenes, who provided their own exegesis of the Old Testament. The Gospel of Truth does not represent itself as having been written by an apostle. We should conclude that the Valentinians regarded apocryphal gospels with much the same caution as that exhibited by more orthodox Christians. One Valentinian teacher, named Marcus, told an apocryphal story about Jesus's infancy; but such quotations are rare among them.

On the other hand, when we come to the Naassenes (late second century or early third), we find significant examples of Gnostic interweaving.

Unless you drink my blood and eat my flesh



you will not enter into the kingdom of heaven


but if you drink the cup which I drink, where I go, there you cannot enter

Unless you eat the flesh and and eat my flesh drink the blood of the Son of Man (John 6.53); my flesh ... my blood (John 6.54 - 56)

you will not enter into the kingdom of heaven (Matthew 5.20; 18.3)

Can you drink the cup which I drink? (Mark 10.38)
Where I go you cannot come (John 8.21; 13.33)

Here the process of interweaving is obvious; the purpose of the process is not so obvious. According to Hip- polytus, the Naassenes taught that "living things" were "reasons, intelligences, men— pearls which the great undifferentiated. Being has cast into the work here below." They held that men were to eat such living things (see Saying 10), but since they understood this "eating" as a spiritual process, they undoubtedly held that drinking the blood of Jesus and eating his flesh was spiritual too (cf. John 6.63 "The Spirit is what gives life, the flesh is of no avail; the words which I have spoken to you are spirit and life"). But why were Jesus's disciples not to drink the cup which he drank? Perhaps the Naassenes were so much concerned with "spiritual" meanings that they wished to reject the possibility of taking part in a real Eucharist, with real wine and real bread. To drink the cup which Jesus drank would mean surrender to the carnal practices of the "psychic" Church. In any event, it is clear that they themselves created this compilation from the Church's gospels.

The Naassenes' version of the parable of the sower (Matthew 13:18-23; Mark 4:3-9; Luke 8:5-8) 7 is based on a combination of Matthew and Luke, though without any really significant alterations. They explained it as meaning that "these mysteries are understood only by the perfect Gnostics"— an interpretation based on Matthew 13:19, which speaks of those who "hear the message of the kingdom and do not understand it. At one point, however, their exegetical ingenuity was responsible for the combination they made. Matthew speaks of the "fair" soil on which some seed was cast; Luke uses the word "good." The Naassenes found both adjectives in Deuteronomy 31:20, where Moses speaks of the "fair and good land" to which he will lead his people. Since they used the Deuteronomy passage to explain the parable, they used both adjectives in speaking of the sower's soil. Once more, their combinations are due to their theological intentions.

Another quotation apparently reflects a different purpose. The Naassenes quoted the question of Jesus, "Why do you call me good?" from either Mark 10:18 or Luke 18:19 (Matthew 19:17 has a different question). They gave his answer, "One is good," from Matthew 19:17, however, since the expressions used in Mark and Luke are different. All three synoptic gospels continue with an admonition to obey the commandments of God. The Naassenes, on the other hand, turned to Matthew 5:45, which urges men to become sons of "your Father who is in heaven," who "makes his sun rise on the evil and the good and makes it rain on the just and the unjust. They changed "your Father" to "my Father" and altered the saying so that the sun would rise on the just and the unjust and the rain would fall on the pious and on sinners. What they were doing was, apparently, simply changing the gospel sequence in order to suggest that their version was more reliable than that of the gospels. They also removed a command to obey the injunctions of the Decalogue.

Another kind of modification is brought out in their rearrangement of Matthew 7:13-14. The gospel mentions a narrow gate; then it speaks of the broad road to destruction; finally it refers to the narrow entrance to life. Partly in order to smooth out the sequence of thought, but also partly to emphasize the narrowness of the way to life, the Naassenes reversed the order of the last two parts of the saying. Similarly, when they quoted John 4:21-24 they rearranged the order of the verses. "God is a Spirit [24]; therefore the true worshipers [23] will worship neither on this mountain nor in Jerusalem [21] but in spirit [23]. The sense of the passage in John is not significantly changed, but the new order leads more precisely to the Naassene conclusion: "For the worship of the Perfect is spiritual, not fleshly." A good example of their work of synthesis is provided in their combination of two gospel sayings. "A lamp is not laid under a bushel but on a lamp stand; we proclaim the proclamation on the rooftops, on all roads and all streets, and even in houses." The saying about the bushel and the lamp stand is found in Matthew 5:15 and Luke 11:33; the word "light" occurs in the preceding verse in Matthew. But "light" is also mentioned in Matthew 10:27/ Luke 12:3; and in both gospels Jesus goes on to say, "What you hear in your ear proclaim [or "will be proclaimed"] on the roofs." The sayings have been combined as the result of a kind of verbal association, though theological interpretation is also a factor.

This example is important because the same combination, though in reverse order, is to be found in Saying 34 in Thomas. "What you hear in your ear preach to another ear upon your roofs. for no one lights a lamp and puts it under a bushel ... but he puts it on the lamp stand." We know that the Naassenes were fond of such combinations, because Hippolytus describes one rather elaborate one. It began with the story of Jesus's transforming water into wine and thus manifesting "the kingdom of heaven" (John 2:1 - 11 though John said that he manifested 'his glory'). The Naassenes claimed that this passage showed that the kingdom is "within you" (Luke 17:21), and they added that the same doctrine was expressed in the parables which compared the kingdom with a treasure (Matthew 13:44) and with leaven in three measures of meal (Matthew 13:33). It is worth noting that, though Thomas contains no miracle stories, in it we find the kingdom within (Saying 2), the parable of the treasure (106), and the parable of the leaven (93) . In Naassene transmission and exegesis of the sayings of Jesus there is a tendency to combine words found in different contexts in order to bring out their Gnostic meaning.

We also find a tendency to rearrange the sequences found in the gospels, even within individual sayings. To some extent, this rearrangement may be due simply to the shortcomings of memory. It is also probably to be ascribed to the Naassenes' desire to show that they possess the true understanding of the sayings, an understanding different from that possessed by ordinary Christians. [Robert Grant the Secret Sayings of Jesus 1960 p. 94 - 101]