PAPYRUS EGERTON 2 AND

THE FOURTH GOSPEL *

John W. Pryor

In his Presidential Address to the SNTS in August, 1986,1 Professor

Raymond Brown noted a trend in recent scholarly research: the challenging
of the canonical gospels by appeal to other extant documents. In particu-
lar, Brown entered into a critical review of the work of J. D. Crossan,
Four Other Gospels,2 concentrating on the latter's claim that "Peter"
represents tradition earlier than that found in the canonical gospels. With
considerable audacity, one more Presidential Address is going to enter the
debate over the relation between the canonical gospels and other written
traditions.
In 1935 H. I. Bell and T. C. Skeat published critical notes on three
fragmentary papyrus leaves of a Greek codex which contained sayings,
miracle stories and controversies of Jesus. 3 Known as Papyrus Egerton 2
(or Unknown Gospel = UG), their exact provenance is unknown, though
they were discovered in Egypt. 4 At first, UG provoked a lot of scholarly
interest, including a doctoral dissertation by Goro Mayeda. 5 Thereafter,
interest subsided somewhat until recent years, when a number of scholars
have revived interest in the document and its relation to the canonical
gospels.6

*The Presidential Address delivered to the Fellowship for Biblical

Studies, Melbourne, 1988.
l"The Gospel of Peter and Canonical Gospel Authority", NTS 33 (1987)
321-343.
2J. D. Crossan, Four Other Gospels (Minneapolis: Winston, 1985).
Crossan has followed up this work with a subsequent study, The Cross that
Spoke (San Francisco: Harper and Row, 1988) in which he claims that the
passion and resurrection narratives of the canonical gospels are radical
revisions of an earlier gospel account, reflected in the Gospel of Peter.
3H. 1. Bell and T. C. Skeat, Fragments of an Unknown Gospel (London:
British Museum, 1935).
4That is, they were bought in Egypt in 1934 from a dealer. Their exact
origin is thus subject to speculation.
5Goro Mayeda, Das Leben-lesu-Fragment Papyrus Egerton 2 und seine
Stellung in der urchristlichen Literaturgeschichte (Bern: Haupt, 1946).
6H. Koester, "Apocryphal and Canonical Gospels", HTR 73 (1980) 119-
126; D. F. Wright, "Apocryphal Gospels: the 'Unknown Gospel' (Pap.
2 A USTRALIAN BIBLICAL REVIEW 37/1989

In this paper, our particular interest is in the relationship between VG

and the Fourth Gospel, for which, over the years, three explanations have
been presented.
(i) The 4G used VG as one of its sources. 7 The proponents of this
position have argued their case on the basis of stylistic criteria: the
Johannine elements appear to be more naturally located in VG than in 4G.
I shall take up this matter later. It needs to be borne in mind that accep-
tance of such a position would not demand a dating of 4G much later than
100 AD, for VG, which Bell and Skeat dated "not very far from the middle
of the second century", 8 appears to be but a copy of an earlier text. Dodd
demonstrates the possibility of this when he discusses the address
"didaskale Jesou" (lines 33,45).9 The term is an attempt to express the
concept "Rabbi Jesus", out of recognition that Jesus was known as a
rabbi. But "Rabbi Jesus" does not correspond with contemporary Jewish
usage. However, after 70 AD eminent teachers were referred to as "R.
Jochanan", "R. Eliezer" etc., but not as a form of address, which remained
simply "Rabbi". Egerton would therefore appear to have arisen some
time after the fall of the Temple in a circle unfamiliar with the precise
details of Jewish custom. Further confirmation is offered by Bell and
Skeat who suggested that the lack of doctrinal bias and the use of both
"Jesus" and "Lord" in the narrative, indicate a date 80-120 AD for the
original composition. lo
(ii) VG was written independently of any of the canonical gospels.
Incidents and terminology common to them lead back to independent use
of oral traditions. I I

Egerton 2) and the Gospel of Peter", in D. Wenham (ed) Gospel

Perspectives: the Jesus Tradition outside the Gospels (Sheffield: JSOT,
1984) 207-232; F. Neirynck, "Papyrus Egerton 2 and the Healing of the
Leper", ETL 61 (1985) 153-160.
7So Bell and Skeat 34-38; and Koester. It is noteworthy that Koester
has chosen to ignore completely Dodd's earlier article which presents
forceful arguments to the contrary.
8H. 1. Bell and T. C. Skeat, The New Gospel Fragments (London: British
Museum, 1935) 10.
9C. H. Dodd, "A New Gospel", in New Testament Studies (Manchester:
Manchester Univ. Press, 1953) 21. The article originally appeared in BJRL
in 1936.
I~ell and Skeat, New Gospel Fragments 19.
11 Goro Mayeda; also R. Cameron, The Other Gospels (Philadelphia:
Westminster, 1982) 73.
 Pryor: Papyrus Egerton 2 3

(iil) VG was written in dependence on the 4G, quoting it either from

memory or, less likely, copying from it,12
Before we can look at John and VG, we need first to clarify the
synoptics-VG relationship.

A. VG AND THE SYNOPTICS

VG has similarities to two synoptic incidents, lines 32-41 to the
cleansing of the leper in Mark 1:40-44 et par., and lines 43-59 to the
tribute money incident in Mark 12: 13-17 et par. Though it is not my
intention to undertake a detailed study of the pericopes, the following
comments are in order:
1. Frag 1 recto has elements of two pericopes, the break occurring at
line 31. Lines 22-31 tell of an attempt by certain people (perhaps
archontes is the correct restoration of line 25) to seize Jesus. The
language is thoroughly Johannine. 13 I would suggest that these lines re-
present the conclusion to the dispute between Jesus and the archontes tou
/aou (116) in Frag 1 verso. That is to say, the fragment which we have
and which gives evidence of 21 lines of text on the verso side is the lower
part of the leaf with perhaps only one or two lines missing. 14 A new
pericope begins on line 32 with kai idou /epros prose/than auto Jegei.
This construction is distinctive of the synoptics and is never found in
John. However, in the synoptics it begins a pericope very rarely (Matt
19: 16, Luke 2:25, 10:25). Indeed, it can plausibly be argued that in the
first two examples the kai idou serves to link the verse with what precedes
and was never the start of an isolated pericope. 15 Normally kai idou is
preceded by an introductory statement which sets the scene of the action. 16
This would indicate that whatever the source of VG, synoptic tradition or
independent tradition, the redactor of the text has himself deliberately
linked the start of the incident with the movement of Jesus at the end of
the previous one.
What all of this editorial activity points to is that in spite of the
strongly Johannine "flavour" of lines 1-31, the author of VG is writing a

12F.-M. Braun, Jean le Theologien (3 vols; Paris: Gabalda, 1959-66) I.

91-94; H. I. Bell, Recent Discoveries of Biblical Papyri (Oxford, 1937)
17-overtuming his earlier judgement; J. Jeremias, Unknown Sayings of
Jesus (London: SPCK, 1957) 18-20, 93-94; Dodd, "A New Gospel".
13See Dodd 27-31 for a precise demonstration of this point. See also
the later discussion of lines 22-31 in this paper.
14 Crossan, 70, agrees with this conjecture.
15Note that Matt 19:16 is very similar to VG 32.
16Matt 9:2,20; 12:10; 17:3; 28:2,9; Luke 5:12; 7:12,37; 9:38; 13:11;
14:2; 19:2.
4 AUSTRALIAN BmLICAL REVIEW 37/1989

gospel much more in the style of the synoptics: relatively short pericopes
consisting of conflict stories, healings and the like, at least some of
which were carefully knit together into a flowing narrative. We lack the
evidence to decide whether VG has any plot development, though lines
28-29 may point to this.
2. Lines 32-41 contain the story of an encounter between a leper and
Jesus. There certainly are features which resemble the narrative in Mark
1:40-'~4 et par. Some of the parallels are exact: ean theles is common to
the synoptics, and VG (which adds su), as is thelo katharistheti. Other
phrases and words in the narrative agree with only one of the synoptics, or
are very similar Thus, while lines 39-41 have a synoptic "flavour", they
in fact agree with none of the accounts (though see Luke 17: 14 in the ten
lepers incident). The language of the pericope shows affinity with no one
of the synoptic accounts. 17
In fact, the differences from the synoptics are more noteworthy than
the points of contact. The leper at no time pays homage to Jesus, a factor
present in all synoptic accounts (Mark 1:40 et par., Luke 17: 16); there is
no reference to Jesus' extending his hand to touch the leper; Jesus is
addressed as didaskale Iesou, a title distinctive to VG; the account tells of
the man's social contact with other lepers, the cause of his contracting the
disease. We cannot tell whether the charge to keep silence, which in the
synoptics precedes the dismissal to the priests, is lacking from VG, as
Dodd believes, or was present in the missing lines.
It is hard to agree with Wright (supported independently by F.
Neirynck I8 ), that these differences from the synoptics do not argue for
VG's independence from them. 19 As noted, the narrative is verbally closer
to none of the synoptics in preference to the others, and all of the synop-
tic similarities are such as could be expected to exist even in divergent
traditions. The core of the incident is the man's expectation and Jesus'

I70odd, 33-34, has listed the synoptic similarities.

I8In a recent article, "Papyrus Egerton 2 and the Healing of the Leper",
ETL 61 (1985) 153-160, F. Neirynck debates a recent claim by M.-E.
Boismard (see his Synopse, vol. 2, 101-105) that apart from lines 33-36 (=
proto-Matthean additions), the Egerton account is more archaic than the
synoptic story of the healing of the leper (Mark 1:40-45 par.). Neirynck
has no trouble in demonstrating a common Lukan vocabulary between lines
33-36 and the rest of the story, which leaves no real justification for the
Boismard theory. But in the process, Neirynck is less convincing in
arguing that the author "probably had some acquaintance with the three
synoptic gospels and almost certainly with Luke" (159). His arguments are
not Rersuasive and do not overthrow the case here presented.
9Wright 217.
 Pryor: Papyrus Egerton 2 5

response, and this is where the accounts are closest. It is inherently more
likely that here we are confronted with independent living tradition.
3. The tribute money incident (lines 43-59) has been evaluated
differently by several scholars. Crossan suggests that it puts us in touch
with a tradition more primitive than Mark. Thus, whereas VG has a
quote from Isa 29:13 as part of Jesus' accusation against his inquirers, a
quote which he applies to himself, Mark has relocated it to 7:6-7 as part
of his treatment about Pharisaic human tradition. And here Mark has the
quote referring to God. 2o But Crossan's reasoning here is particularly
curious. The introductory "Well did Isaiah prophesy ... " is found only
here in the gospel tradition (Acts 28:25b is dependent on Mark, as is the
Matthean parallel), and this is a sign that Mark has taken it from VG!
Moreover, while both VG and Mark omit from Isaiah the phrase "with
their mouth", in VG's case this is because in lines 52-53 he makes
allusion to it in Jesus' words. Mark, however, has no good cause for
omitting the phrase, and this also demonstrates he is dependent on VG!
The force of this logic escapes me completely. I would have thought it
more likely that with Isa 29:13 originally referring to Yahweh as speaker,
the context in Mark 7 is more suitably pre-Easter, and that in VG more
suitably post-Easter in location. It is after Easter that the church begins
the practice of applying to Christ sayings of God in the OT. As to
Crossan's second point, it simply does not follow that the absence of the
phrase in Mark's quote supports his case. This presupposes that early
Christianity was normally faithful to the text of the OT when quoting it,
a presumption that is demonstrably false. Furthermore, as Wright asks,
can Crossan say that "well did ... " is characteristic of VG since it is found
there only once also? And may not Mark 7:37, 12:28,32 be partial
parallels to the phrase?21
An opposite view is held by David Wright, namely that there is an a
priori probability that VG is dependent entirely on elements in the
canonical gospels, a probability he is inclined strongly to favour.22 The
main problem is that elements in the pericope are scattered in the gospels,
as follows:
lines 45-47: John 3:2
48-50: Mark 12:14 et par.

20Crossan 83-85.
21 D. F. Wright, "Four Other Gospels: review article", Themelios 12
(1987) 57. Wright further comments that Crossan has failed to notice the
omission of "with/in their mouth" from the LXX of Codices Aleph and A,
thus indicating a textual variant which may have influenced Mark's
tradition. But this is not a strong point, as the tradition in Aleph and A
may have itself been influenced by the gospel citation.
22Wright, "Apocryphal Gospels", 217-219.
6 A USTRALIAN BIBLICAL REVIEW 37/1989

50-51: cf. Mark 1:43-embrimesamenos in the leper incident

52-54: Luke 6:46
54-59: Mark 7:6 (= Isa 29: 13)
It is Wright's belief that this collection of scattered verses in no way
weakens the case for canonical dependence. What he claims is that they
are associated by link words triggered in the memory of the author of VG
as he recollected the gospel stories.
For example, the introduction to the question didaskale oidamen
hoti (Mark 12: 14 par.) is similar enough to John 3:2 (rabbi
oidamen hoti .. didaskalos) to explain how the latter came to be
prefaced to VG's version of the tribute money exchange. (Would it
be far-fetched to remark on the similarity between what follows
hoti-i.e., alethes in Matthew and Mark, apo theou eleluthas in
John? Note that apo theou is reconstructed in VG line 45.) As the
pericope develops, didaskalon recurs, stomati is picked up by
cheilesin (some MSS of Is 29: 13 LXX and Matt 15:8 have both
stomati and cheilesin), and perhaps epropheteusen recalls prophetas
in line 47. 23
While the case for catchword connections is not implausible, it is
weakened by the following observations:
(i) The supposed parallel which Wright draws with lines 1-31, where
J ohannine verses appear to be so connected, fails to reckon with the
difference that in the latter case the quotations from 4G are far more exact.
As we shall suggest below, VG's use of 4G can hardly be ascribed to
vague reminiscence alone, but seems to demand quotation from the text
itself. Such can hardly be the case for the tribute money incident, where
VG's allegiance to synoptic traditions is not at all strong. Nor are the
recollections from 4G in lines 1-31 as random as they appear to be in 43-
59.
(ii) Catchword association cannot explain every case. Thus, Jesus'
question in 52-54, parallel to Luke 6:46 can hardly be explained as derived
from Luke to link with the ascription didaskale Iesou, for in Luke Jesus
uses kurie. Nor is there any obvious reason to change from the Lukan
"do what I say" to something ("hear"?) else, especially since pOiein has
already been used in line 46, thus providing a second naturallinkword. I
believe it is equally likely that at some stage in the oral tradition prior to
the writing of VG, the question and Isaiah quote response of Jesus were
added to the tribute money confrontation. All of this happened under the
influence of the favoured ascription of Jesus as didaskalos Iesous. These

23Wright, "Apocryphal Gospels", 218.

 Pryor: Papyrus Egerton 2 7

lines were added to the pericope before the (missing) answer of Jesus to
the question.
Other factors support the contention that this incident is independent
of the gospels. There seems no reason for the change from Kaisari to
basileusin. Supporting the case for independent tradition is the comment
of Crossan: "basileus was often used for the Roman Emperor in the
eastern provinces so this could easily be a specific reference to Roman
imperial taxation even without the name Caesar being used".24 Indeed,
apart from the repetition in the question, there is little similarity to the
Markan (or synoptic) version of the question. Again, Dodd's listing of
vocabulary shows that "nothing but the barest minimum of words is
common to the canonical Gospels and the papyrus"25. Finally, Wright's
claim that Jesus' words in lines 52-54 look odd in that Jesus has not yet
given any teaching to obey or disobey26 does not carry much weight for
we do not have much of what went before in the codex. It is not at all
impossible that the author (or even the tradition) could be thinking of
teaching of Jesus already given (or presumed to have been given).
4. Finally, we need to take note of Dodd's detailed analysis of the
vocabulary of VG in relation to that of the canonical gospels. 27 His
analysis revealed that VG has "a much closer affinity with the Lucan
writings than with the Gospels according to Matthew, Mark and John".
This unassailable conclusion must be evaluated alongside two other
considerations: (i) the narratives in VG do not bear any close resemblance
to Luke's incidents; (ii) Luke's style and vocabulary are the most
"literary" of all four gospels. All these factors would suggest that the
literary affinities between VG and Luke are purely co-incidental, the result
of the closeness of cultural background between the two writers rather
than of direct literary relationship.
I would conclude this brief study by affirming the position of Dodd
and Mayeda: 28 there is no established literary relation between VG and the
synoptic gospels. On the contrary, the evidence suggests that VG has
drawn on independent oral traditions which bear some resemblances to
known synoptic accounts. I would have no confidence at all in the claim

24 Crossan 79.
25Dodd 37.
26Wright, "Apocryphal Gospels", 217, repeated in "Four Other
Gospels", 57.
27Dodd 37.
28 Mayeda's work was not available to me, but it was reviewed at length
by H. I. Bell in HTR 42 (1949) 53-63.
8 AUSTRALIAN BIBLICAL REVIEW 37/1989

of Jeremias that the "author (of VG) knew all and every of the canonical
Gospels" .29

B. UG AND THE GOSPEL OF JOHN

Frag 1 verso and a segment of Frag 1 recto contain lines which are
thoroughly Johannine in character. The opening lines of the fragment are
too obscure and fractured to be reconstructed with any certainty, but what
is known bears no relation to any part of 4G. At line 5 the text becomes
more secure: an editorial comment paves the way for words of Jesus
which are close to words in 4G. We can make the following observations
about lines 1-20.
(i) In editorial comments, VG has no close contact with 4G. Thus,
lines 5-7 are only superficially Johannine: tou laou is never used after
archontes; ton logon touton is found in John (6:60; 7:36; 10:19; 19:8;
21:23, but it is also found in Matthew (x 3), Mark (x 1) and Luke (x 3);
and the construction strapheis + indic. verb is more Lukan (x 8) than
Johannine (x 2).
(ii) Lines 7-10 are very close to 5:39. The closeness is even more
marked when one recognizes that in the fragment there is a conspicuous
space after echein. The translation thus becomes: "Search the scriptures,
the ones in which you think you have life. They it is which bear witness
to me". This corresponds to the variant Western text as represented by a,
band Syrcu.30 Even the absence of "eternal" is paralleled in one element
of the doublet in a.
(iii) The second saying has two differences to 5:45. Instead of
kategoreso VG has elthon kategoresai, and it has added mou after patera.
The second is of no consequence. The fIrst variation is not a J ohannine
construction but is common in the synoptic tradition and presumably in
the oral traditions. However, erchomai is a very common verb in 4G (x
155) and many of these occurrences bear the theological sense of the
coming of Jesus/the Son for his divine mission. It is not at all unlikely
that the author of VG has modified the text of John 5:45 under the
(unconscious?) influence of both these factors.
(iv) The third quote, derived from John 9:29, is modified to suit the
direct address from the rulers to Jesus. It also has replaced hemeis before

29In E. Hennecke (ed), New Testament Apocrypha (2 vols; London:

Lutterworth, 1963-65) I.95.
30The text of a reads "scrutate scripturas in quibus vos existimatis in
illis vitam aetemam habere illae sunt quae testimonium dicunt de me in
quibus putatis vos vitam habere hae sunt quae de me testificantur". See the
comment on this doublet in Bell's review of Mayeda, HTR 42 (1949) 55-
56.
 Pryor: Papyrus Egerton 2 9

oidamen with (presumably) suo Neither of these changes is of great

significance when weighed along with the exact verbal sequence followed
by VG.
(v) Lines 18-19, words of Jesus, are not the same as any saying in
4G, but they may be an attempt to summarize the sense of John 5:46-47
with their reference to lack of belief in the testimony of Moses. 31
How shall we evaluate this evidence? I believe that the closeness of
word order on three occasions eliminates as an option the possibility that
4G and VG are independently using common oral tradition. 32 Either John
is dependent on VG or the reverse is the case. The issue will be resolved
by reflecting on two questions: whether the language of the fragment is
more characteristic of John or of VG, and whether the sequence of verses
fits more naturally in VG than in their separated contexts in 4G. As to
the first criterion, the more objective one, I believe Dodd has demonstrated
conclusively that the vocabulary and style are thoroughly Johannine but
they are not characteristic of the rest of VG.33 Dodd's case has been
ignored by those who, like Koester and Crossan, would argue the opposite
conclusion, and even by Mayeda who opts for independence. But aspects
of style and language in Frag 1 are characteristic of John, for example
marruroun, ekeinos estin + substantive participle, estin + substantive
participle. One would have to suppose that the author of 4G has so
imbibed the style of this tradition that it has become his own. The
second criterion is admittedly more subjective, and here again Dodd,
Wright, Braun, and Bell (in 1937) line up against earlier Bell and Skeat
(1935), Mayeda, Koester, and Crossan. Again, however, I cannot but feel
that the sequence in VG is not as natural as we at first feel, and that had
we no knowledge of the discourses in 4G we would be somewhat puzzled.
In the first saying, Jesus exhorts his antagonists to search the scriptures
to find divine witness to his own person and actions. But then, in line
10, Moses (i.e. the Torah) is appealed to for a quite different purpose: to
condemn the rulers. The third verse is even more removed from the flow
of the "argument", for the rulers' response does not quite make sense.
The question of Jesus' origins have not been the subject of debate, nor has
the authority of Moses been questioned, and yet both of these are implied
in lines 15-17. In John 5 and 9 the verses do fit their context. It very
much looks as though the author of VG, in constructing this conflict

31S 0 Dodd 28.

32S ee the article by J. Bradshaw, "Oral Transmission and Human
Memory", ET 92 (1981) 303-307, for a valuable discussion of the neuro-
logical limitations of and tendencies in memorization and the implications
for studies in the gospel tradition.
33Dodd 24-25.
10 AUSTRALIAN BIBLICAL REVIEW 37/1989

situation, has drawn upon verses known to him in 4G and which in his
memory are linked by conceptual (scriptures-Moses) or verbal (Moses-
Moses) associations.
Lines 22-31 also are Johannine in style and content. Though the first
lines are obscure,34 the general sense is quite clear: the rulers try to lay
hands on Jesus to hand him over to the people. This occurs after they
attempted to stone him. As earlier mentioned, these lines should be
thought of as the conclusion to the incident in lines 1-21. As to Johan-
nine style, lithazein, piazein, and eJeJuthei autou he hara are all demon-
strations of this point. 35 The synoptics know of no attempt to stone
Jesus. And though lines 22-31 have no exact parallel in the 4G, they do
bear resemblance to several situations described-7:30-44, 8:31-39 and
8:59. The closest synoptic verse is Luke's comment at the end of the
Nazareth incident in 4:30: autos de dielthan dia mesou auton eporeueto.
The strongest evidence that these lines are a conflation of J ohannine
elements has to do with the role of the crowd: presuming that paradasasin
is correct (lines 26-27), we may well ask why the rulers want to hand
Jesus over to the crowd. If the crowd desired to seize Jesus they could
have done it themselves. Such a confusion does not enter the Johannine
accounts, and we may be left to suggest that VG has modified the
Johannine presentation in the light of the general Jewish hostility to
Christianity at the beginning of the second century.
How shall we assess the relation to 4G revealed in lines 1-31?
Firstly, we must say that though the incident is so decidedly Johannine in
content and style,36 the author of VG has constructed a narrative that is
synoptic-like in structure. He is not interested in presenting Johannine
discourses but in portraying a conflict incident in Jesus' ministry.
Secondly, I have suggested that VG betrays a conscious dependence on
4G. But is the dependence from memory or is it literary? Scholars tend

34Bell and Skeat in their two works in 1935 first suggested ochlo in line
5 but then revised their recommendation to helkostn in the second work.
35 But the addition of tes paradoseos interprets the hour of Jesus in a quite
unjohannine way: instead of being the hour of his glorification, it is now
the time of his arrest (cf. Luke 22:53). What this indicates is that though
the author of UG may have known the 4G and respected it, he has not
really understood its deeper theological insights. His thinking on the
ministry of Jesus is still fashioned by what he knows from the synoptic-
like oral tradition.
36Wright has convincingly demonstrated, against Koester and Jeremias
(New Testament Apocrypha, 1.95), that the few synoptic features do not
demand synoptic-type traditions here. See "Apocryphal Gospels", 213-215.
 Pryor: Papyrus Egerton 2 11

to suggest that memory alone is at work here. 37 But memory and slavish
literary dependence are not the only two options. It is possible that the
author in rather freely constructing his incident is consciously drawing
upon Johannine material alone, and that for certain sayings he conforms
almost verbatim to a known text. 38 That, of course, leads to the obvious
question: why is he at times so free and at other times so controlled by
the text? Here we can only guess at an answer. Perhaps he feels less
liberty to tamper with the words of Jesus than with narrative details.
Thirdly, apart from lines 45-47 to which we are about to turn, VG
displays no tendency to mix Johannine and other traditions. We simply
cannot say, with Jeremias, that "the Johannine material is shot through
with Synoptic phrases and the Synoptic with Johannine usage ... "39 The
only other possible Johannine fragment is Frag 3 recto, but its evidence is
so minimal that we cannot conclude anything definite from it. There is a
reasonable chance that it is the tail end of a saying of Jesus ("I and the
Father are one") with a consequent attempt again to stone Jesus. If so,
then it is a recollection of John 10:30-31 with apokteinosin instead of
lithasosin. 40

C. JOHN 3:2 AND UG

In the light of the above, what are we to make of lines 45-47 which
remind us so much of John 3:2? In view of what I have concluded, it
would be foolish to deny that the author was mindful of John 3:2 as he
wrote these lines. But is it a total fabrication adapted from 4G for the
occasion? This would be a possibility were we to have suggested, with
Wright, that the whole pericope is a blend of synoptic and Johannine
elements and nothing more. But such is not my advice, but rather that
the incident derives from oral tradition independent of the synoptics. That
being so, the pericope as VG knew it must have had some flattering
introductory words similar to what we have in VG and in the synoptic
accounts. As we address this problem, we can with reasonable confidence
make the following assertions:
(i) Nowhere else has VG blended synoptic and Johannine elements.
This raises the possibility that this has happened here only because the
Johannine and the oral traditions were quite close anyway.

37Thus Wright, "Apocryphal Gospels", 214, Braun 1.92, Jeremias New

Testament Apocrypha. 1.95.
38Though Dodd does not take up the question of memory as opposed to
literary dependence, his silence leads me to suspect that he may well have
thou~ht along similar lines to me.
3 Jeremias, New Testament Apocrypha. 1.95.
40So Dodd 85, Wright, "Apocryphal Gospels", 219.
12 AUSTRALIAN BmLICAL REVIEW 37/1989

(ii) If the author of VG only had John 3:2, it is unlikely he would

have modified the verse so much. In Frag I sayings of both Jesus and the
rulers more faithfully reflect the text of John. Here the change from rabbi
in 3:2 to didaskale Iesou may be thought natural in the light of both the
equivalence of rabbi and didaskalos and also the seeming preference of VG
for this title (line 33). But equally, the difference may be traced back to
the oral tradition behind VG, for it is noteworthy that Mark 12:14 et par.
also begins with didaskale. Moreover, we must bear in mind that if
Jesus' response in lines 52-53, with its conclusion of didaskalon, is
integral to the tradition, then it demands that the opening salutation
contain didaskale. The last phrase of the address is also unjohannine:
huper + accus. is not found in 4G (though see 12:43); and while John
does occasionally refer to "prophets" (1:45; 6:45; 8:52,53), he never
speaks of "all the prophets" (x 3 in Luke). We need also to remember
that marturei is far from certain as only the "m" is clearly visible. It is
not impossible that the reading was something like meizona estin.41
(iii) We may presume that oidamen hoti was present in the non-
J ohannine traditional introduction to the story. It is found also in the
synoptic version. Had these words not been present in the original oral
form of the saying, it is difficult to imagine what would have prompted
the author of VG to think of John 3:2.42
All of these considerations drive me to the conclusion that the form of
the introductory remark in the tradition known to the author/compiler of
VG must have included the following: "Teacher Jesus, we know that ...
for ... more than all the prophets". Beyond that we cannot go, for it
looks as though Egerton's knowledge of John 3:2 has shaped his words,
without its being his only or primary source.43

CONCLUSION
We cannot have any confidence that the synoptic gospels are either
known to or treated with respect by the author of VG. There is a much
stronger case for supposing total ignorance of the synoptics, so that
synoptic-like incidents derive from living oral tradition. Otherwise we are

41 If line length makes this reconstruction an impossibility, the general

point being argued is not thereby weakened.
42Here I disagree with Wright. See the passage from Wright quoted
earlier, with my comments.
43If this is correct, the implications for the pre-literary form of the
Nicodemus story may be considerable: was it originally, like the tribute
incident in the synoptics and Egerton, a controversy story?
 Pryor: Papyrus Egerton 2 13

left with a document which treats the 4G source far more conservatively
than the various synoptic sources.
John's Gospel, on the other hand, is known to the author of VG, and
he clearly respects it as providing authoritative guidance on the person and
mission of Christ. But he does not feel that it is an untouchable holy
book. On the contrary: as he writes up his own account of the incidents
in Jesus' life, drawing largely from living oral tradition, he is also
influenced, particularly in the recounting of sayings of Jesus, by what he
knows from John, and he makes careful yet liberal use of it. He is thus a
witness to the early Egyptian knowledge and reception of 4G in what we
can only presume are "ordinary" Christian circles.44
VG is a witness to one other facet of early Christianity, the collecting
of gospel units into a consecutive narrative similar to the synoptic style
and pattern (Luke 1:1). We have no knowledge of what kind of literature
it represents, whether there is a beginning to Jesus' ministry and whether
it contained a passion narrative. My own suspicions, in the light of the
mention of "his hour" (1:29) and of the Jewish opposition which is so
strong, is that there may well have been plot development and even a
passion narrative. This would make the synoptics and VG to be indepen-
dent witnesses to the same trajectory. But this can ever remain nothing
more than speculation.4s

44That is, the document betrays no gnostic or other distinctive

tendencies. So also Braun 1.94, Dodd 45,51.
4S Such speculation has recently been furthered by D. Wright, who
suggests for consideration the hypothesis that the Gospel of Peter and UG
are one and the same document. Discussion of Wright's article is outside
the scope of this paper, as is also the question of the precise intention of
the gospel. See D. F. Wright, "Papyrus Egerton 2 (the Unknown Gospel)-
Part of the Gospel of Peter?" The Second Century 5 (1985-86) 129-150.