THE TEXTUAL STRUCTURE
OF JOHN'S GOSPEL
TABLE OF CONTENTS
(Click on items to go directly to that text.)
NOTES ON TABLE I
DEDUCTIONS BASED ON THE TABULATION
INSIGHTS FROM REVELATION
Textual analysis suggests the Gospel was first a collection of notes on specific topics. The notes were then organized into incremental codex half-page and page segments. These segments were later woven into a continuous narrative. This structure led to dislocation of text sometime before publication. The evidence suggests final organization and editing were done without the influence of John, perhaps after his death.
Since early Christian days various individuals
expressed dissatisfaction with the chronology of John's Gospel. A disturbed
sequence in Chapters 4 to 7 was noted by Tatian, who copied large sections
of the Gospel into his Diatessaron. He placed the order as Chapter 6,
4.4-45, 5 and 7. If we follow the traditional text we find Jesus with ministry
in Judea, 3.22-30, going up to Samaria and Galilee in 4.4-54, coming back to
Jerusalem in 5.1-40, and back to "the other side of the Sea of Galilee" in 6:1.
How could he go to "the other side" if he was in Jerusalem and not already in
Galilee? Many have commented on his cleansing of the temple in 2.13-25. Did the
cleansing take place during his last visit when he denounced the scribes and
Pharisee, rather than at the beginning of his teaching ministry, when he would
be more concerned with public reaction? (See Matthew 23, a section not included
in John's Gospel.) The most notorious dislocation is in 14.31 where Jesus gives
the command "Arise, let us go hence," as though the group should disburse, but
then continues to discourse through Chapters 15 and 16. Other early Christians,
including Irenæus and Origen, also made remarks which indicated dissatisfaction
with the text(1).
In 1928 J. H. Bernard and A. H. McNeile published a comprehensive analysis of the apparent dislocations. They proposed a cause that was dependent upon the structure of the text. According to their study, the remarkable aspect of the proposed dislocations is that they follow integral segments of size. By counting Greek letters, and noting that the text was apparently written originally as a codex, they proposed the movement of integral codex leaves, either accidentally or perversely to unknown criteria(1).
(Manuscripts arranged in codex leaves began to appear near the end of
the first century. They gradually replaced the older system of scrolls.)
The oldest known manuscript evidence of the New Testament is a short fragment of John's Gospel containing Pilate's famous question, "What is truth?" Paleographic comparisons date this fragment to the latter part of the first century or beginning of the second. Except for an iotacized dipthong, there is no difference between the text in that papyrus and our modern text(2). The recto contained portions of lines from 18.31-33 while the verso contained 18.37-38. Reconstruction of the lines shows that the manuscript averaged 31 or 32 letters per line, and that a page had approximately 20 lines. Thus a page contained between 620 and 640 letters, while a leaf of the codex had double that number.
Other ancient manuscripts and fragments show
other page sizes. The Egerton 2 fragment had 17 lines/page with about 26
ltrs/line; the Bodmer II papyrus had 25 lines with about 28 ltrs/line(2).
These two codices would have had about 880 and 1400 ltrs/leaf respectively.
Bernard and McNeile cite the Oxyrhynchus papyrus (#208 and #1781) at about
710 ltrs/page for 1420 ltrs/leaf, and the papyrus codex #1780 at only about
Bernard and McNeile proposed leaf movements in John's text, based on 750 ltrs/leaf, from an original manuscript that then became the Nestle standard. (This standard was revised and updated under the coordination of Kurt Aland in 1966(3), cataloguing and compiling all known and newly discovered manuscript materials.) Examples given by Bernard and McNeile are:
The average error at 750 ltrs/leaf is -27 ltrs. If the average number
of ltrs/line was 30 the difference from integral leaf segments was less
than one line for three of the cases, and four or five lines for the other
three, (out of 25 lines per leaf.) Certainly, this is strong support for
The proposal invokes two assumptions.
One, the scribe conservatively filled all leaves; he did not leave large
blank spaces in pages. The desire to conserve writing materials is witnessed
in available manuscripts by the stringing together of words without space
separation, and the breaking of words from one line to the next when a
line was filled.
Two, he maintained the same style and size of writing throughout the
text. In some manuscripts letters are crowded together in one line, while
other lines are more spaced(2). For example, in a letter from
an official under Hadrian in 135, Berlin papyrus #173, the average number
of letters over nine lines is 26. But the first line is crowded with 35
letters. The first page of the Bodmer II papyrus has 36, 39, 32, and 31
letters respectively in the first four lines but in subsequent lines the
scribe settles down to an average of only 27 letters. Thus we cannot rule
out the possibility that deficient or excess letter counts affect leaf
calculations due to scribal inconsistencies.
If the proposal for the movement of leaves is valid it necessitates interchange with other integral leaves. However, Chapter 6, if switched with Chapter 5, has a calculated value of 7.5 leaves at the 750 ltrs/leaf proposed by Bernard and McNeile, short by one page. If 3.22-30 made up one leaf, and was later inserted before 3.31-36 to make our present text, the latter section should be an integral leaf also. (This assumes that 4.1 continues as a different episode with its own dedicated section.) But 3.31-36 makes up only one-half leaf (439 letters), one page (plus two
lines). Obviously, a total count of all letters of all sections is necessary to demonstrate rigorously that the proposed dislocations are due to leaf movements.
In order to evaluate the assumption of 750 ltrs/lf used by Bernard and
McNeile, and to obtain quantitative values, I estimated letter counts for
all segments of the Gospel text, divided according to the thematic sections
given by Bernard and McNeile.
I asked if all the thematic sections were designed in similar manner,
as integral page or leaf segments . Why would Chapter 6 be moved as an
integral unit unless it was written according to an integral leaf design?
If other sections were composed as integral leaves they should appear also.
Next I used values from 710 to 790 ltrs/lf, in increments of 20 ltrs/lf,
to determine how individual sections would appear when tabulated in comparison
columns. I determined that a more realistic value for leaf design was 770
In fact, the calculations provided truly startling
To first recognize integral segments of text consider sections which are easily distinguishable because of their independent subject matter. The prayer of Jesus in Chapter 17 is famous. It is two lines short of three leaves, based on my assumed value of 770 ltrs/lf. The story of the woman caught in prostitution in 8.1-11, hotly contested as an extraneous insertion into the text, and not included in the Bernard-McNeile tabulation, is a mere 8 letters more than one leaf. Jesus' teaching in the temple, 7.14-24, is two lines more than an integral leaf. Jesus' entry into Jerusalem in 12.12-19 is two lines short of an integral leaf. Jesus' foretelling the betrayal and Judas' departure in 13.21-30 is again a mere 8 letters above one leaf. And so on. Obviously, the text is structured in many places into integral leaf segments. But not always. See Table I.
NOTES ON TABLE I
The sections given by Bernard and McNeile are shown in Column I, except as follows:
I condensed Chapter 5.1-40 into one unit to save table space.
I follow the traditional sequence in Chapters 10 through 12.
These differences do not alter conclusions drawn from the Table.
DEDUCTIONS BASED ON THE
Could the divisions into page and half-page segments be fortuitous or the result of my methods of analysis?
In Table I and the above discussion I gave criteria for acceptance of nominal leaf, page, and half-page counts based on the number of letters to a leaf. These translate to lines per leaf or page, but we do not know the number of letters per line in the original manuscript. Using available manuscript evidence the minimum value would be approximately 25 ltrs/line and the maximum about 32. This translates from 31 to 24 lines per leaf, 16 to 12 lines per page, and 8 to 6 lines per half page. My criteria for acceptance into the incremental values then would be +/- 2 to 3 lines per leaf, less than two lines for a page, and just slightly more than one line for a half-page. Beyond 3 lines the incremental value shifts to the next half-page level.
Of proposed 25 sections 19 fall within two lines of the incremental leaf and page hypothesis. Of proposed 18 sections 14 fall within one line of the incremental half-page hypothesis. Certainly, these are fair limits for acceptance. The remaining 16 sections fall slightly beyond these criteria.
Such tight criteria prohibit the leaf, page and half-page assignments from being mistaken for the next incremental page value.
A histogram of the calculated leaf counts normalized to incremental leaf values, (for example 4.22 to 0.22) shows a definite clustering around the 0.25, 0.5, 0.75, and 1.0 leaf values. 75% of the sections fall within the tight limits I used to demonstrate my hypothesis. It is highly improbable such large proportion of the sections would fall within those tight limits strictly by chance. To propose chance distribution voids the rewards of this analytical study that provides such deep insights.
I examined various statistical methods to determine if there were other underlying causes to the segmentations. These included distribution of the number of groups by group size, size of the sections according to sequence in the text, and deviation
of the sections from nominal values. I could discover no subtle natural or accidental causes which might have created the integral segmentations.
The uncertain sections becloud a clean demonstration of the half-page design. But those sections were deliberately interwoven into larger integral groups. 7.10-13 and 7.25-36 add to 1.72 leaves for a nominal value of 1.75. 10.1-6 and 10.7-18 add to 1.97 leaves, for a nominal value of 2.0. The uncertain sections are related in other ways.
Evaluation of the raw data does not expose the creative design used by the scribe for groups. This design may be seen more readily by calculating the group distributions in macroformation.
Jesus' prayer in Chapter 17, and the woman caught in prostitution in 8.1-11, both as independent subjects with integral leaves, serve as boundaries for a larger grouping of sections between those two. Where 8.12-59 and 9.1-41 add to 9.5 leaves, and the Last Discourse is 10.75 leaves, the Commentary in 12.36b-43 is 0.75 leaves, to make that macrogroup exactly 40 integral leaves. Thus the scribe may have adjusted his Commentary to create such a large macro integral value. Another macro group may be seen from 18.1 to the end of the Gospel where the count is exactly 17 integral leaves. The section at 20.19-29 is 1.5 leaves, but the two commentaries are both 0.25 leaves to provide the integral count. The group from 7.14 to 8.11 is 5 leaves, but this cannot be considered strictly as a macro group. The section from 1.1 to 5.47 add to 27.5 leaves. This is the only exception to an integral leaf count in the macrogroups, but it is an integral page count.
Note that this one exception is caused by 6.60-71 and 5.41-47, which add to 1.77 leaves for a nominal value of 1.75. Note also that this exception falls within a portion of text which has been subject to doubt since early Christian days.
With a page short from an integral leaf in this macrogroup, the scribe could easily have pushed all following text up by one page to end with a blank page on the last leaf.
This analysis profoundly impacts upon the task of devising an explanation for the evidence. I propose that the Gospel was originally composed in small segments, by subject, as a group of notes. The notes were then sized to incremental half-pages, pages, and leaves. This complex was then woven into the complete document.
Did John write the notes?
He could have prepared and collected them over a considerable length of time. Or he could have dictated them over a shorter period to a personal scribe. They certainly show first-hand knowledge of events. But the confusion in the text was due to the working of the assemblage into a complete document that did not have John's direct guidance.
Textual comparison between the Gospel and the Book of Revelation shows considerable differences in vocabulary and style. These differences are convincing that two different scribes composed those works. If John used a scribe for collation and editing of the Gospel notes, that scribe was educated beyond the literary level displayed in the Book of Revelation. Thus we have a more obvious explanation for the controversy over the respective authorships. John could have used two different scribes for the respective works.
Given that the Gospel notes were sized to fit incremental leaf and page values, we are forced to ask how a scribe may have edited the material. Since the evidence speaks so strongly to a hand other than John's we must ask how far that hand went in editorial creativeness.
It is seems incredible that John went to the trouble of designing his notes around incremental leaf and page values. We do not normally think, nor write, with such segmentation. Editing to half-page, page, or leaf segments also is not natural. Normally, when one writes, one places on paper one's thoughts or presentation, not heeding where it will end. If a partial page hangs over, so be it. Editors for journals will edit an entire document for brevity, or perhaps remove an unimportant section, and size to fit their design goals, but they will not segment such a large document to the criteria displayed in John's Gospel. The half-page segments suggest more than a mere relating of accounts. They suggest a rigorous editing to limit to those discrete size segments. Thus it is very possible the scribe used a free hand, not only in design toward integral page and leaf values, but in text content to create those integral values. This could take the form of both additions and deletions.
The Commentaries are telling evidence. Although they are related to adjacent text, were they part of the note assemblage which the scribe felt had to be incorporated into the text? 3.16-21 and 3.31-36 suggest he respected the materials although he may not have fully understood the original purpose. When faced with them he may have felt it awkward to place both comments as contiguous; they may have seemed repetitive to him. Perhaps he solved the problem by assigning them to two different sources, John and the Baptist.
But other views of textual design may be derived from the Commentaries. If a scribe did not understand the continuity of events, as evidenced by the confusion in Chapters 5 and 6, in the Last Discourse, and so on, he may not have recognized it in Jesus' Passion remarks. Thus he could easily slip 12.36b-43 into the text in its present location to fill out that macrogroup. The Commentaries on the scope and purpose of the Gospel and its authentication, in Chapters 20 and 21, now appear more naturally as compositions by the scribe to fill out the last macrogroup. Perhaps all the commentaries were scribal creations to meet the design requirements of the macrogroups, although I personally wish that John was the original inspiration behind their creation. This possibility raises serious question of their authorship, including the fundamental definition of Christian hope in 3.16.
Why would a scribe devise such a design?
It may have been a peculiarity of his personality, a way to clearly organize and track the material from subject to subject. Or it may have been his method of computing the length of the document and keeping order in the different episodes. Perhaps there were other reasons which escape us.
The dislocations may not have been perverse text movements. It may be erroneous to view them in that light. The scribe who wove the segments into the final text may not have understood the sequence of the original composition. He may have had within his hands a loose assemblage of notes and was faced with the task of making a sensible account from them. This
person could have been the one who made the original notes, or another person who came into possession of them before publication.
All of this implies that John was not present when the assemblage of notes was woven into our present text. Although the scribe generally followed John's sequence in composition, he did not appreciate the chronology, or logic, for all sections of the text. When faced with the task of weaving it together he did not have John's advice to guide him.
If John had rehearsed chronology with the scribe it seems doubtful he would have confused sequence. This lends support to the proposition that the assemblage of notes came into possession of another scribe when John's influence was no longer present.
The assemblage of notes may have become confused merely by accidental shuffling before a later scribe came into possession of them. This placed a burden upon him to determine correct sequence. He may not have been of great analytical mind, although displaying good literary style.
Other deductions can be drawn from the evidence. Jesus' prayer in Chapter 17 now becomes suspect as a literal, verbatim report. It may reflect John's memory but we cannot say with certainty how much it may have been edited, or even created by a later scribe.
The temple cleansing in Chapter 2 seems to be out of place chronologically. We would not expect Jesus to create animosity for himself and his teaching mission before he barely gets started. The cleansing probably took place at the end of his mission, along with the condemnation of the scribes and Pharisees, the announcement of his Passion, and other terminal events. Examination of this segment in relation to neighboring segments, now easily recognized in Table I, suggests a movement to fill out an integral page increment.
The Commentaries, gross misplacement of the Temple cleansing, the sizing of Jesus' prayer, and other elements lead one to suspect that the integral page, leaf and group design may have strongly influenced the scribe in his choices, and his editorial freedom.
This analysis offers much better appreciation for composition of the text, and dispels much myth centered around it. The Commentaries could easily have been later scribal creations to fill in integral text segments. The section on the woman caught in prostitution, 8.1-11 also now shows more clearly as a possible later insertion into the text, borrowed from some strong early Christian tradition.
It seems highly doubtful that a text which had the blessing of John would have been altered. Neither could a mind capable of remembering such details of history have been too weak to know the proper sequence of events, unless the notes were created much earlier and later given to a scribe to collate. Therefore the weaving into a continuous, but erroneous, narrative had to be done after John's influence no longer existed. The manuscript was removed from his presence, he became too feeble, or he died.
I have not attempted analysis to determine deeper logic which may lie behind the design, nor is it the place of this paper to engage in further textual study that may be derived from this analysis. I leave that to others.
The results of this study obviously have profound ramifications on our understanding of the creation and organization of the Gospel, as well as the Book of Revelation, and John's letters. If John was an uneducated man, he may have felt impelled to rely on scribes to compose his works. This relationship then left him exposed to the vagrancies of those other persons. If he asked them to assist him in creation they may have felt privileged to participate in such activity. But their feeling of privilege may have been amplified into self-esteem that gave them a sense of implied control over the documents. Perhaps John's Revelation was subject to even greater corruption in text; if a less educated or less astute scribe helped in that composition he may have felt less constrained in his final editing.
While all of this is speculative, we now have quantitative grounds for such speculations.
A whole new light is now cast on the proposed divine inspiration of the New Testament. John's Gospel vividly displays purely human elements in its design. It would be idiocy to assign those to God. We now have a much more realistic grasp on the creation of John's Gospel and, by extrapolation, perhaps all New Testament documents.
INSIGHTS FROM REVELATION
The foregoing discussions and deductions are based strictly on analysis and interpretation of the textual structure of the Gospel. The Urantia Papers offer this information:
"The Gospel according to John relates much of Jesus' work in Judea and around Jerusalem which is not contained in the other records. This is the so-called Gospel according to John the son of Zebedee, and though John did not write it, he did inspire it. Since its first writing it has several times been edited to make it appear to have been written by John himself. When this record was made, John had the other Gospels, and he saw that much had been omitted; accordingly, in the year A.D. 101 he encouraged his associate, Nathan, a Greek Jew from Caesarea, to begin the writing. John supplied his material from memory and by reference to the three records already in existence. He had no written records of his own. The Epistle known as "First John" was written by John himself as a covering letter for the work which Nathan executed under his direction."
The dramatic difference between John's Gospel, and his Book of Revelation shows that John used scribes to create those documents, and that at least two different scribes were used. This would be natural if twenty years separated their creation.
The evidence of John's Gospel shows that at least two difference scribes were involved in that creation. The first worked directly with John, to create the main chronology we now find in that document. We do not know if that scribe created the integral segments of text, but the evidence suggests so. A later scribe, faced with confusion in the displacement of leaves, then reconstructed according to his best view, but without the hand of John to guide him. The point at which the Commentaries, or the woman in prostitution, or Jesus' prayer, or other segments, were inserted is impossible to say. The first scribe may have inserted those pieces as part of the original integral design, or a later scribe may have done so, or a combination of both. If the latter the evidence shows that he recognized the integral design an obeyed its principles, although confused by the chronology.
Thus the textual analysis supports the assertions made by divine Revelation.
The statement shows that John worked from memory, not from notes. The notes, which were later reworked into integral segments of text, would then have been created by Nathan, working directly under John.
1. -- J. H. Bernard and
A. H. McNeile, Gospel According to St. John, International Critical
Commentary, T. & T. Clark, Edinburgh, 1928.
2. -- V. Salmon, translated by M. J. O'Connell, The Fourth Gospel, The Liturgical Press, Collegeville, MN, 1976.
3. -- Kurt Aland, et al, The Greek New Testament, United Bible Societies, 1966.
Ernest P. Moyer