Internal Evidence

Why δε
in John is Irrelevant

Updated and Expanded from: Nazaroo, Internal Evidence for John 8:1-11, (1982)

Page Index

Last Updated: Feb 21, 2009

Prologue: - Introduction to Problem

Analysis of δε : - Why its Irrelevant:
    Part I Frequency Tables and Commentary
        Section 1: John Chapters 1 - 3
        Section 2: John Chapters 4 - 7
        Section 3:Pericope de Adultera
    Part II: The Signs Narrative (cont.) chs. 8 - 10
    Part III: The Signs Narrative (cont.) chs. 11 - 12
    Part IV: The Last Supper Monologue chs. 13-17
    Part V: The Passion Narrative chs. 18-21
    Conclusion: Why de is Meaningless

Return to Index



"[Many] sentences are connected with 'δε' in the pericope,
and this is unparalleled in John..." 1

"John does not use the Greek word 'δε' ("but"/"and") near as often as the other gospel writers, but in this passage (Jn 8:1-11),
it is found MUCH more often than in the rest of John. ..."

This has been a standard argument against John's authorship, (and therefore the authenticity) of John 8:1-11 since the 1850's.

Ironically, conservatives who consider the passage authentic are sometimes mocked for having an 'outdated' viewpoint. Yet the stylistic arguments against the passage were pretty much finalised by men like Samuel Davidson (1848), and Tischendorf (1860) over 150 years ago.

Such presentations were commonplace in the 1800's, and appeared sophisticated and convincing in their time. However, most of this work, in spite of its 'political correctness' and popularity, was too naive in method and subjective to have any enduring value.

Statistical analysis was still in its infancy, and modern linguistics was simply non-existant.

Yet to this day, the same old 'assured results of modern criticism' are trotted out as though they were really the latest findings of linguistic and historical science.

Yet nothing could be more inaccurate: the 'science' of Biblical Criticism has largely remained in the Dark Ages. Very little of the current corpus of work has used truly modern scientific techniques.

The field remains a backwater where the most inept and methodologically flawed conclusions are passed off as scientific 'fact'. Here is a fair example of the high-school level presentations usually given the public today as if it were serious scholarship:

"John does not use the Greek word 'δε' near as often as the other gospel writers, but in this passage, it is found MUCH more often than in the rest of John.

Out of the other 867 verses in the gospel of John, the word 'δε' is found 203 times, or in an average of 23% of the verses, while in these 12 verses, John 7:53 to 8:11, the word 'δε' is found 11 times, or an average of 92% of the verses.

Another change in style is an increase of participial phrases. For these reasons and others, I feel no uncertainty in flatly declaring that the passage is not written in the style of the apostle John."

(taken from a popular website with its own translation of the NT, which omits John 7:53-8:11)

To an ordinary person without any specialized training or knowledge of statistics or analytical technique, this appears quite legitimate and impressive. However, we will show shortly that it is simply meaningless nonsense.


1. Gary M. Burge, A Specific Problem In The New Testament Text And Canon: The Woman Caught In Adultery (John 7:53-8:11), (JETS Volume 27, p 144. June, 1984; 2002).

Return to Index

Analysis of δε

The Critical Text Chosen

We begin with the W-H / UBS critical text. We don't want people complaining that the 'Byzantine' or Majority text-type is a source of error in the analysis. Secondly, we follow the paragraphing identified in those versions. Again, if the case can be made at all, it must be made on the basis of the critical text advocated by those who claim significance to the 'evidence' they say exists.

As it turns out, the textual variants however, will not significantly affect the results found in our table, since those involving 'de' are less than a dozen. In this case, methodology is far more important than the critical text chosen for the survey.

We offer the basic Freqency Table used at the end of this article/webpage. Any observations can be checked there, and doubled-checked with a UBS text and critical apparatus.

The Targets and Scope of Analysis

We have stepped through the entire Gospel of John, paragraph by paragraph, counting the frequency of και, αλλα, δε, and ουν. There are several reasons for including all four of the most common connecting particles. One is that it gives us a good perspective on the way all such particles may vary from section to section, depending upon content, style, and other influences.

Second,και, being by far the most common particle, shows much greater stability throughout the length of the document. This helps to place the true significance of the range of variation for other words.

The focus on paragraphs has a twofold advantage. In many cases, there will be possible sources being used by John, as in the Prologue, the passages paralleling Mark (John 6f, feeding of the 5000 etc.), and the passages from Synoptic traditions (John 2:12, 4:43, 12:12, Temple Cleansing, Nobleman's Son, Triumphal Entry etc.). The paragraphs into which the Gospel naturally falls may be good indicators of where the influence of John or another source may start and stop.

Overall word-counts of paragraphs offer a straightforward and convincing way to calibrate short-term frequency of key-words in their local context.

A second question revolves around narrative versus dialogue, or more often monologue, and where other characters 'end' and John's commentary begins. These issues can possibly be illuminated by a study of language and style.

Return to Index

Table and Commentary

Relative Frequency of και, αλλα, δε, ουν in John

The following counts are based upon the W-H / UBS critical text of the Gospel of John. Paragraph sections follow those of the UBS text. Some interesting sections are marked: Jesus' monologues are marked in yellow, and John the Evangelist's commentary is in green.

Return to Index

Section I :
John ch. 1-3

John Part I: The Signs Narrative

Paragraph και αλλα δε ουν
1:1-5 6 0 0 0
1:6-13 3 2 1 0
1:14-18 6 0 0 0
1:19-28 11 0 0 3
1:29-34 7 2 0 0
1:35-42 8 0 2 1(0)
1:43-51 11 0 1 0
2:1-12 17 0 4 0
2:13-22 18 0 1(2) 3
2:23-25 1 0 2 0
3:1-15 13 1 1 0
3:16-21 2 2 3 0
3:22-30 10 1 2 2
3:31-36 4 1 1 0

The first section we examine here will largely set the tone for what is to follow, although this part of John has some unique features of its own.

The Exceptions: The first 34 verses of John are unique. Many commentators have thought that the Prologue is part of an older psalm or poem, incorporated by John as a fitting beginning to his gospel, and represents a kind of catechism of the Johannine Community on behalf of whom he writes.

Not surprisingly, this section has unique features unseen in any other part of John. One of these is the almost complete absence of both 'δε' and 'ουν. Although this can partly be accounted for by the subject matter and content of the Prologue, it is not a total explanation.

One possibility is that this section is a more direct 'translation Greek' from an Aramaic or Hebrew original, although the evidence is too scanty for a firm pronouncement.

What is of interest for our discussion, is that AFTER this section, both 'de' and 'oun' exert their presence in entirely different ways:

' δε' - This connective particle is a soft 'but', lying halfway between the strong adversive 'alla' ('But') and the purely associative 'kai' ('And'). From 1:35 forward, there aren't two paragraphs in a row without its presence. (with 1 exception, 19:28-37) Although the frequency of ' δε' goes up and down more or less randomly, it remains a permanent part of John's expression throughout the Gospel. There is hardly a section in which John can't find a natural use for this connective particle.

' ουν' - This connective particle could hardly be more different in its usage by John. When it does appear, it is in short bursts, followed by multiple paragraphs, long sections of its absence.

Half the time it pops up, it is in the dialogue of antagonists. Its main narrative use is in connecting events a short distance apart, usually speech, where a logical sequence of cause and effect is meant.

Yet it does seem to be affected, even in its relative scarcity, by the nature of the content. It is almost entirely absent from the long monologues of Jesus at the Last Supper. Here we must suspect that the material simply comes from a different source than John's narrative style. In a superficial way, we could say that 'John prefers ' ουν '" but this would be a misleading distortion:

Unless a speaker is presenting a logical argument or retelling a story himself, it is unlikely that ' ουν' will be a feature of his monologue. In the case of short exchanges, such as those between John the Baptist and the Pharisees, ' ουν' appears for a reason. The Pharisees are responding to each of John's short answers with a new question, spawned from the previous answer. This must be classed as 'special dialogue', and not natural speech or monologue.

In summary then, ουν only gets used when it is functionally necessary, and hence appears in blips and bursts. δε on the other hand, is a very handy general connective particle, and seems to be a permanent part of John's writing style, in narrative, commentary, and even dialogue.

A Specific Example: John 3:1-21

We have highlighted in yellow the parts of John which are mostly Monologue (i.e., Words of Jesus in Red in many bibles). Because the main contrast between parts of John has focussed upon narrative versus dialogue/monologue, this allows us to see any peculiarities in the usage of the chosen particles in these sections.

We mentioned that the prologue has been made a potential candidate for 'translation Greek' or a Semitic Source document. Interestingly, modern analysis has NOT supported the claims of John's stylistic usage of either 'de' or 'oun'.

Instead, the only solid and convincing observations regarding the usage of various particles have been those surrounding the relative usage of 'δε' and 'kai':

"The frequently occurring syntactical features found to be significant indicators or translated and original Greek are as follows:

(a) Syntactical features which are less frequent in 'translation Greek' from Hebrew or Aramaic:

...(2) the use of 'δε';
...(6) the use of the adverbial (circumstantial) participle;

(b) Syntactical features which are more frequent in 'translation Greek' from Heb. or Aram.:

...(2) the use of 'kai' to join main (independant) clauses;

(R.A. Martin, Syntax Criticism of the Synoptic Gospels (1987, Edwin Mellen Press))

Now John 3:1-21 is an exciting section, since it is controversial as to exactly where Jesus' speech leaves off, and John's narrative comments begin. Many older 'Red Letter' bibles mark the whole section as Jesus' speech, but most analysts hold that John's comments begin at 3:16.

Now looking at the chart, we see several interesting points:

(1) the use of KAI is unusually high in the part accepted by all as Jesus' speech.

(2) by contrast, the use of δε is lower.

(3) the use of ουν is absent from both sections.

There appears then to be a good case that John is the author/commentator speaking in 3:16-21, and that Jesus' speech in 3:1-15 may be dependant upon a translation from either Hebrew or Aramaic.

But the side-bar to this is that 'ουν' is apparently not an active or useful indicator of John's style at all, and rather only appears when it is functionally called for.

Return to Index

Section II:
John Ch. 4 to 7

Paragraph KAI Alla DEOUN
4:1-15 9 2 2 4
4:16-26 8 1 0 0
4:27-38 11 0 1 2
4:39-42 3 0 1 1
4:43-54 6 0 3 6
5:1-9a 5 0 2 0
5:9b-18 7 1 4 2
5:19-29 8 1 1 0
5:30-40 7 2 3 0
5:41-47 2 1 1 0
6:1-15 7 1 4 5
6:16-21 4 0 2 1
6:22-33 6 4 0 5
6:34-40 4 3 1 1
Paragraph KAI Alla DEOUN
6:41-51 7 0 1 1
6:52-59 8 0 0 2
6:60-65 2 1 1 2
6:66-71 5 0 1 1
7:1-9 4 0 3 2
7:10-13 2 1 2 1
7:14-24 5 3 2 2
7:25-31 6 2 2 3
7:32-36 7 0 0 2
7:37-39 2 0 2 0
7:40-44 0 1 2 2
7:45-51 5 1 0 2

Next we observe the similar behaviour of 'oun' to 'alla' (the strong form of 'but'). Yet here now, it is the similarity in occurance that is striking!

Both particles appear and vanish for large stretches, quite independantly of the nature of the content (i.e., narrative, dialogue, monologue, commentary).

Quite obviously, both particles vary completely independantly of each other also, indicating that the only factor in dictating the presence of either of them is their functional necessity in the specific statements being made in various sections of John.

The theory then, of 'oun' versus 'de' as indicators of the style of John versus another writer of a different identity, has fallen flat on its face.

Return to Index

The Pericope de Adultera

Paragraph KAI Alla DEOUN
7:52-8:11 10 0 11(9) 1

It will be immediately apparent that in comparison to the rest of John, the values for John 7:53-8:11 only show a moderate increase in the number of 'de's. All other particles show a normal or average count.

Yet even here, we need to note the difference between the sources of the two texts in this case. In the Critical Editions such as Westcott & Hort, or the UBS/Nestle-Aland text, the text for John is primarily based upon Codex B / Aleph, or simply put, the Alexandrian (Egytian) text.

Yet here, the UBS text and the rest are actually relying upon a 'Byzantine/Western' text here reconstructed by Hort largely from Codex D (Bezae). Naturally, the actual count of 'de's has been skewed!

Hodges & Farstad's Majority (Byzantine) Text by comparision only has 9 'de's, and our own reconstruction only counts 10 also.

As Mark Twain would say,

"Rumours of the number of 'de's have been exaggerated."

Return to Index

John Part II:
The Signs Narrative (cont.)

Paragraph KAI Alla DEOUN
8:12-20 8 1 3 3
8:21-30 8 2 0 4
8:31-38 3 1 1 2
8:39-47 5 1 2 1
8:48-59 14 2 2 3
9:1-12 13 2 0 5
9:13-23 8 0 3 5
9:24-34 10 1 1 3
9:35-41 8 0 2 0
10:1-6 5 2 3 0
10:7-18 16 3 0 1
10:19-21 1 0 1 0
10:22-30 9 1 0 1
10:31-42 7 1 2 1

The section following the Pericope de Adultera is identical to that of preceding (chapters 4 to 7). The only remarkable feature is that again the discourses of Jesus appear to be totally independant of the frequency or even appearance of the most common connecting particles.

Return to Index

John Part III:
The Signs Narrative (cont.)

Paragraph KAI Alla DEOUN
11:1-16 5 3 7 5
11:17-27 3 1 3 3
11:28-37 11 1 3 4
11:38-44 6 1 3 1
11:45-53 9 2 3 3
11:54-57 3 1 2 2
12:1-8 3 1 4 3
12:9-11 3 1 1 1
12:12-19 5 1 1 2
12:20-26 6 0 3 1
12:27-36 8 2 1 3
12:36b-43 7 1 1 0
12:44-50 5 3 1 1
Paragraph KAI Alla DEOUN

Chapters 11 and 12 seem to mark a bridge-point in the usage of connecting particles. Not only are both 'de' and 'oun' much more frequent, but 'kai' has correspondingly gone down in frequency. Here there might be a case for either a different hand, or else a different source material in use by John. These two chapters are more similar to the Passion Narrative, and may be derived from the same source.

Note however, that in a very similar manner to the Pericope de Adultera, the first paragraph of chapter 11 has an unusually high number of 'de's (7). The author of 11:1-16 seems quite capable of a long string of 'de's, irregardless of what critics may view as 'normal'.

Return to Index

Part IV:
The Last Supper Monologue

Paragraph KAI Alla DEOUN
13:1-11 11 3 2 1
13:12-20 7 1 1 2
13:21-30 5 0 1 4
13:31-35 5 0 0 1
13:36-38 0 0 1 0
14:1-14 17 0 3 0
14:15-24 13 1 2 0
14:25-31 5 1 1 0
15:1-10 9 0 0 0
15:11-17 7 3 4 0
16:1-4 1 2 0 0
16:4b-11 5 2 5 0
16:12-15 2 2 1 0
16:16-24 15 1 3 2
16:25-33 7 2 0 0
17:1-5 3 0 1 0
17:6-19 14 2 1 0
17:20-26 8 1 2 0

The Last Supper Speechs: John 13 to 17

Here we see something that indeed appears significant. 'oun' is all but absent throughout the entire Last Supper Monologues by Jesus. It only puts in an appearance twice, out of apparent necessity.

This also seems to indicate that John is relying upon sources for the content of the Last Supper. This is a good thing, for it shows that rather than 'making up' these unique monologues (as John has been accused of doing, since they differ so much in content and detail from the Synoptics), John is here apparently relying upon a previously recorded and written speech, probably taken down as minutes by one of the Twelve. The bursts of large numbers of 'KAI' may indicate that some of Jesus' speech was given in native Hebrew rather than Greek.

Return to Index

Part V:
The Passion Narrative

Paragraph KAI Alla DEOUN
18:1-11 8 0 4 4
18:12-14 4 0 1 1
18:15-18 9 0 4 2
18:19-24 3 0 2 2
18:25-27 3 0 1 2
18:28-38 7 1 2 4
18:38b-40 2 0 1 2
19:1-7 12 0 0 3
19:8-12 3 0 2 1
19:13-16 2 0 0 2
19:16b-22 4 1 2 3
19:23-27 7 1 2 4
19:28-30 1 0 0 2
19:31-37 8 1 0 2
19:38-42 6 0 4 3
20:1-10 14 1 2 4
20:11-18 9 0 2 1
20:19-23 8 0 0 3
20:24-29 12 1 2 1
20:30-31 2 1 0 0
Paragraph KAI Alla DEOUN
21:1-14 15 1 5 5
21:15-19 4 0 1 1
21:20-23 2 1 2 2
21:24-25 3 1 1 0

With the Final Passion Narrative, we probably get something like a 'normal' usage of both 'oun' and 'de'.

Yet even here, the instances of 'oun' are probably a bit high, not because of stylistic considerations, but due to the compelling sequence of events, which John wants to present in a way that appears necessary and inevitable.

That is, once Judas betrays the Lord, the following 'steps' fall firmly into place in a rather 'predestined' manner. Now we see John's use of 'oun' multiply, yet it is for a fully functional purpose.

So again, although we find interesting things in the changes in frequency of the various particles, there is nothing that cannot be adequately accounted for by the subject matter itself, and the need to present it a certain way.

The only thing close to a 'preference' for 'oun' over 'de' in John appears in the last two chapters (19 and 20). Here it is 'de' which disappears for one or two paragraphs at a time, while the frequency of 'oun' is almost doubled from that found in the rest of the Gospel.

Yet again the answer is not hard to fathom: Here John is probably again relying upon an earlier source for his Passion material, and working it into his book to create an 'official' Gospel. In this case, that source appears to have the traits that an early textual critic like Davidson believed were representative of the entire Gospel. This however cannot be maintained in the light of our tables.

There's nothing really Johannine about it, but the Greek, although heavily Semitic, is not stilted 'translation Greek' anywhere, but simply good Greek. There are traces of possible translation, but John has done such a good job that little solid evidence remains.

Return to Index

Reasons why
'de' is Meaningless

The fluctuations in John's usage of 'de' have no relevance for the authenticity of John 8:1-11.

This will be our opening thesis. We will back this up with many good reasons why in a moment.

Failure #1

The Failure to do the preliminary work necessary to flesh out the different aspects of the problem, and the failure to hold back on premature judgements until the results of analysis of were available to researchers.

When the internal evidence for the case against John 7:53-8:11 was put together, textual critics were already largely convinced of its inauthenticity on textual grounds, and ideologically committed to its removal from the text.

That these men were biased, poorly trained and ill-suited to the task of evaluating internal evidence was hardly perceived in the 19th century. It was assumed to be a simple job, requiring little specialized knowledge or acumen.

Today with hindsight, we can see that the task is an order of magnitude more complex, subtle and difficult than previously thought.

A rather large and significant amount of evidence and research was either in an unknown or totally disorganized state, or else wholly nonexistant. Let us review just a few of the major developments completely unforeseen by 19th century critics:

(1) The discovery of the early papyrii in the early 1900's, and the beginning of their contribution to the new understanding of NT & Koine Greek, resulting in the complete rewriting of the Lexicography and Grammar of the New Testament. 1

(2) The birth of the science of modern linguistics, beginning with serious statistical methods of analysis of writing styles. 2

(3) New breakthroughs of Cryptography during the Second World War. Deep syntactic structure, semiotics and morphology was advanced dramatically. 3

(4) The discovery of the Dead Sea Scrolls in 1948, taking the knowledge and science of Hebrew and Semitic influence on the Greek language forward by an inestimable amount. A whole new period of work in this area was begun. 4

(5) The advance of computer power, and the development of tools for the analysis of language and literature to an unsurpassed degree of detail and accuracy. 5

1. See for instance A. Deissman, Light from the Ancient East (1922) Eng. Xlat. 1978 Baker Books

2. A good example of early fruits in this area is Turner, A Grammar of NT Greek Vol. IV: Style, 1976, T & T Clarke

3. Anyone who doubts the new scientific footing achieved at this time need only investigate the story of the cracking of the Enigma encoding machine.

4. One may consult with profit the summary of the history of the advances in Biblical Semitic studies found in M. Casey, Aramaic Sources of Mark's Gospel, 1998, Cambridge Press

5. An early pioneer in computers and NT Studies was Prof. J. Heard at U of T, using SNOBOL 4 on mainframes in the 70's.

All of this simply underlines the fact that whatever the textual critics thought they were doing in the 19th century, it wasn't useful science. That would have to await the linguistic knowledge and analysis tools of a different century.

Return to Index