Internal Evidence

ει μη
in John 8:1-11

Exerpted from: Nazaroo, Previously Unconsidered Evidence for John 8:1-11, Jan-Dec 2006
(thread carried on: , )

Page Index

Last Updated: Feb 21, 2009

Introduction: - Davidson's Claim
Lesson 1: - Names, Nouns, Exclusionary Modifiers
Lesson 2: - Compound Sentences
Lesson 3: - Conditional Sentences
Lesson 4: - Negative Conditionals and Debate
Lesson 5: - Various Matters and Examples
Conclusion: - 'ει μη' is not a stylistic marker

Return to Index


Davidson (1848): Internal Evidence against John 8:1-11

In 1848 Samuel Davidson published his Introduction to the New Testament, a text-critical reference work intended to introduce English academics to the latest German higher-critical research and opinion regarding the NT.

In his chapter on the authenticity and integrity of John, Davidson presented several pages of evidence regarding John 7:53-8:11. Here he gave about a page of internal evidence from the passage, examples of style and diction alleged to be unusual for John.

One of the more interesting and difficult examples of 'style' alleged by Davidson to be foreign to John is the following case. According to Davidson, πλην ("except"), which is found in our passage, is unlike the style of the Evangelist (John), and he would rather have used ει μη ("if not"). [Davidson, ibid, pg 359-360].

This claim is utterly baseless, but in order to demonstrate this plainly to the English speaking reader, we must first explain properly the meaning and usage of both of these terms in the Greek NT, especially in the Gospel of John.

The extra difficulty comes from the fact that ει μη ("if not") is not just a simple word that can be looked up in a dictionary or lexicon. Rather it is an idiomatic phrase that connects conditional clauses. That is, it is a rather complex piece of grammar, and to thoroughly understand what it is, we need to relate it to similar structures in English.

So then to have a clear understanding of what Davidson claimed, and why it is an illegitimate claim, we have to divert first into a short discussion of Greek and English grammar and usage. This is unusual, and requires extra effort on the part of the reader.

However, as a reward for this extra effort, the reader will gain a better understanding of both Greek and English usage and idiom, the mechanics of translation, and especially, the meaning of several important passages involving Christian dogma and doctrine.

To understand why Davidson made this error in the first place requires a review of the state of knowledge at the time he wrote, and the trends in scholarship as they waxed and waned in the 1900's. We have done this elsewhere in dealing with Davidson's other evidences, and so will not repeat that here.

One helpful article which we can recommend for review before delving into this complex puzzle is the following one:

ει μη Clauses in the New Testament: <--click here for this article

The reader should understand that we don't necessarily agree with everything contained in this article, as they will notice when continuing to read our work here. Nonetheless, treatments like this are useful in presenting much of the technical side of the problem of mapping grammar and idiom, and how it affects translation.

Many readers, especially experts in either English or NT Greek or both, will be able to skip much of what we present below. The material is not meant to be condescending or pedantic, but simply to be exhaustive and gather in one place all that is needed to fully understand the discussion.

This thoroughness should not be interpreted as 'preachiness' but is simply offered in a fun format so that anyone can access this topic regardless of their education or background. All that is required is some knowledge of English grammar and the contents of John's Gospel.

Return to Index

Lesson 1:

Adjectives and Adjective modifiers:

Many common nouns are too inclusive, and ambiguous in typical situations.

If I say "pass the book" in a library, it might be confusing. To narrow down which book, and exclude others which are irrelevant, I might simply add an appropriate adjective, forming a phrase: "the green book".

The adjective is the word "green", and it is not merely descriptive, but also exclusive. In gesturing to a pile of books, "pass the green book" excludes those that aren't green.

With a pile of objects on the table, "pass the book" excludes the objects that arent books, and with a pile of books on the table, "pass the green book" excludes other books.

In both cases, "the book" or "the green book" do the same job. They simply identify an object and exclude other objects that might accidentally be included.

These groups of words are called phrases, and they simply act like names, or nouns. In fact they have a special name, "substantive phrases".

Phrases aren't sentences. They don't really describe actions or tell a story by themselves. They just identify something, a person, a thing, or even an idea.

"Except" and "But": Words that narrow meaning by exclusion

As well as simple adjectives, like "green" we can have other ways of excluding the objects we don't want to refer to:

"Pass all the books except the red ones."

This is indeed a sentence, but the last seven words (underlined) in it all act as one single name, or "substantive phrase". The sentence really just says, "Pass something." and that something is a group of objects, identified by excluding others.

That group of objects could be a bit ambiguous, so an adjective modifier or "adjective phrase" (another group of words) is added to exclude some objects we don't want: "except the red ones".

This particular "adjective phrase" is an exclusionary phrase: it excludes some objects from the group we want to talk about. This is done with a special connecting word, "except". This word is normally used just like it is here, and quite often, to build a more specific name out of the common words at hand, by adding an excluding phrase to the name or substantive phrase.

We also commonly use another connecting word in just the same way:

"Everyone but John went to Jerusalem."

Here again, 'everyone' would be too ambiguous, so an exclusionary phrase, "but John" is added to show who isn't in the group.

"but" is the connecting word making the following phrase exclusionary. This is just like the case with the word "except".

Both words can connect an excluding phrase to a name, making it more specific, and both words are often interchangable in ordinary English.

*This usage of these connecting words is simply to join two phrases together, to make one big (and less ambiguous) phrase, which acts as a noun, or name.

But the group of words formed by these connections are just bigger phrases. They are NOT sentences in themselves: they will be used as Subjects or Objects later in a larger structure which has a verb, called a clause or sentence.

*KEY POINT: Notice that in Greek the word for joining phrases to give the exclusive meaning is "plen" πλην ("except", "but"),. The other popular word with this function is "αλλα" αλλα ("but").

"plen" πλην ("except", "but") is actually a general purpose word, most easily rendered 'but' in most contexts in the Synoptic gospels, where it appears dozens and dozens of times.

A good example of its use as an exclusive modifier is in Acts 8:1, where not much else other than "except" will fit as a translation:

"...and they were all scattered abroad throughout the regions of Judaea and Samaria, except the apostles."

... παντες δε διεσπαρησαν κατα τας ξωρας της Ιουδαιας και Σαμαρειας πλην των αποστολων. (Acts 8:1)

Allthough this is statistically rarer as a usage generally than πλην as a more general connector of contrasting clauses, it is perfectly normal, and good Greek. The statistical preponderance in the NT really doesn't reflect normal usage as such, so much as it reflects the subject matter, namely the large sections of Jesus' speechs in the Synoptics, where it mostly appears.

Note these sections are to help ordinary people follow what is being discussed. There is no point in critiquing these sections for their lack of technical terms, or some simplifications and inaccuracies regarding 'elipsis', 'apodosis' or 'optative moods'. The whole idea is to lay out the basics without introducing a whole pile of needless technical terms. - Nazaroo

Return to Index

Lesson 2:

Simple Clauses and Compound Sentences:

We call a simple statement with a single Subject and predicate (and one basic verb) a 'clause'. This is one step up in size and complexity from a 'phrase', discussed in the previous section. (Recall that 'phrases' are just complicated names of people, objects, places and ideas.)

"The disciples went to Jerusalem." is a simple sentence, i.e., a "clause".

"Jesus stayed behind." is another simple clause.

If we join two clauses together, we get one long sentence:

"The disciples went to Jerusalem, and Jesus stayed behind." - This is a "compound sentence".

We can connect two clauses with a variety of words, each with slightly different nuances, but all essentially doing the same basic thing:

"The disciples went to Jerusalem, but Jesus stayed behind."
"The disciples went to Jerusalem, yet Jesus stayed behind."
"The disciples went to Jerusalem, while Jesus stayed behind."

This allows us to relate two statements or contrast them, in telling a story.

In Greek these functions are handled similarly, with a few connecting words, like
"kai" και ("and, but") "de" δε ("and, but") "αλλα" αλλα ("but") (stronger form).

*KEY POINT: Notice that in Greek these words for joining clauses are different than the word used for phrases - "plen" πλην ("except", "but"). (In English there is an overlap in the use of 'but' and even 'except' because these words can do both jobs. English is less precise than Greek in this respect, but both have overlaps.)

Return to Index

Lesson 3:

Conditional Sentences:

Of course if all we could do was make flat statements, stories would be boring, and we would not be able to do much reasoning or science. But we can also ask questions, make exclamations, and most importantly make hypothetical statements and proposals. This allows us to express certain facts and 'truths' which are more abstract, not based on what did or didn't happen, but what 'might' have happened, and what we know about what the consequences might be if this or that happened.

And so we come to 'IF-statements', or Conditional Sentences. A Conditional Sentence has the basic skeleton-form of:

If __(STATEMENT A)__ , then __(STATEMENT B)__.

A Conditional Sentence is a COMPOUND sentence, made up of two clauses, each with a verb. Each clause is a statement of its own right, and asserts an action or state.

On the 'IF' side, we put a hypothetical idea, perhaps something that did happen, or could happen. The actual truth of the statement, whether or not it is a historical fact, either at the time we say the sentence or later, is for the moment irrelevant and put aside.

On the 'THEN' side, we put what we know about the consequences of the statement found for the 'if' side. Let's see how two people can share information that neither of them possesses when they start the conversation(!):

Disiple A: "Did John go to the market?"

Disciple B: "I don't know. But, IF John went to the market, THEN Judas went to the market. I know this because Judas has the money pouch."

Disciple A: "Well, I know this: IF Judas went to the market, THEN John went with him. I know this because Judas said he would not go without John. "

Disciple B: "Well if so, then John went to the market. I know this because Judas left with John just now."

Using some 'if-statements' (conditional clauses), the disciples were able to share their partial knowledge about the situation and combine it to arrive at a conclusion regarding the actual facts. They were able to do this by considering some hypothetical cases and their consequences.


Notice that in the third bold Conditional Sentence above, the word 'so' appears. It is not a real clause, but acts as a 'place-holder' for the clause that is implied:

"If so (if it is true that Judas would not go without John), then..."

This 'dropping out' of various ideas, like verbs, or even whole clauses, is a natural time-saver in conversational speech. It is called 'elipsis'. (a skip or hole).

KEY IDEA: A whole clause (with verb) can be implied, even when it is actually left out, and this unspoken clause will be assumed by the reader or listener familiar with these shortcuts and conventions.

Greek Conditional Sentences:

In Greek, the 'if - then' conditional Sentence uses the following words:

ει (if) __(hypothetical statement)__ ουν (then)__(consequence)__.

"Lord: If he sleeps, (then) he does well." (John 11:12)

" κυριε, ει κεκοιμηται, σωθησεται. " (John 11:12 Greek text)

In both Greek and English, the coordinating word "then" is optional, and is often omitted when the structure of the conditional sentence is clear.

Return to Index

Lesson 4:

Negative Conditional Sentences and Debate:

The most common occurance of the Conditional Sentence in the New Testament is in dialogue (or monologues like Epistles) during an argument or debate, or to make a point. In this case, the speaker will make an assertion, and then justify it, or prove it by using a NEGATIVE Conditional sentence. The form is a three-part statement, a coordinated 'sentence' made of three clauses.

Here the basic structure is this:

(1) ___(Factual Assertion)___ :

(2) For___(Counter-factual Consequence)__ ,

(3) if not ___(Previous Assertion)__ .


(1) "Rabbi, we know that you are a teacher come from God:

(2) For no one is able to do these signs,

(3) if God were not with Him." (John 3:2)

The speaker makes an assertion, a claim, or a statement of fact. Then he proves it by pointing to an obviously missing consequence that would result if the first statement he made were NOT true. The lack of any evidence for the alternative shows that the first assertion is true.

The Negative Particle:

In English, the 'negative particle' ("not") floats around and is most comfortable either immediately preceding the verb, or inserted between the parts of a 'compound' predicate.

"John would have no money, if Judas did not go with him."

Here we see the word 'not' stuck between the two verb-parts (underlined). Again in English (only), since it is not highly inflected, many special forms like past and future tenses are made by combining a verb with the verb 'to be' or 'will/shall'.

In Greek however, this is not the case. a typical verb has all the information packed into its form.

The negative particle 'mh' ("not") tends to stick to the word 'ei' ("if") and both sit between the two clauses being joined together. At most, sometimes the word 'de' ("but") by force of its own natural place in a sentence in Greek will insert itself between the two.

Greek Form of Negative Conditional:

So typically in Greek also we will see a three clause grouping:

Clause 1: (leading thesis) "_(assertion A)_:

Clause 2: (counter argument) ει μη [if not]__(assertion A)__,

Clause 3: ουν [then]__(counterfact)__."

Just as in English, the opposing counter-assertion is often just left out entirely, by 'elipsis', but it is still there in force, being assumed by the speaker and hearer. It is often represented in English by a simple comma!:

"In my Father's house are many homes.
But if not, (that is, "if there were not many homes there," )
(then) I would have told you." (John 14:2)

" εν τη οικια του Πατρος μου μοναι πολλαι εισιν.
ει δε μη [elipsis of counter-assertion],
ειπον αν υμιν!"

Often, it is not the whole clause that is left out, but just the main predicate, and the Subject (a noun or phrase) of the missing clause functions as the placeholder:

"And no one ascends to heaven,
if not the One descending."
(that is, "if the One descending does not ascend.") (John 3:13)

" και ουδεις αναβεβηκεν εις τον ουρανον,
ει μη ο εκ του ουρανου καταβας

Here the phrase is a placeholder for the entire clause or thought.

Almost any part of a complex clause or sentence can stand in for the whole thing, depending upon the emphasis or focus. Just like English, Greek is equally flexible and similar in the thought-construction:

"No one comes to the Father,
if not through Me."
(that is, "if they do not come to the Father through Me")
(John 14:6)

" ουδεις ερξεται προς τον Πατερα
ει μη δι εμου

Here a secondary modifying phrase for the action is standing in for the whole complex sentence. Why? Because this is the essential point and focus of the argument at hand.

We have seen that although the expression 'ει μη' is common in the gospel of John, it almost invariably appears in dialogue, specifically during an argument, debate, or special instruction or teaching. Its function is to coordinate whole clauses to create conditional statements.

KEY CONCEPT: In Greek and English, the Conditional Sentence, and especially the Negative form, with 'if not' almost always appears in dialogue or monologue, not narrative! It is used in argument and debate to express hypothetical conditions.

The Optative Mood:

In a purely hypothetical statement, the speaker can emphasize the hypothetical-ness or falsity of the statement by using a rare 'optative' form of the verb. In English, we use the plural form of the past-tense for this job.

We might say,

"Were John going to the store, he would have taken Judas."

Here people use the plural past tense of the verb 'to be' to indicate the actual falsity or hypothetical status of the assertion in the (in this case abstract) 'if' clause.

Notice that the word 'if' is actually left out completely here. (even 'if' is optional!) The full version is:

"If John were going to the store, he would have taken Judas."

This is an English idiom only, since English doesn't really have an 'optative mood' for verbs, although we accomplish the function of one. This is because English is a much less 'inflected' language, where things like idioms and word order take the place of special verb forms.

Return to Index

Lesson 5:

Excursion into Various Matters

This next post will gather up some miscellaneous matters that naturally arise in the discussion of the grammar of these passages.

(1) Translation and Dynamic Equivalence

Some readers may already have perceived quite a difference in the renderings for the examples we have been discussing under Conditional Sentences. An explanation is in order.

The differences for the most part reflect that fact that here when carefully examining grammatical issues we will almost exclusively render the Greek components as 'literally' as possible. The purpose here is (1) to directly convey the physical form of the Greek structures as well as the meaning, and (2) to do as little 'interpretation' as possible. It is a 'ground up' process.

Specifically, those familiar with the KJV etc. will notice that these translations render the particles "ει μη" ('if not') frequently as "but" or "except", and occasionally as "unless".

This is partially the result of the act of "dynamic equivalence" in translation / interpretation. Here the translators feel that in many cases the literal "if not" gives a more clumsy appearance, and more common English idioms can safely be substituted without loss of meaning, and give a more natural expression.

These arguments are often strong and appealing, but we should be cautioned that they still involve an element of assumption, interpretation and sometimes modification of meaning. That new meaning may not be present with the same certainty as what a more literal rendering offers.

Sometimes a deliberately more inclusive, more general, or more ambiguous choice of words has been made by the original author or the Holy Spirit for a good reason. There is always some compromise or loss of meaning in translation. Here in the minute study of grammar we want to minimize the introduction of artifacts or misleading effects as a result of too free a rendering.

(2) The NON-Equivalence of "if not" and "except"

A really good example of the potential loss of meaning can be shown in the case of the word "except" as a substitute for "if not".

There is one often critically important piece of meaning completely lost in the substitution of "except" for "if not" in a Conditional Sentence:
The sentence is no longer conditional!

It actually becomes a flat statement of fact. This can often be unimportant, where it simply hides a Greek idiom or form of expression.

However we should never lose sight of the fact that a Conditional Sentence in Greek or English is NOT a statement of fact.

Its a hypothetical statement presented to the reader for consideration. In order to consider the hypothetical truth of a Conditional Sentence, one's current belief or knowledge of the historical facts must often be suspended entirely.


Sometimes converting a Conditional Sentence can add things or imply things that are simply not true, and are not present in the original Greek:

"A prophet is not without honor,
except in his own country, and in his own house."
(World English Bible xlat., Matt.13:57)

However the Greek only says,

" ουκ εστι προφητης ατιμος,
ει μη εν τη πατριδι αυτου και εν τη οικια αυτου."
(Matthew 13:57)

"A prophet is not without honour,
(even) if he is not (honoured) in his own country and in his own house." (Literal)

Falsely converted from a conditional sentence to a statement of 'fact', this could be mistakenly applied as a prediction or a test for true prophets: i.e., A true prophet is always rejected by country and family. But this is not what the Greek implies at all.

Again here:

Those whom you have given me I have kept. None of them is lost,
except the son of perdition,
that the Scripture might be fulfilled. (Jn 17:12)

ους δεδωκας μοι εφυλαξα, και ουδεις εξ αυτων απωλετο,
ει μη ο υιος της απωλειας
ινα η Γραφη πληρωθη
. (Jn 17:12)

"Those given to Me I have kept, and no one out of them is destroyed,
if not the son of destruction,
in order that the scripture might be fulfilled."

That is, "I have kept all those you have given to Me, and not even one of them is destroyed, if even Judas has not been (truly) destroyed, in order that the scripture might be fulfilled."

In the original Greek there is still a chance for Judas to repent and be saved, at least at this point in time in the narrative. While this may not be an exciting or welcome discovery for those who have already written off Judas, other cases can be more severe.

Often the conversion of a Conditional Sentence into a flat statement can actually introduce a contradiction into the translation, where there was none in the Greek.

"Have ye not read what David did, ...
how he ... ate the showbread,
which it was not lawful for him to eat, ...
but only for the priests?" (Matt. 12:3-4, Amer.Stand.Ver)

Here the ASV (along with others) has Jesus actually contradicting the O.T. Law, which permitted not only priests, but their wives and children to eat the showbread, i.e., any descendant of Aaron or relative by marriage. (Obviously the families of priests also lived off of the temple offerings. For instance, see Lev. 22:11-13: There could be no laws regulating women eating dedicated food if they were actually forbidden to eat it at all times.)

Left as a Conditional Sentence however, Jesus' statement makes perfect sense without contradicting the Law:

ουκ ανεγηωτε τι εποιησε Δαβιδ,...
πως ... τους αρτους της προθεσεως, ...
ους ουκ εξον ην αυτω παγειν, ...
ει μη τοις ιερευσι μονοις ;
(Matt 12:3,4)

Have you not read what David did?...
How... of the showbread he did eat,...
which is not permitted for him to eat,...
(even) if not for the priests only?"
(i.e., "even if (it was) not just for the priests to eat? " ) (Literal)

That is, the Greek has Jesus correctly pointing out that David's act was not permitted by law, not because only priests could eat the bread, but because David was not a descendant of Aaron, or a family member. There is no need for a rendering which makes Jesus look ignorant or inaccurate in His own speech.

Nor is it always a 'minor' doctrine: Consider the well-known but often misunderstood verse here, commonly rendered as a flat unconditional statement:

And he said to him, Why callest thou me good?
there is none good but one, that is, God:
(Matt.19:17 Noah Webster Bible)

This rendering has Jesus inadvertantly exclude Himself from the class of those who are 'good'. (Unless you already know Jesus is God, you will assume He is admitting he is a sinner.)

If this is what the Rich Young Ruler thought he heard, then one can understand why he walked away after being asked to hand over his fortune to a confessed sinner!

But the Greek offers no such excuse for a dispute between Christians and others who have lower christological viewpoints:

τι με λεγεις αγαθον?
ουδεις αγαθος ει μη εις ο θεος
. (Matt 19:17 Greek)

Why do you call me good?
No one is good, if not One, God. (Literal translation)

Here Jesus is not saying that other beings cannot be good, (which would leave him contradicting Luke and others) but rather that the source of all goodness is God, including His own goodness. This corrected translation removes at least two difficulties from the typical English rendering.


Modern idiomatic or "dynamic equivalent" renderings of these verses that actually preserve the intent of the Greek without adding anything might run like this:

(1) "Whether or not a prophet is dishonoured in his own country or home, he is not without honour."

(2) "If even the Son of Destruction, Judas is not yet lost, then I have lost none of those given to Me, so that the scripture is fulfilled."

(3) "Didn't you read what David did? How he ate the showbread? That wasn't allowed for David, even if others besides the priest could eat it."

(4) "Why are you calling me good? If God isn't good, then no one is good!"

Key Point: "except" and "if not" are not equivalent renderings of "ει μη" and important information can be lost or added by a poor choice in translation.

The phrase "ει μη" is not a simple stylism, but is used carefully and intelligently by all the Evangelists, including John.

Return to Index


πλην ("except")

We illustrated both the form and function of πλην ("except") in 'Lesson 1' above. There we mentioned that this word is used almost exclusively to connect nouns or substantive phrases to form bigger ones, which then function as exclusionary identifying substantives. πλην ("except") seems never to be simply used to coordinate two clauses. Even when a clause is involved, the verb is usually in a participle form, and so the whole clause functions as a substantive identifying name.

αλλα ("but")

The often synonymous αλλα ("but") also can function to connect substantives to form larger identifying word clusters, but has other meanings and performs a wider range of functions. αλλα ("but") is truly the 'general purpose' word here. This explains fully why it is far more frequent in all the NT writings including John than πλην ("except"). We did not illustrate the usage of de ("but, and"), but actually the word αλλα has more connection and synonymous usage to de than it does to πλην. Along with kai, these two are often interchanged in parallel passages and in manuscript copies by scribes.

ει μη ("if not")

The coordinating connecting cluster ει μη ("if not") has a totally different function. It is used exclusively to connect multiple clauses to form Conditional Sentences. ει literally means "if", and combined with μη ("not") does exactly what its counterpart phrase in English does. The 'moveable "not", which is sometimes placed between two verb parts is just a morphological quirk of English usage, and has no lexical content. The usage and form of this word cluster took longer to illustrate, because of its entirely different grammatical function (Lessons 3-5). In many cases, ει μη can be translated idiomatically into English as "unless", which is a syntactic word that has taken on this specialized meaning.

In some cases, the word "but" can be substituted for "if not". This always causes the loss of the original Conditional structure. The result is that unintended meanings or implications to the original can be added, without authority or warrant from the original text. In some cases, this is harmless, because the context or additional information indicates the added 'definiteness' is plausible or correct. But as a rule it is unreliable and misleading to translate ει μη as "but" or "except" without additional warrant provided from the text.

Davidson's Alleged Internal Evidence

Let us return now to Davidson's claim that John would have used ει μη instead of πλην, which is found in the passage at verse 10:

"And Jesus seeing no one but the woman..." (8:10, some MSS)

Ισους και μηδενα θεασαμενος πλην της γυναικος

This is obviously a simple exclusionary phrase, meant to qualify and limit the 'no one' to others, exempting the woman still standing there. There is no Conditional Sentence here, nor is there even two clauses to coordinate. ει μη simply cannot stand in this verse as it has come down to us in any version.

Davidson has completely failed to understand the completely different function of the two grammatical constructs. This betrays that he is largely relying upon the opinions of others, and lacks the basic grasp of NT Greek to evaluate the case himself.

It is simply not a stylistic option at all, and there is no freedom to choose between ει μη and πλην in a case like this. We can say categorically that John would not, and could not have chosen ει μη instead of πλην. This would have resulted in a nonsensical construction.

What about αλλα?

The only other word which actually could have a synonymous function here would be αλλα, although this would be a more unusual usage. Of course a superficial word-count naturally shows that John, like all other NT writers, uses αλλα more often than πλην. But this is a meaningless statistic, since all the other Evangelists including John obviously know and use the word πλην. There is nothing here in the passage that indicates John would favour the more awkward αλλα here, for the comfortable πλην, which he quite competantly uses elsewhere.

Technically speaking, " ει μη" actually does appear in the narrative although very rarely. However, it is so much less frequent than either "πλην" or "αλλα" which on the surface appear to have a similar function, that it begs for a closer look. (in fact, αλλα appears almost 700 times in the NT, while πλην only occurs 30 times, and ει μη is also relatively rare. There are about 500 'if'-statements in the NT, with about 10-15% of those being in conjunction with the negative particle.)

The one possibly significant case of "ει μη" in the Johannine narrative, is in John 6:22:

"...there was no other boat there,
if not simply that one into which His disciples had entered.' (John 6:22)

" πλοιαριον αλλο ουκ ην εκει,
ει μη εν εκεινο εις ο ενεβησαν οι μαθηται αυτου
" (Jn 6:22 Greek)

Here I think the explanation for John using such an unusual phraseology is the complication of the description of the boat. He is compelled to say there was a boat there, and so this requires almost an explanatory aside to explain why it was still an obvious miracle even so. Thus he carefully enumerates the boat as the one the disciples were in, which everyone knew Jesus had not entered upon its departure.

Note that if John had simply used "πλην" or "αλλα" it would problematically imply that the miracle was dependant upon the witness of the people, which John is not interested in credencing. It is his own testimony that he wishes to provide. While converting this to a flat unconditional statement makes for a smoother English rendering, I think something subtle is actually lost here. If John had simply wanted to say "no boat except the one the disciples had entered" he should have done so. Here we have the paradoxical case of a Conditional Sentence being more definite as to the reality of the miracle than an unconditional flat statement about the observation of witnesses!

No 'but's about ει μη:

In any case, it turns out that John uses 'αλλα' about 20 times as often as 'ει μη'! (1:8, 1:13, 1:31, 1:33, 3:8, 3:16, 3:17, 3:28, 3:36, 4:2, 4:14, 4:23, 5:18, 5:22, 5:24, 5:30, 5:34, 5:42, 6:9 etc.)

But what can we take this to mean, except that Samuel Davidson's assertion that John was more likely to use ει μη than πλην was based on pure smoke! If any argument could be made, here, it would be that John might have been tempted by habit to use 'αλλα'

Once again the 'assurded results of textual criticism' based upon expert judgment is apparently more like a bad guess than a good intuitive knowledge of the author's style and habits.

Key Point: John would not, and could not have chosen ει μη instead of πλην. This would have resulted in a nonsensical construction. Nor is there any other plausible synonym that John would have preferred here. Davidson has no case here for any kind of credible internal evidence of 'style' against Johannine authorship. QED

Return to Index