Verse of the Week - Ephesians 5:21

 This week's variant is an interesting one.

Many Greek texts, primarily Byzantine and TR, read:

21 ὑποτασσόμενοι ἀλλήλοις ἐν φόβῳ Θεοῦ. (Hodges Farstad, GNT Acc. to f35, TR)

While most others read

21 ὑποτασσόμενοι ἀλλήλοις ἐν φόβῳ Χριστοῦ. (Robinson Pierpont, all Critical texts)

The story

I first came across this variant in a stream by Dwayne Green (YT channel) where he mentioned the variant amongst others. This piqued my interest and I decided to look into it. The next thing I knew I had checked over a thousand separate Greek New Testament, medieval commentaries, Latin Vulgate, and 'Vetus Latina' (Old Latin) manuscripts and compiled a spreadsheet in Excel to track the variant from the second century AD to the 18th century AD.

The result was, to say the least, interesting.

When it comes to determining the 'originality' of a variant there are a variety of approaches:

• Those who in the 'Alexandrian' (Category I, II) camp generally favour the earliest, most difficult readings, often excluding the majority readings.
• The TR only camp will generally ignore anything not found in the various Textus Receptus editions.
• For the Byzantine Priority camp the majority reading is most commonly the reading favoured over any reading agreed to be 'Alexandrian'.

Obviously, this I am painting with a broad brush and this is not meant to demean any of the positions, however, this is the broad approaches of each.

So how would each determine the validity of this one?

I've already basically given the answer away above. For the most part the Byzantine and TR camps choose θεοῦ over Χριστοῦ.

θεοῦ or Χριστοῦ?

I started my research on this variant by looking at the CNTTS apparatus as it is one of the most comprehensive apparatuses available. It's not perfect, I've found errors, but for the most part it's highly accurate and often has more GNT witnesses recorded than the UBS Apparatus. This began to paint a picture that I didn't expect. However, there were not enough witnesses and so I went to the list of Greek New Testaments on Wikipedia and one by one opened them up in the Muenster Virtual Manuscript room to look at this verse and noted the variant on a spreadsheet I created for this project.

I started this research with about 40 MSS and ended up with 668. I also found several new variants that had not been noted (though perhaps they had been found before me). These were κυ, κν, θεου, θω, and Omit NS. 

When we look at the following results table we see my findings:

At first glance it would appear that the correct variant is θεοῦ - it is found in about 65% (rounded to the nearest whole %) of all catalogued and available manuscripts while Χριστοῦ is found in a mere 35%.

It is quite clear cut, it has to be θεοῦ right? It is found in the vast majority of the witnesses, so it must be that one.

Well, you would be wrong if you thought so.

When we look at the next table, we can see a clear pattern forming (click to enlarge):

So, what are looking at here? Simply put, there is a pattern that I believe means Χριστοῦ, the minority reading, is the true original meaning. How so? Well, when we look at the above table we see that Χριστοῦ first appears in the 3rd century AD (III). It is first found in a papyrus P46 in the form of the Nomina Sacra χρυ. There are three other papyri (49, 92, and 132) with Ephesians, but they are lacunose (missing) in 5:21 and are of no help.

Next, in the 4th century (IV) we see 2 witnesses with χυ a slight variation on P46's form. These are the manuscripts Codex Sinaiticus (GA 01) and Codex Vaticanus (GA 03). Moving to the 5th century there are two more manuscripts, Alexandrinus (02) and Freerianus (aka Washingtonianus, 016). In the 6th (VI) century there is a single witness, Codex Fuldensis (F), one of the earliest Vulgate manuscripts and one of the earliest complete Diatessarion manuscripts (a Gospel Harmony by Tatian).

Unfortunately, the 7th (VII) century is barren of any witnesses. However, with the founding of some of the earliest monasteries we see witnesses beginning to appear again. In this case it is another Latin witness, Codex Amiatinus (A), the earliest known complete Latin Vulgate.

In the 9th century (IX) a spate of new variants appear, these being ιυ χυ, ιυ χρυ, κυ, and κν. These variants are found only in 6 manuscripts, some being the Latin of a Greek/Latin diglot or interlinear. Apart from singular appearances in the 10th (κν), 11th (κυ), 12th and 13th (χω) these variants disappear from the record and can be safely dismissed as intrusions.

It is not until the 10th century that θεοῦ first appears in 22 manuscripts. The next century there are 52 and it reaches a peak in the 14th century (XIV) after which it tapers off until the 18th century (XVIII) when some of the last handwritten manuscripts were copied.

Through these centuries the variant χριστοῦ continues steadily and reaches a peak of 35 in the 12th and 13th centuries and is last seen in the 17th century (XVII).

Categories, Commentaries, and Translations

In the image above you can see another pattern clearly laid out. Apart from appearing in nearly every century from the 2nd to the 17th, the variant χριστοῦ also appears in literally every Aland Category. It's found in 18 uncategorised witnesses, 102 of unknown category, 17 commentaries, Latin Vulgates, and three slightly different variants (χρυ, ιυ χυ, and χυ ιυ) in Vetus Latina manuscripts.

In contrast the variant θεοῦ is found only in Category II (2), III (8), V (81), uncategorised (32), unknown category (250), and commentaries (75). Unlike χριστοῦ there are no Alexandrian, Western, or Latin witnesses.

Conclusion

It is true that θεοῦ is found in the majority of the manuscripts. But the evidence that stands against its originality is the fact that there are no witnesses existing before the 10th century. It suddenly appears and then explodes in numbers as it is copied in manuscript after manuscript, mostly in monasteries on Mount Athos.

Χριστοῦ is found in manuscripts ranging from Egypt in the south to Anatolia in the north east to Italy in the West. It is found in manuscripts copied in Egyptian monasteries such as St Katherine's, by individual scribes, and even 7th century British monks (Amiatinus). The categories range from pure 'Alexandrian' to 'Free' to 'Western' to 'Byzantine'. It is also found in every century from the 3rd to the advent of printing.

Based on the evidence, we can only conclude that χριστοῦ is the original reading. True, it forms only one third of the known witnesses but the centuries long consistent appearance of this variant across all categories versus the late appearance of θεοῦ in the 10th century, seven centuries after χριστοῦ, and the restricted categories weigh strongly in favour of χριστοῦ.

Unless one is working with strictly contained families of manuscripts such as the Textus Receptus or Family 35 there is no reason to accept θεοῦ as being the original reading.

Why Θεοῦ Though?

We could understand perhaps why θεοῦ appeared - in all other verses where we are told to 'fear' someone it is God (θεοῦ) who is the subject. But, in Ephesians 5:21 it is Christ (χριστοῦ) we are told to fear. Given the number of verses telling us to 'fear God' it is almost reasonable to think that the verse specifying Christ instead of God might be an error and that χριστοῦ should be 'corrected' to θεοῦ. If you are a monk copying a manuscript in a monastery and you don't have access to a computer and a spreadsheet app nor to the hundreds of copies of Ephesians, there is no way to check which is correct. You have to either stick with what is written your exemplar or make an executive decision to change the reading to what you think it should be. It seems that at least one monk made this decision, wrongly as it turns out.

But, we can be confident that χριστοῦ is in fact the correct reading and that, barring the discovery of a first or second century manuscript of Ephesians reading otherwise, was the name used in Paul's autograph of Ephesians.

If you are feeling a bit nerdy and want to dig through a spreadsheet of my findings, you can find it on the Downloads page. If you are using MS Excel the second tab is filled with check boxes while in LibreOffice Calc (and maybe others) these are shown as either 'TRUE' or 'FALSE', but the data doesn't change.