Meanwhile, reporting on the manuscript evidence in favor of “washing the robes of them” (plunontes tas stosas auto) or “keeping the commandments of him” (poiountes tas entolas autou).
The spread of witnesses is pretty wide (unfortunately, since that makes it proportionately more difficult to weigh for which is the original text).
I haven’t been able to find any evidence from either side that either translation can be traced back to the several papyrus fragments we know to exist. This probably only means that not enough of any papyrus has survived to be counted as a witness to the text, one way or another, in that place.
The two phrases look only trivially similar in written Greek, so it’s practically impossible that this is a mere copy error one way or another; other typical mere copy error explanations can be similarly ruled out, though I’ll skip the technicalities there.
The two phrases sound more similar in spoken Greek than they look in written Greek. Since scriptographies tended to operate with a room-leader reading from a text and workers copying what they heard, sometimes errors slipped in that way (from mishearing similar sounding words or phrases). I’m inclined to think that the differences aren’t that similar, though.
ploo-OHN-tehs tahs STAW-sahs ow-TOE
POY-ee-OON-tehs tahs EN-toh-lahs ow-TOO
This leaves over two plausible explanations for the variance, which I’ll get to later.
Next, listing the known text witnesses (so far as I can find them). For each text category, I’ll list the estimated (or known) century date of the text’s production. Occasionally estimates are lateX/earlyY. If an exact year is known and accepted in the field, I’ll mention it in parenthesis.
Texts in favor of “washing the robes of them” (or some close variant, of which there is at least one):
Unical Greek texts (tend to be older, before the invention of small letters and spaces between words; or later very conservative imitations of the style): 4th, 5th, 9th/10th,
Miniscule Greek texts (tend to be younger, a lot more in existence than unicals): 11th, 12th (1107), 13th (minor grammatic varient), 13th. (Plus about 11 of the very many Koine miniscules later than this.)
Vulgate Latin: 4th/5th
Old Latin: 9th
Ethiopic: 6th
Coptic Sahidic: 4th
Fathers: Ambrose, late 4th (397?); Apringius Pacensis early-mid 6th; Fulgentius early 6th (527? 533?); pseudo-Athanasius (most copies have a grammatic variant)–but no general consensus has come forward for dating his work (I mention him only for completeness, of course).
Texts in favor of “keeping the commandments of him”
Unical Greek: 10th
Miniscule Greek: 12th (but “supplied”, i.e. written into the text by someone substantially after the original 12th century production, according to handwriting, letter-style, ink, etc.), 15th, 15th, nearly all (with about 11 exceptions) of the extremely numerous “Koine” group (dating from the Byzantine Empire to the present day).
Old Latin style (not the Vulgate): 13th
Syrian: 6th, 7th
Bohairic Coptic: 9th
Church Fathers: Tertullian (with a slight variation), early 3rd
Assessment of textual spread: “keeping the commandments of him” (hereafter “keeping”) scores the earliest known date of witness (early 3rd), but doesn’t do so well after that. Moreover, he is writing in Latin, and (in keeping with his writing style) may have been paraphrasing what he thought the text meant.
From 4th through 7th centuries,“keeping” can only post witnesses in two Syrian texts; whereas “washing the robes of them” (hereafter “washing”) scores two unical Greek, one early Coptic, one Ethiopic, three Fathers (all Latin), and the Vulgate Latin tradition generally. No less than five (and maybe seven) of these witnesses predate the earliest surviving Syrian witness.
8th through 10th centuries, both variants have a unical witness (the first and only unical text for “keeping”; whereas the text with “washing” has some indication of being earlier); and both score a 9th century copy (Coptic for “keeping”, Old Latin for “washing”).
The witnesses significant for reconstruction purposes after the 10th century are all in miniscule Greek, with “washing” starting off somewhat stronger than “keeping”, and then fading out as “keeping” later achieves final dominance in the Koine tradition.
Tertullian’s “keeping” reference is clearly the odd data point out, but probably at least testifies to a respected interpretation (if not exactly a translation) of the text. Insofar as surviving evidence goes, “washing” is otherwise far more widespread and numerous until late in the first millennium, at which point “keeping” achieves something like a parity–except in Greek, although later “keeping” grows to strongly dominate the Koine tradition (even manually “fixing” a text to comply) with only a few surviving outliers of “washing”.
Based on the surviving textual spread, the evidence strongly favors “washing” being the original text, with “keeping” being a popular doctrinal interpretation that eventually took over the tradition.
It may also be noted that, at more textually settled places in RevJohn itself (12:17, 14:12), the author shows a preference for using the phrase “terein tas entolas” rather than “poiountes tas entolas” when talking about faithful people keeping the commandments of God. (The Textus Receptus agrees with this, too.)
Internal style thus also weighs (though not conclusively, of course) toward “keeping” not being the original text at 22:14.
Textual critics make a point of keeping in mind the cultural emphases of the scribes doing most of the copying of scripture in the ages before printing. These were typically monks living according to monastic rules of varying strictness; and monastic traditions, especially in the East, have long had a documented tendency to emphasize the notion of earning salvation by personal discipline. (Definitely a key theme in the Desert Fathers I have studied over the years, as I can attest to myself!) There would be a natural tendency to favor a notion of earning access to heaven by discipline, when choosing between textual variants to follow and promote, even aside from the question of the external evidence of dating and spread–and “keeping” can certainly be more easily understood along that line than the “washing” variant (which thematically depends on accepting the freely given grace being first and authoritatively provided by God, although still acknowledging human responsibility.)
Finally, we come back to the question of replacement method, most of which can be (for various technical reasons) eliminated from consideration. There are two plausible options left over for transpositions of this type.
The first option is that the two readings witness either to an early Aramaic/Hebrew original, or at least to early expectation that the author, even if he composed originally in Greek, was familiar with thinking and teaching in an Aramaic idiom. Double-meaning puns are popular in that idiom, especially among Jewish rabbis. RevJohn shows some interesting evidence elsewhere of at least some prevalent families of the Greek texts being based off written Syriac/Aramaic; however, I have not yet been able to locate any evidence that an underlying original Hebrew-Aramaic phrase at 22:14 could easily mean either or even both of the variant phrases (where the Greek was mistranslated to mean one instead of the other, or where the author actually intended BOTH meanings but translators had to pick one or the other because the double-meaning didn’t hold up in Greek). Moreover, this kind of explanation tends to result in equally strong textual witnesses across time (until one or the other perhaps achieves dominance), and that doesn’t fit the existent manuscript evidence for this verse. Consequently, while I cannot rule out this option, it’s worth checking to see if the other option makes more sense.
Phrase or term replacement commonly occurred when comments or marginal glosses were added in the margins to scripture, without sufficient indication of whether this was only a comment on the text or a more accurate reading which ought to replace an error on the page. The sufficient differences between the two phrases lend themselves (though not altogether conclusively, due to some real if also trivial similarities) to this method of replacement. As it happens, we have direct evidence of the earliest Greek miniscule in favor of “keeping” having been corrected to “keeping” by just such a manual override. This can only weigh (so far as it goes) in favor of “washing” having been replaced by “keeping” through this method. The spread and progression of surviving texts fits this profile rather than counting against it as well.
External manuscript evidence; internal stylistic evidence; cultural proclivities; and modes of replacement: all weigh together (with varying strength, the external evidence being strongest) toward “washing” being the original phrase.
Now back to the original discussion. 
)
agree
(For which there is also abundant testimony, not least Rev 2:1-7.)
(If it wasn’t for verse 4 and the consequent threat of verse 5, they’d have a super-positive obedience report card, yes?)