John 5:3b-4: Original or Not?

            Metzger’s observation that 5:3b contains two “non-Johannine” words is lightweight, considering that John had few other occasions to use either ἐκδέχεσθαι or κίνησις.

            (I commend to readers both the article written by Zane Hodges in 1979 in Bibliotheca Sacra 136, pp. 25-39, and the article by Gordon Fee which appeared in Evangelical Quarterly 54 (pp. 207-218.)              

            Before reaching a conclusion about John 5:3b, let’s investigate 5:4.  Dr. Bill Mounce addressed this variant briefly, but his treatment is extremely oversimplified.  More is required.  First, we must get an idea of how much textual variation there is within this verse.  In A K L Y Δ Π, κυρίου (ΚΥ) appears after αγγελος γαρ (or, in L, αγγελος δε).  And instead of κατέβεινεν, A K Π Ψ 579 have ἐλούετο.  And A (supported by some Bohairic manuscripts) has ουν between δήποτ’ and κατείχετο.  Instead of δήποτε, K and Π have δ’ αν.  In C^c H M U Y Δ Λ Π 078 and at least 17 lectionaries, instead of ἐτάρασσεν, the text reads ἐταράσσετο.  The Ethiopic version also supports ἐταράσσετο.   Swanson erroneously lists Δ as if it reads ἐταράσσετο and ἐτάρασσεν; a check of the manuscript show that it supports ἐταράσσε το (the το being the το before ὕδωρ).  

            Plus, in S Λ Π 047, and 72 minuscules, the passage is marked with asterisks.  The Harklean Syriac also features the verse marked with asterisks.

             The external evidence mostly aligns with the external evidence for 5:3b – but not quite. D W^supp 33, 2718, and the Armenian and Georgian versions, which include 4:3b, do not imclude 5:4.  5:4 is supported by Tatian’s Diatessaron (as demonstrated by a comment by Ephrem in his commentary ), by Ambrose, by Tertullian, by Chrysostom (who was listed in UBS1 as a witness for both inclusion and non-inclusion), and Cyril.  

            Tertullian, in De Baptismo 5, near the end of the chapter, wrote, “If it seems an unheard-of thing that an angel should interfere with water, there was a precedent for that which was to be. The pool of Bethsaida ‘was stirred’ by the intervention of ‘an angel.’  Those who complained of their health used to watch for him. For anyone who had first descended there ceased to complain after a bath. This picture of bodily cure was prophetic of spiritual cure, according to the practice by which things carnal always precede, being a picture of things spiritual. As, therefore, the grace of God spread among men, greater power was added to the waters and the angel.”

            Tertullian goes on to say, “Those who healed bodily defects now heal the spirit.  Those who worked temporal salvation now restore for us everlasting salvation.  Those who freed one once a year, [this indicates how Tertullian understood κατά καιρόν] now daily save communities, death being destroyed by the washing away of sins.”  Tertullian clearly had no problem reading this verse and applying it to the life of the church.

            Chrysostom commented on 5:3b-4 in detail in his commentary on John, perceiving in the paralytic’s healing a thematic template of baptism and salvation. 

            Tertullian, in Latin, and Chrysostom, in Greek, demonstrate the antiquity of the passage in the text, as early as two papyri from c. 200 and c. 400 would.  Chrysostom also shows that John 5:4 was read in the text of the church in Byzantium during his bishopric.  Amphilochius of Iconium (340-403; bishop after 374) – cousin of Gregory of Nazianzus – does not include 5:4 in the text he used.  Both the non-inclusion and inclusion of 5:4 are very early readings.

            What phenomenon, occurring sometime between 90 (when the Gospel of John was written – unless John Robinson’s redating to pre-70 – in light of (among other things) 5:2 – is adopted) and 200, could elicit one transmission-stream to include John 5:4 (in the case of Tertullian’s text of John), and another transmission-stream to not include John 5:4 (in the case of P75, À, and B)? 

            I am willing to posit that an anomaly in the autograph of the Gospel of John itself elicited different treatments of John 5:3b-4.  Picture John reading chapter 5 to his listeners from the autograph for the very first time – without 5:3b-4.  Inevitably, someone would ask, “John, why were these sick, blind, lame, and paralyzed people waiting near the pool instead of swimming in its water?”  And I can imagine that John added an explanatory note in the margin, “waiting for the moving of the waters.”

            And then someone asked, “What agitated the pool’s water?”.  And John, realizing that his listeners in Ephesus were oblivious to the background of the pool at Bethesda, added another note – and thus verse 4 came into existence as a second marginal note.  When John died, the autograph was entrusted to the Christian community at Ephesus – and they treated the annotations in three different ways in the next generation:

            In the ancestor of Byzantine manuscripts, the notes were either blended into the main text (as John 21 has been), or else copies just the way they appeared in the autograph, in the margin with symbols to connect them to John 5:3-5.  In the ancestor of Alexandrian manuscripts, receiving the text of the autograph slightly later (being in Egypt, not Ephesus), the notes were assumed to have originated with someone other than John, and were therefore not considered worthy to be included in either the main text or in the margin. 

            Another consideration might have been in play in the mind of the early Alexandrian scribe who decided not to include verse 4:  a desire to protect John from the charge of promoting superstition.  A scribe who thought he knew that water in the pool of Bethesda was agitated by entirely natural forces could easily persuade himself that the marginal note in his exemplar, stating that an angel of the Lord bathed in the pool of Bethesda, could not have been written by an inspired author; in addition, he did not wish to appear to commend Asklepieions.

            The testimony of P and its relatives which have John 5:4 with asterisks commends family P as an excellent representative of the autograph of the text of the Gospels.  The form of verse 4 that appears in Codex P is the form which should be adopted, instead of the readings found in the majority of manuscripts.

            An addition question is sure to be asked:  what should English Bible editors do with John 5:3-4?  I have no objection to the inclusion of 5:3-4 in the main text, or in the margin, with a note stating that the passage appears in the margin, or not at all, in a few early manuscripts.  But to omit it entirely would guarantee that English readers would perpetually ask, as John’s first listeners did, “Why weren’t they all swimming?” or, “Who or what stirred up the waters?”

            Another question may be on the minds of some readers:  Would an inspired author expand on his own narrative in this way, adding marginalia?  I see no reason why not.  Many a Spirit-led preacher reading from a manuscript he wrote has spontaneously clarified himself mid-sermon.  Even Saint Paul, in First Corinthians 1:16, clarified that he had baptized the household of Stephanas (who, according to tradition, was the jailor in Acts 16), right after saying, “I baptized none of you except Crispus and Gaius.”  I Cor. 1:16 may have originally been a note in the margin added by Paul as he proof-read the letter; no one at Corinth, however, would have doubted its veracity.

 

Meet GA 473

             GA 473 is the same manuscript that was known to F. H. A. Scrivener in 1853 as Lambeth Palace MS 1178 when he collated it (as Gospels-MS “d”) for his book about A Full and Exact Collation of About Twenty Greek Manuscripts of the Holy Gospels.   

            “It is one of the most splendid manuscripts extant,” Scrivener wrote about GA 473.  He also mentioned that GA 473’s text “contains many remarkable variations, such as Scholz would refer to his Alexandrian recension.”  Robert Waltz has briefly updated Scrivener, noting that GA 473 is part of von Soden’s “family I^KC,” and thus an imperfect representative of Family Π, which constitutes, for the most part, the earliest stratum of the Byzantine Text of the Gospels. 

          Scrivener stated that Luke 22:43-44 is obelized, but the passage is accompanied by asterisks, not obeli, in the margin.  And the manuscript features more than what Scrivener shared with his readers:  in the margin of GA 473, alongside Luke 22:43-44, there is also a note referring to part a Maundy Thursday lection.  (This lection begins in Matthew 26; jumps to John 13:1-17, jumps back to Matthew 26, and then jumps to Luke 22:43-44, and then jumps back to Matthew 26:40 (i.e., to Section 296 of Matthew).  The hundreds-unit in the margin of GA 473 is an error; the expected hundreds-unit is 200 (ς) rather than 100 (ρ).  Other than this misprint, or rather, miswriting, the marginalia in GA 473 shows how the text was used at Eastertime in the lectionary.  (The asterisks need not convey scribal doubt.)  (In the core members of family 13, Luke 22:43-44 is embedded within the text of Matthew, which is one of the strongest proofs of lectionary influence upon family 13’s text.)              
            Scrivener listed 32 notable unusual readings attested by GA 473.  Because he collated the entire text, there is no need to go through them all here, but a few may be pointed out as examples:   

            At Matthew 23:27, GA 473 reads τοῖς ἀνθρώποις after φαίνονται, a reading shared by F and 33 (sorta). 
            At Mark 16:19, GA 473 reads ἐν δεξια instead of ἐκ δεξιῶν, a reading shared by C and Δ.

            At Luke 14:32, GA 473 reads εις before πρεσβείαν, yielding a minor tweak of the meaning.

            At Luke 17:17, GA 473 reads ευκοποτερον (relieved) instead of the usual εκαθαρίσθησαν (cleansed).  Perhaps an overly scrupulous scribe did not think that those who did not express thanks for the miraculous healing Jesus had given merited being called clean.

            At John 2:14, GA 473 is missing the words καὶ περιστερὰς (and doves).  This was very extremely probably due to simple parablepsis from the καὶ before περιστερὰς to the καὶ that comes after περιστερὰς.

            At John 13:15, GA 473 reads ποιω instead of ἐποίησα.

            At John 19:21, GA 473 is missing the first occurrence of τῶν Ἰουδαίων.  This is probably not accidental, but a case of stylistic adjustment to remove perceived superfluity.

        Digital page-views of GA 473 at the CSNTM website have been fully indexed.  Readers are invited to look over this manuscript for themselves to view its other interesting features (such as the obeli in the margin alongside John 5:4, as well as the readings within John 5:4 itself).

 

 

A Moment Please, Dr. Holmes!

             In a recent article at the Bible Odyssey website, What Are English Translations of the Bible Based On?, Dr. Michael W. Holmes made some claims about the external support for some verses that appear in the text of the New Testament in the King James Version but not in the New International Version, the New Revised Standard Version, or the English Standard Version.  He claimed that Matthew 17:21, Matthew 18:11, Matthew 23:14, Mark 7:16, Mark 9:44, Mark 9:46, Mark 11:26, Mark 15:28, Luke 17:36, Luke 23:17, John 5:3b-4, Acts 8:37, Acts 15:34, Acts 24:6b-8a, Romans 16:24, and 1 John 5:7b-8a were all added to the New Testament; that is, that they are not original. 

From beginning to end, Holmes’ article seems designed to convey that these verses are not in the NIV, NRSV, and ESV because they are not supported by early evidence.  That is undoubtedly the impression that many of Dr. Holmes’ readers derived from the article.  However, the real state of the evidence is quite different.  Readers of Bible Odyssey may find their impressions modified by the following observations:

● Although it is claimed that the oldest manuscript used by Erasmus was “from the 10th century,” Erasmus also used the quotations made by patristic writers who wrote long before then, including Irenaeus (who wrote in the second century).  Also, in the course of making his revised editions, Erasmus learned of readings in other manuscripts, including some readings from Codex Vaticanus (made around 325). 

● Dr. Holmes claimed that to scholars in the late 1700s and 1800s, it became clear, as they studied ancient manuscripts, that “the text of the New Testament had “grown” slightly as it was copied by hand century after century.”  However, while this was indeed the impression that those scholars had at the time – and they were so sure of it that a guideline was developed, Lectio brevior potior (the shorter reading is to be preferred) – a series of more recent studies (including a particularly extensive piece of research by James Royse) has shown that copyists’ tendency to omit was much more dominant than their tendency to add.  Dr. Holmes no doubt already knows that one should not casually adopt short readings from ancient manuscripts; he retained Matthew 12:47 in the text of the SBL-GNT although it is not supported by the early manuscripts Vaticanus and Sinaiticus.  (The ESV, meanwhile, does not have Matthew 12:47 in its text.)

 

Matthew 17:21 in Codex W.

● Matthew 17:21 is not only in over 99% of the Greek manuscripts of Matthew; it was in the manuscripts used by the early church writer Origen (early 200s-254).  One can consult Origen’s Commentary on Matthew, Book 13, chapter 7, to see this.  It is also in the Vulgate, which was translated by Jerome in 383.  (Jerome stated in his Preface to the Vulgate Gospels that he had consulted ancient Greek manuscripts in the preparation of the Gospels’ text.)  Codex W, found in Egypt, also includes the verse.  The Latin manuscripts used by Ambrose of Milan in the 300s also included this verse, and so do several Old Latin manuscripts.  Thus the support for this verse does not only come from the vast majority of Greek manuscripts; it comes from a patristic quotation earlier than the earliest manuscript of this part of the Gospel of Matthew, and it comes from witnesses in at least four different parts of the Roman Empire. 

● Matthew 18:11 was in the Greek manuscripts used by John Chrysostom (late 300s, in Constantinople) and in the Latin manuscripts used by Augustine (early 400s, in North Africa).  The verse is not only in the vast majority of manuscripts, but also in Codex Bezae (“one of our oldest witnesses,” according to Bart Ehrman) and Codex W and the Vulgate and the Peshitta – quite a diverse quartet from the late 300s and 400s (depending on what production-date is given to Codex D). 

● Matthew 23:14 – the verse in which Jesus mentions widows’ houses and long prayers as He denounces the scribes and Pharisees – is in most manuscripts, and in most manuscripts it appears before the verse that is, in the KJV, verse 13.  The inclusion of this verse (whether before or after verse 13) is not just supported by a strong majority of Greek manuscripts; it is also supported by the testimony of Hilary of Poitiers (310-367, in France), John Chrysostom (late 300s-407, in Constantinople), and the Peshitta version (in Syria).

            Another consideration regarding this verse is that verse 13 and verse 14 begin with precisely the same word (“Woe,” Greek Οὐαὶ), which would make it easy for a verse to disappear if a copyist’s line of sight accidentally drifted from one occurrence of the word Οὐαὶ to another occurrence of the word Οὐαὶ.   

● Mark 7:16 is supported by the vast majority of Greek manuscripts and also by Codex Alexandrinus, by Codex Bezae, by the Gothic version (made c. 350), by the Vulgate, by some Old Latin copies, and by Augustine.  By no possible stretch of the imagination is its support limited to late evidence. 

● Mark 9:44 and 9:46 are supported by the vast majority of manuscripts, and also by Codex A, Codex D, the Gothic version, the Vulgate, and some Old Latin copies. 

● Mark 11:26 is supported by the vast majority of Greek manuscripts, and by Codex Alexandrinus, Codex Bezae, by Old Latin copies including Codex Vercellensis (probably made in the 370s), by the Gothic version, the Vulgate, and by Augustine. 

Another consideration is that this verse ends with the same phrase (“your trespasses,” Greek τὰ παραπτώματα ὑμῶν) as the verse that comes before it, which made this verse vulnerable to accidental omission in the event that a copyist’s line of sight drifted from one occurrence of the phrase to the occurrence of the same phrase further along in the text (thus skipping all the words in between). 

● Mark 15:28 is supported by a large majority of Greek manuscripts, and by the Vulgate, the Gothic version, the Peshitta version, and very probably by Eusebius of Emesa (in the mid-300s).  Eusebius of Caesarea, in the early 300s, seems to have known it and included it in his Canon-tables as a cross-reference with Luke 22:37.

● Luke 17:36 is not supported by a majority of Greek manuscripts, but it has support from Codex D, a dozen Old Latin copies, the Vulgate, the Peshitta, both forms of the Old Syriac (that is, both the Sinaitic Syriac and the Curetonian Syriac), and from Ambrose of Milan, and from Augustine. 

Another consideration is that this verse ends with the same word with which the previous verse ends (“shall be left,” Greek ἀφεθήσεται), which made this verse vulnerable to accidental omission when a copyist’s line of sight drifted from the ἀφεθήσεται at the end of verse 35 to the ἀφεθήσεται at the end of verse 36, skipping the words in between.      

● Luke 23:17 is supported by the vast majority of manuscripts, and by Codex Sinaiticus (note that at the Codex Sinaiticus website, the English translation is completely bogus at this point), Codex W, Old Latin copies, and the Vulgate, and in Codex Bezae and in both Old Syriac manuscripts, the verse is found after verse 18 rather than before it.  
            Another consideration is that this verse begins with the word Ἀνάγκην and the following verse begins with the word Ἀνέκραξαν; a copyist’s line of sight could feasible drift from one Ἀν- to the next Ἀν, omitting the words in between.   

● John 5:3b-4 is supported not only by the vast majority of manuscripts but also by the Vulgate, by Ambrose of Milan, by Chrysostom,  and by Tatian’s Diatessaron (as cited in the commentary on the Diatessaron by Ephraem the Syrian in the mid-300s) and by Tertullian (in De Baptismo, that is, On Baptism, chapter 5).  Tatian lived in the mid-100s, and Tertullian wrote in the early 200s.  Chrysostom also used this passage.    

● Acts 8:37 is only supported by about 15% of the Greek manuscripts of Acts, but it was used by Irenaeus (in Book 3, 12:8 of Against Heresies, written c. 180, long before the earliest existing manuscript of this part of Acts was made), by Cyprian (in the 200s, in Book Three of his Testimonies, Treatise 12, chapter 43), by Pontus the Deacon (in the mid-200s), by Pacian of Barcelona (late 300s), and by Augustine (early 400s); it is also in the Coptic (Egyptian) manuscript known as the Glazier Codex (made in the 400s or 500s).

            Now I set aside the remaining four passages mentioned by Dr. Holmes, in the interest of brevity (and to avoid getting distracted by the “Comma Johanneum,” the interpolation in the Textus Receptus in First John 5 that originated as an interpretive note in a branch of the Old Latin version).  Even the evidence-descriptions I have given are far from complete.  My goal today was simply to demonstrate that the historical details about most of the passages in Dr. Holmes’ list of “added verses” do not sustain his presentation, or any presentation which conveys that in these textual contests, the shorter reading is supported exclusively by early evidence, and the longer reading is supported exclusively by late evidence.
            Contrary to the picture that Dr. Holmes has painted, the real-life scenario here is not one in which all the ancient evidence pertinent to these passages points one way, and all the young evidence points a different way.  It is not a simple matter of older-versus-later, as anyone can see by observing that evidence for both readings can be found in testimony from the 300s – and in some cases, evidence for the longer reading comes from sources earlier than the earliest evidence for the shorter reading. 

In closing:  this is not intended to represent a full defense of the genuineness of all or any of these passages.  My sole point is that in most cases, there is ancient evidence on both sides.  If you read anything that gives any other impression – whether a Bible Odyssey article, or a commentary, or vague Bible footnotes – I recommend that your suspicions should be alert to the possibility that instead of reading a disinterested and balanced description of the evidence, you are reading a well-disguised campaign speech for the Alexandrian Text – a speech that cannot be persuasive unless you remain uninformed about the actual state of the evidence.

[Readers are encouraged to explore the embedded links in this post for additional resources.]