The difficult textual variant in John 1.18 concerns whether Jesus is described as o μονογενης υιος (“the only begotten Son”) or [o] μονογενης θεος (“the only begotten God”). Some of the important early critical editions (e.g. von Soden, Bover, Tischendorf), preferred the variant o μονογενης υιος. On the other hand, in 1876 Hort authored a paper which discusses the reception of μονογενης θεος in scripture and tradition. Hort believed this variant should be accept on grounds of the documentary evidence, of transcriptional probabilities, and of the intrinsic fitness of the variant. Adjudicating the variant in a similar manner to Hort are the critical editions of Merk, Holmes, and Nestle-Aland.
In English translations of the New Testament, μονογενης θεος is the preferred reading of the ESV, NASB, (T)NIV, and the NET, whereas o μονογενης υιος is the reading of KJV, NKJV, RSV, and the HCSB.
A Brief Critical Apparatus
Manuscripts: P66, 01*, 03, 04*, 019, 028*, 423
Versions: Diatessaron, Syriac (Pe and Hamg )
Fathers: Arius (acc. to Epiphanius), Basil, Clement, Cyril, Didymus, Epiphanius, Gregory-Nyssa, Hilary, Irenaeus, Jerome, Origen, Pseudo-Ignatius, Ptolemy, Theodotus (acc. to Clement), the Valentinians (acc. to Irenaeus and Clement), and the author of the Apostolic Constitutions
o μονογενης θεος
Manuscripts: P75, 011, 05, 33
Versions: Coptic (Sa & Bo)
o μονογενης υιος
Manuscripts: 02, 043, 011, 017, 029, 033, 034, 037, 041, and Majority text
Versions: Vulgate, Syriac (Cu, Ha), Armenian, Ethiopian, Georgian, Slavonic
Fathers: Alexander, Ambrose, Ambrosiaster, Athanasius, Augustine, Basil, Caesarius, Irenaeus, Clement, Cyril, Chrysostom, Hippolytus, Origen, Eustathius, Eusebius, Serapion, Gregory-Nazaianzus, Theodoret, John Damascus, Tertullian, Hilary, Fulgentius, Jerome, Victorinus
A few things immediately leap out from this data. First, more than a few of the same names appear on lists of patristic witnesses to both variants. This, of course, takes into account the fact that a witness to o μονογενης υιος only counts if it is clear that it is John 1.18 being referenced (considering the phrase appears elsewhere in the Johannine corpus – John 1.14, 18; 3.16, 18; 1 John 4.9). Second, the earliest appearances of μονογενης θεος is found almost exclusively in Alexandrian manuscripts. And third, the anarthrous form of μονογενης θεος is to be preferred over the articular (perhaps the articular was a scribal emendation to the confusing anarthrous).
The Case for μονογενης θεος
In his discussion of this textual variant, Bart Ehrman says that μονογενης θεος is “found almost exclusively in the Alexandrian tradition.” That is probably not entirely accurate considering that Sinaiticus is regarded as a Western text in John 1.1-8.38. But nevertheless, μονογενης θεος is primarily an Alexandrian reading, found in manuscripts of the Alexandrian text-type (including the Alexandrian uncials 01, 03, 04, and the papyri P66 and P75), as well as in early Alexandrian Fathers (e.g. Clement of Alexandria witnesses to both variants in his Stromata; Origen likewise has μονογενης θεος in his Against Celsus, but also has o μονογενης υιος in his Commentary on John).
Interesting, though I guess not surprising, is the fact that μονογενης θεος is the exclusive reading in the Coptic tradition. It is also found in some Syriac manuscripts of both the Harklean and Peshitta traditions, which is significant considering the Harklean is associated with the Western text-type and the Peshitta with the Byzantine.
In his text critical commentary on the New Testament, Philip Comfort prefers μονογενης θεος, saying that the papyri P66 and P75 “tip the balance”. Yet, as Ehrman has pointed out, the discovery of these papyri, while not being trivial discoveries, do not actually change the picture regarding this textual variant. An interesting argument Comfort mentions in favor of μονογενης θεος is that Jesus is described as θεος in John 1.1, then μονογενης in 1.14, and then μονογενης θεος in 1.18 (which is typically seen as the end of the prologue of John’s Gospel).
Metzger gives μονογενης θεος a “B” rating in his textual commentary, which signifies that the reading is “almost certain.” He puts o μονογενης υιος down to a scribal assimilation to John 3:16, 18. One member of the USB committee, Allen Wikgren, dissented from the majority’s opinion and opted for o μονογενης υιος. Wikgren notes in Metzger’s commentary that a “D” rating would have been preferable for μονογενης θεος.
The Case for o μονογενης υιος
As Ehrman notes, virtually every text type (Western, Caesarean, Byzantine, Alexandrian) all witness to o μονογενης υιος. Additionally, the reading is found in the Greek, Latin, and Syriac traditions, as well as the Armenian, Ethiopian, Georgian and Slavonic traditions. A couple points need to be noted at this juncture. First, despite the fact that secondary witnesses to the Alexandrian text type attest to υιος (e.g. 029, 037, 044), one has to keep in mind that this reading could hardly be claimed to go all the way back to the Alexandrian archetype. And second, it needs to be noted that the Caesarean grouping (e.g. f1, f13) has largely fallen out of favor as being a legitimate text type.
Along with the strong external evidence for this variant is the strong internal evidence. As mentioned earlier, the phrase o μονογενης υιος is used elsewhere in the gospel (John 1.18, 3.16, 18; also, 1 John 4.9). The phrase μονογενης θεος is not found elsewhere in John, or the New Testament. In fact, the term μονογενης is used almost exclusively in the New Testament literature with υιος (the lone exception being Luke 8.42). Theologically speaking, o μονογενης υιος is much more kosher than μονογενης θεος, which could rightly be regarded as an aberration of Johannine thought on God; where else is the concept of a God dwelling alongside another God so readily apparent? Regarding literary style, it could be said that o μονογενης υιος is more elegant considering it meshes nicely with the use of “Father” in the verse (“the only-begotten Son, who is at the Father’s side”). These arguments, of course, also add weight to the superiority of μονογενης θεος due to the fact it makes the principle of lectio difficilior potior apply to that variant.
An Excursus – The Meaning and Rendering of μονογενης θεος
In the New Testament literature μονογενης only occurs only in Luke, John, and Hebrews. Older translations, such as the KJV, render it as “only begotten.” This is due in part to the Latin rendering of μονογενης as uni-genitus, as well as the fact that μονογενης was thought to have be derived from γενναο (“begotten”) instead of γινομαι (“to become”) [which, I think, is now recognized as where μονογενης is actually derived from].
Μονογενης can be used adjectivally, stressing quality, rather than stressing the concept of genealogical descent or derivation. So instead of “only begotten”, the meaning of the word can warrant a translation of “unique” to convey the properly nuanced meaning. For example, in Hebrews 11.17 Isaac is described as the μονογενης son of Abraham. Was this because Isaac was the only son of Abraham? No (remember Ishmael)! But Isaac is the μονογενης son in the sense of “unique” son because he is the son of the promise.
So if we were to assume that μονογενης θεος was the original reading of the text, how then shall we render it? A couple of examples are:
- the unique God
- the Unique One, God
The first example uses μονογενης in an adjectival sense, but if the author intended this meaning, he could have more clearly conveyed it by saying o μονογενης θεος or o μονογενης o θεος. The second example is the most interesting considering that it sees μονογενης as a substantive. This substantival sense is the most promising rendering of this variant as it sees 1.18 as containing a series of appositions, which could be translated as “the Unique One, [who is] God, [who is] in the bosom of the Father.” Compare to the translation of the NET: “The only one, himself God, who is in the presence of the Father…”
Ehrman, however, claims this proposal is “entirely implausible.” His claim is not that μονογενης can not be used as a substantive (which it is in John 1.14), but that it is not a substantive when followed by a noun that agrees with it in gender, number, and case. He declares, “No Greek reader would construe such a construction as a string of substantives, and no Greek writer, would created such as inconcinnity.” I wonder, though, how thoroughly Ehrman culled through Greek literature of the period (or even the NT) to see whether an adjective attached to a noun of the same case, gender and number has to necessarily function adjectivally and not substantivally.
A plausible case can be made for the originality of μονογενης θεος. Assuming this was the original reading, it could have caused confusion leading to the definite article being added to give o μονογενης θεος. This, then, could perhaps lead a scribe to think the verse originally said o μονογενης υιος. Another explanation is that the a scribe may have blundered by reading a nomina sacra θεος (ΘΣ) as a nomina sacra υιος (ΥΣ). Yet, even though θεος was one of the four key words abbreviated as nomina sacra, υιος was not a primary nomina sacra which could raise the question as to why an early scribe would write a ΥΣ instead of a ΘΣ. Regardless, the idea that this textual variation resulted from a nomina sacra mishap makes more sense when arguing for θεος being changed into υιος considering that the former was abbreviated as a nomina sacra from quite early on, whereas the latter wasn’t.
On the other hand, I would have to probably agree with Bart Ehrman that o μονογενης υιος is the superior reading. This leads to the question as to why μονογενης θεος arose as a variant reading. Some (e.g. Ehrman) would posit that it is a modification to the text that occurred during the adoptionist Christological controversy. That is to say, the variant was created to support a higher Christology by affirming the Son as deity. While I can not claim to know whether this is indeed the reason that the θεος variant was introduced into the textual tradition, I think that the external and internal evidence do point in favor of υιος. For while neither reading is contained to a particular text type, though μονογενης θεος comes fairly close to be a uniquely Alexandrian reading, it is manifestly clear that o μονογενης υιος enjoys an earlier and clearer geographical distribution. This reading is also supported better internally, both stylistically and theologically.
 Fenton John Anthony Hort, Two Dissertations, (Cambridge: MacMillan and Co., 1867), 1-72
 Two important studies on this textual variant are: Paul McReynolds, “John 1:18 in Textual Variation and Translation,” in New Testament Textual Criticism: Its Significance for Exegesis: Essays in Honor of Bruce M. Metzger, (Oxford: Clarendon Press, 1981); D. A. Fennema, ‘John 1:18’, New Testament Studies 31 (1985): 125-26.
 Bart Ehrman, The Orthodox Corruption of Scripture: The Effect of Early Christological Controversies on the Text of the New Testament, (New York, NY: Oxford University Press, 2003): 72-82, 265-66
 ibid., 79
 Philip Comfort, New Testament Text and Translation Commentary (Tyndale House, 2008), 255-56
 Bruce Metzger, Textual Commentary on the Greek New Testament, 2nd edition (Deutsche Bibelgesellschaft, 1994), 169-70
 Ehrman, op. cit., 79
 For support of the traditional “only begotten”, see John Dahms, ‘The Johannine Use of Monogenes Reconsidered,’ New Testament Studies 29 (1983) 222-32. For support of removing the idea of “begotten” from the phrase, see Dale Moody, “God’s Only Son: The Translation of John 3.16 in the Revised Standard Version’, Journal of Biblical Literature 72 (1953): 213-19.
  Ehrman, op. cit., 81