In a comment on my latest post, Tim Joseph asked a fair question about where I stand on the date of Bezae’s text. Since this question gets right to the heart of my research, I thought it a worthwhile topic for a post.
In a number of posts over the past few weeks, I have suggested that various features of Bezae’s text seem consistent with a date in the late fourth century, including:
- its parallels with the Old Latin version
- its parallels with early writers
- its bilingual tradition
- its so-called tendencies
- its resonance with late-antique antiquarianism
- the work of its correctors
In short, it seems little prevents Bezae’s distinctive final text form from having come together as late as the end of the fourth century. Of course, like most multi-layered texts, Bezae reflects influences from different periods. In terms of Bezae’s layers, my provisional dating in John 4:1-42 is:
- with Alexandrian (probably initial)
- with BYZ (probably 2nd)
- with Origen, family 1 and other Greek (3rd)
- with African Old Latin (3rd)
- with European Old Latin (4th)
- with Irenaeus and other authorities (perhaps 4th)
- singular (4th)
It seems that some of the above layers were likely combined before entering Bezae’s tradition:
- a Greek base (layers 1 through 3)
- a Latin base (layers 4 and 5, possibly 6)
Moreover, certain features of Bezae’s tradition may have entered through multiple layers:
- Bezae’s harmonizations may have entered through layers 2, 3, 4, 5, or 7
- Its citations of early writers, such as Irenaeus, may have entered through layers 4, 5, or 6
- Elements of an early “free” text may have entered at least through layers 3 or 4 
It is clear that many of the mainstream readings Bezae shares with the Byzantine tradition originated in the second century, because we see them already in P66. But while Bezae’s most distinctive readings are often assumed to derive from the second century, I think in many cases this assumption is at least questionable.
In my view, it has not been demonstrated that Bezae’s distinctive Greek text could not have been substantially influenced by Old Latin readings, not from Bezae’s own Latin column, but from other Latin versions, and not necessarily confined to so-called “Latinisms,” but to an entire layer of Old Latin readings. Such a retroversion of Old Latin readings into the Greek text, assumed by the above dates, must have post-dated the mid-fourth-century development of the European Old Latin tradition, which puts us not far from ca 400. 
Of course, the dates and model are not definitive.
References and Notes
 B. Aland, “Entstehung, Charakter und Herkunft des sog. westlichen Textes” ETL 62 (1986) 5–65 at 20-22 and 49, what Aland refers to as stage-1 changes.
 The Latin and Greek columns of Bezae’s bilingual tradition would presumably have been corrected afterwards in both directions to agree as closely as possible. I discuss this controversial matter further in posts on Latinization and the relation of Bezae’s columns.