James Charlesworth writes (The Pseudepigrapha and Modern Research, pp. 224-226):
The nature of the present monograph precludes a discussion of the literary history of this work. For the present a brief outline of an hypothesis must suffice. The work consists of twenty-two chapters (cf. James, pp. 96-108; and Craigie, pp. 220-24) of which possibly the first and certainly the last were appended later because only they are written in teh third person, while chapters two to twenty-one are in the first person. It is not surprising to perceive accretions appearing at the end and beginning of a text; in fact the Armenian recension (cf. Zanolli, p. 153) even adds to chapter one the idea that Yovsimios lived on a mountain on Schizia, an island at the end and beginning of a text; in fact the Armenian recension (cf. Zanolli, p. 153) even adds to chapter one the idea that Yovsimios lived on a mountain on Schizia, an island in the Ionian Sea. Also belonging to this latest level is the last sentence of chapter twenty-one, which defines the work as Zosimus' testament (he diatheke aute). Chapters nineteen to twenty-one were appended earlier to the work, because Zosimus' name does not appear in them and the narrative is out of character for this powerless monk. These chapters appear to be a remnant of an early account of Jesus' conquest of the Devil (ho Diabolos) during the forty days of temptation, because forty days are mentioned more than once, bcause only the traditions attributed to Jesus aptly fit the Devil's lamentation ("Woe is me that by one man I have lost the world [These chapters are under the influence of Rom 5.], for he has conquered me by his prayer."), and because the judgment of the Devil suggests Jesus' authority ("Then I dismissed him, dispatching [him] and the demons with him into the eternal fire."). Attributing chapters nineteen to twenty-one to another literary stratum explains why Zosimus' tablets (tas plakas) are called "the book" (ten biblon) in chapter nineteen, and why Satan (chaps. 6, 18) is called the Devil only in chapters nineteen, twenty, and twenty-one. Chapters two and fifteen-b through eighteen are earlier prefixed and suffixed additiosn to the core because the name Zosimus appears in them nine times. These appear to be by the same scribe since the river is called Eumeles only in chapters two and fifteen-b. Either the scribe of this stratum was a Christian, or his work was redacted by a Christian. The remainder of the document, chapters three through fifteen-a, is the core inw hich the name of Zosimus does not occur and which appears to be Jewish with frequent indications that the original was composed in a Semitic language (viz. "lamented with great lamentation," chps. 6 and 7; "rejoiced with great joy," chp. 7). In the core, which is an apocalypse, the seer is called "a man of God" (chp. 4), "the man of vanity" (chp. 5), or simply "man" (chp. 6). The allocation of chatpers one and six to two different literary strata explains the contradiction between Zosimus' unworthiness (ouk ei axios) and the man's worthiness (kai katexiosen me). The parallels in the core with The Lost Tribes indicate that it may have been composed around A.D. 100. Behind these chapters, however, there seems to be a very ancient core, chapters seven through nine, which concerns the history and present abode of the descendants of Rechab, the son of Jonadab, who were not scattered over the earth but are in a place encircled by an abyss and a cloud (chp. 9). Chapters six and ten, with their impressive similarities, appear to reveal that the evolution moved centrifugally from chapters seven through nine. Since the ancient core, the Rechabite text, claims that God turned away his anger from Jerusalem (chps. 7 and 8) and that God's mercy came to Jerusalem (chp. 7), it would be unwise to ignore the possibility that this oldest section is a Jewish work that predates the fall of Jerusalem in A.D. 70.
If the above analysis is generally correct, then it is possible that the ancient core and the core, because of their Semitic flavor and concern for Jerusalem, were written somewhere in Judaea. This suggestion is corroborated, but of course not proved, by the superscription in a British Library Syriac Manuscript of the work (B. M. Add. MS 12174, f. 209v); "But it was translated from Hebrew into Greek, and then from Greek into Syriac by the Holy Mar Jacob of Edessa." It is difficult, therefore, to agree with K. Kunze (no. 1475), who claims that this work was composed in Greek in the sixth century. We can be relatively certain that the original is Jewish and has been redacted by Christians (so also G. Graf, Geschichte, p. 214; J.-C. Picard, no. 1476; Nau, RevSem 6  265; L. Ginzberg, Legende, vol. 6, p. 409).
The above hypothesis may be outlined as follows:
V. Testament of Zosimus (chps. 1 and 22)
IV. Jesus' Conquest of the Devil (chps. 19-21)
III. Christian Additions (chps.2,15b-18)
II. Apocalypse (chps. 3-6, 10-15a)
I. Rechabite Text (chps. 7-9)
An unexpected confirmation of some of this hypothesis comes from the Syriac tradition. This version ends with chapter 16 and is entitled "The History of the Blessed Sons of the Rechabites" (B. M. Add. MS 12174, f. 209v).
James Charlesworth writes: "The date of the History of the Rechabites is the crucial issue, and it is related to the Jewish or Christian character of the various sections. In its present form the work may date from the sixth century A.D., as M. R. James contended. Comparison of the Syriac manuscripts reveals that the document, like many pseudepigrapha (viz. 4Ezra), has received interpolations by Christians; the same observation results from a mere cursory examination and comparison of the Greek manuscripts, and by the recognition that the Greek is expanded by chapters 19 through 23, which are certainly Christian. The Ethiopic, moreover, has been extensively expanded by scribes who were obviously Christian. Some of the present document is Christian, but the Christian interpolationssometimes found in only one manuscriptraise the possibility that 12:9a-13:5c and 16:1b-8 are not original but a Christian insertion into an earlier document. This hypothetical earlier writing could be a Christian revision of inherited Jewish traditions, or it could be a Christian expansion of an original (partly preserved) Jewish document. James, A. Zanolli, Nau, G. Graf, L. Ginzberg, J.-C. Picard, and B. McNeil have perceived evidence of a Jewish original behind the present Christian document. Nau even used such terms as 'the Christian translator,' 'the primitive text,' 'the Hebrew text,' and 'the Hebrew author.' Working with only the Greek document generates the impression that the beginning and end are Christian and that the central chapters, 3-15, are originally Jewish. Focusing upon the Syriac document leaves the impression that only 12:9a-13:5c and 16:1b-8 are clearly Christian and appear to be interpolated, because they interrupt the flow of thought and contain intrusive ideas. The mention of the name 'Zosimus' in the latter section (16:8) suggests that perhaps all passages connected with this name may be from a later stratum, hence chapters 7:12-16:1a, which do not identify the traveler as 'Zosimus,' would be earlier and possibly Jewish. It is only in these chapters, and specifically in 8-10, that mention is made of the Rechabites and their history in Jerusalem during the days of Jeremiah. At this stage in our work it is best to suggest only that sections of this document are Jewish or heavily influenced by Jewish traditions, and that they may antedate the second century A.D." (The Old Testament Pseudepigrapha, vol. 2, pp. 444-445)
Please buy the CD to support the site, view it without ads, and get bonus stuff!