Let’s just admit it. We rarely pay attention to the final greetings that Paul offers at the end of his letters. Such personal statements are, well, too personal—they just don’t seem meant for us. However, our unfortunate neglect of these passages can leave a variety of treasures undiscovered. One such passage may even bring unexpected illumination about the origins of the New Testament canon.
In 2 Tim 4:13 Paul says to Timothy, “When you come, bring the cloak that I left with Carpus at Troas, also the books, and above all the parchments.” Paul makes a curious distinction here between “the books” (ta biblia) and “the parchments” (tas membranas), suggesting they are two different kinds of writings. Scholars widely regard the first of these as a reference to books of the Old Testament, most likely on scrolls. We do not know how many of these Old Testament books Paul had in mind, but it must have been limited to a reasonable number that Timothy could have borne during his travels.
But, what is Paul referring to when he mentions “the parchments”? The term membranas is significant because it is not a Greek word, but a loan word transliterated from the Latin membrana. The history of this term in the first century makes it clear that it is a reference to a parchment codex. The codex book format was different than that of a scroll. Whereas the latter had writing on only one side (and was rolled up to protect that writing), the codex had writing on both sides of the page and was bound at the spine—like our traditional leaf books today. What is interesting about early Christians is that they vastly preferred the codex book format over the scroll even though both the Jewish world and the Greco-Roman world preferred the scroll. Indeed, the reason for the widespread and early use of the codex amongst Christians is a great mystery that scholars have sought to solve for a very long time.
As for the content of the codices which Paul mentions in 2 Tim 4:13, a number of suggestions have been made over the years. Given that Paul distinguishes these codices from the Old Testament writings, many scholars have rightly argued that they likely contained some sort of Christian writings. This may have included a variety of things such as excerpts of Jesus’ teachings or early Christian testimonia (Old Testament proof texts supporting Messianic claims about Jesus). Given the fact that Paul appears to cite Luke’s gospel elsewhere (1 Tim 5:18), and has an established relationship with Luke (Col 4:14, 2 Tim 4:11), we must even consider the possibility that these codices contained the gospel of Luke (!).
However, one of the most compelling possibilities is that these notebooks contained (among other things) copies of Paul’s own letters. It was not at all unusual in the Greco-Roman world to keep copies of (and even publish) one’s own letters. Cicero exemplifies this practice as his personal secretary, Tiro, kept extensive copies of his letters. Cicero would occasionally receive a complaint from friends that one of their letters (from Cicero) was lost or damaged; on such occasions Cicero would quickly dispatch a replacement copy from his own collection. And where did Cicero make/keep copies of his letters? He tells us: “I am jotting down a copy of this letter into my notebook.” In other words, Cicero kept copies of his letters in a codex.
If these “parchments” in 2 Tim 4:13 contained copies of Paul’s letters in a codex, then this opens up fresh insights the development of the New Testament canon. Such a scenario might begin to answer the question of why early Christians preferred the codex over the scroll. Since Paul had already begun to use the codex to contain his letters it is not difficult to imagine that early Christians would have retained that format when it became desirable to circulate a defined Pauline letter collection more broadly to the churches. Moreover, this scenario provides a compelling explanation for why some letters of Paul were preserved for the church and some letters were ultimately lost (1 Cor 5:9). The answer appears to be that some letters were lost because Paul, for whatever reasons, did not make a personal copy of them before sending them out. Thus, they were not available when Paul’s completed letter collection was circulating more broadly to the churches.
Most importantly, 2 Tim 4:13 provides additional support to the idea that, at a very early time period, Christians conceived of their religious writings in two parts: the Old Testament writings (ta biblia) and their Christian writings (tas membranas). In Paul’s day the latter would have still been fairly undefined, including not only copies of his own letters, but possibly excerpts of Jesus tradition, Christian testimonia, and the like. However, even though the content was undefined, there are hints here of a “proto-canon” of sorts, where valuable Christians texts are gathered into one place, in the form of a codex, with some even written by apostles.
If so, then perhaps the beginnings of the New Testament canon can be traced back to Paul himself.
 If Acts is dated before Paul’s death (given the odd, truncated ending), then it is reasonable to put Luke in the early to mid 60’s. This would allow Paul to know about it before writing 1 Timothy.
 E. Randolph Richards, “The Codex and the Early Collection of Paul’s Letters,” BBR 8 (1998): 151-166; David Trobisch, Paul’s Letter Collection: Tracing the Origins (Minneapolis: Fortress, 1994); Gamble, Books and Readers, 100-101.
 Fam. 9.26.1.