Botha, P.J.J. "Chapter Six: Authorship in Historical Perspective." Orality and Literacy in Early Christianity. Eugene, OR: Cascade, 2013, 126-144.
Botha observes that not only the concept of composition, but also the definition of authorship is easily treated anachronistically by modern scholarship (Botha 2013, 126). When considering authorship, we rightly consider what an author was thought to be in the time and place of the composition of the work in question. Following Foucault, Botha takes the author prior to the Enlightenment (and, for that matter, after it) not to be the sole magisterial creator of a work. There are always other forces at work, a cultural and sociological context, for instance, which take a hand in governing creative work. Botha describes this in some detail, then briefly discusses how the author's name recognition became an important data point only in the Enlightenment (Botha 2013, 128).
The terminology of authorship did not exist in Greco-Roman antiquity, nor did the concept. It was not until the Renaissance that people began to be thought of in terms of materials they wrote, though previously the individual might be recognized as someone whose ideas could be used in arguments. It was not until the Enlightenment that "literary property properly emerged" (Botha 2013, 129). Botha therefore briefly reviews some elements of the conceptual and practical elements of authorship in antiquity (Botha 2013, 130). Pliny the elder, a very prolific (and therefore unusual) author, spent extensive time in research, being read to and dictating notes and extracts. Lucian recommended extensive gathering of notes, then some process of organization. Apparently notebooks of some sort were kept, at least by some (Botha 2013, 131). Eventually, a text would be dictated for transcription on a scroll. A text would be revised, sometimes multiple times, before it was considered a finished work (Botha 2013, 132). Because of the nature of dictation, in some instances it would result in notes which would be filled in later by the scribe (Botha 2013, 133).
Publication, or really, release of a literary work, was often done either without the name of an author, or under some other name (Botha 2013, 133). A dedication of a work is often made to someone. This serves as an indicator that the work is finished, and also associates the author in some way with a more prominent individual. It further implies that a copy of the work has been furnished to the person to whom it is dedicated, hence the work is published (Botha 2013, 134). Some publishers are known to have existed, receiving texts and arranging for copies to be made and distributed. It was through public reading that the work would become known.
Distribution of books occurred normally not initially by making copies, but by oral presentation in public and by lending a book to a friend (Botha 2013, 136). Any additional copies were normally made by a private copyist, and could be subject to revision, emendation and error.
Authors, so as to establish their identity, tended to insert statements which could serve as personal markers (Botha 2013, 137). This could assist a reader in knowing the actual source of a work.
Botha concludes that a concept of the author as understood in antiquity is necessary to properly interpret works (Botha 2013, 137). The author in that world is not the same as an author in our world. Further, the communal nature of presentation influenced the content. An awareness of the traditions and expectations of the setting is crucial (Botha 2013, 138). The work of composition, also communal in nature, would influence what would appear. The way material would be included or excluded at all stages of the composition process and the way a listener would gather information would all be influenced by other social contexts.