Wenham, John. "Chapter Ten: How Were the Gospels Written?" Redating Matthew, Mark, and Luke: A Fresh Assault on the Synoptic Problem. Downers Grove, Illinois: InterVarsity Press, 1992, 198-216.
Wenham considers it essential that we remember that Jesus was an oral preacher and teacher in a culture where orality and detailed memorization of words and order of events was common. At the same time, it was also common for teachers and scholars to make notes for mnemonic purposes (Wenham 1992, 199). Jesus and others would have been quite capable of using Aramaic or Greek, and possibly Hebrew as well. Wenham suggests the possibility of both specific recollections of particular wording of interactions and written recollections which may have been consulted (Wenham 1992, 200). Written records would have become more common as the need for a definitive account which could be consulted in the absence of the eyewitnesses arose. This would be seen as a means of preserving the truth.
Matthew was traditionally understood as the first to write a definitive gospel account (Wenham 1992, 201). Wenham does not consider the composition to have been done hurriedly. He even considers that Matthew may well have composed the greater discourses first, then arranged other events around them to guide the overall emphasis of the gospel to the passion account. In doing so, Matthew may have used papyrus transcripts of the teachings of the other apostles, at least as a prompt to his memory (Wenham 1992, 202).
Wenham has more uncertainty about Mark's gospel. He takes it to use Matthew as a model and also to be based on Peter's preaching (Wenham 1992, 202). As Peter was an eyewitness to the events, he would not have been dependent on Matthew, though he may well have had access to a copy of Matthew (Wenham 1992, 203). Wenham is open to theories of literary dependence or dependence on an oral tradition which had some specific idiomatic phrasing.
Wenham reviews accounts of manuscript preparation, noting that copyists did not use a writing table or desk until the late Middle Ages. This would have made a task of composition which drew from and adapted multiple written sources very difficult (Wenham 1992, 204-205). Rearranging and revising texts was considered high level scholarly work. Wenham considers it highly unlikely that any of the canonical Gospels would have been composed in this way (Wenham 1992, 206).
Luke's Gospel, from the prologue, was intended to set testimony in order (Wenham 1992, 209). Wenham considers Luke to have had access to Matthew and Mark. While Matthew would be an "eyewitness," Mark would have been an "assistant," each representing an element of Luke's research. Luke seems to have narrated most of the material which was used by both Matthew and Mark, but to have used Mark's organizational structure for the most part. Many of Luke's differences have to do with geography, about which his record tends to be more accurate than that of the other evangelists (Wenham 1992, 211).
Wenham finally moves to a brief discussion of the different genealogies of Jesus. The two defy harmonization. early Chrsitian traditions suggest the genealogies were provided by different branches of the family (Wenham 1992, 214). Wenham tends to think the genealogy in Matthew is intended to show royal succession while Luke intends to show natural familial succession. Another popular view suggests that Luke has Mary's family while Matthew has Joesph's, thus the legal heritage (Wenham 1992, 216).