Garrow, Alan J.P. "Chapter Six: The Prophet Document: 10.1-7; 11.7-9, 12; 12.1-5." The Gospel of Matthew's Dependence on the Didache. New York: T&T Clark International, 2004, 107-112.
Garrow has concluded from a redaction criticism perspective that the eucharistic prayers in Didache 9.1-5a come from a different layer of redaction than the prayers in 10.1-6. He now observes that 10.7 and 11.7 show a strong affinity (Garrow 2004, 107). This leads him to conclude that the two verses appeared together in a different document, then were separated and brought into the text of the Didache. 10.1-6 lead naturally into 10.7, so he takes those verses to belong together. Likewise, 11.7-9 make coherent sense together, so Garrow takes them to be part of the document which was brought in (Garrow 2004, 107). Garrow calls this the "Prophet document, because of its concern for prophets and prophecy" (Garrow 2004, 108). In contrast, Garrow considers 11.10-11 to interrupt the logical flow, so as to belong to a different source. The chapters do differ in some ways. Of note in Garrow's mind is a different setting. "In 9.5, there is a presupposition of the eucharistic president's knowledge of each member of the congregatin, while 10.6 implies the presence of a larger, more anonymous gathering, suggestive of a later development in the Church's life" (Garrow 2004, 108-109).
Garrow notes that Didache 12.1-5 shows a change in situation as well. While 11.4 speaks of receiving apostles, 12.1 does not limit the welcome to apostles (Garrow 2004, 109). This suggests to Garrow that the material may belong to a separate layer of redaction, now generalizing instructions about itinerant prophets to be applied to all itinerant people. The limit placed on the time of hospitality suggests to Garrow a climate of prophets making specific journeys rather than living permanently as itinerant propheets. Those people would need more opportunity to receive hospitality (Garrow 2004, 110).
Garrow takes the material to be a "prophet document," not merely a layer. He considers it to be a text that was interpolated without significant loss or change (Garrow 2004, 111).