Literary evidence of peculiar strength may be adduced in favor of the view that the Book of Acts was really written, as tradition affirms, by a companion of Paul. This evidence is based primarily upon the presence in the book of certain sections where the narrative is carried on in the first person instead of the third. It is generally or even universally admitted that these "we-sections" are the work of an eyewitness, an actual traveling companion of Paul. But according to the common-sense view--according to the first impression made upon every ordinary reader--the author of the we-sections was also the author of the whole book, who when he came in his narrative to those parts of the missionary journeys of Paul where he had actually been present with the apostolic company naturally dropped into the use of the first person instead of the third. If this common-sense view be incorrect, then a later author who produced the completed book has in the we-sections simply made use of an eyewitness source. But this hypothesis is fraught with the most serious difficulty. If the author of the completed book, writing at a time long after the time of Paul, was in the we-sections using the work of a companion of Paul, why did he not either say that he was quoting or else change the "we" of the source to "they." The first person plural, used without explanation by a writer of, say, 100 A.D. in a narrative of the journeys of Paul, would be preposterous. What could be the explanation of so extraordinary a procedure? Only two explanations are possible. In the first place, the author may have retained the "we" with deceitful intent, with the intent of producing the false impression that he himself was a companion of Paul. This hypothesis is fraught with insuperable difficulty and is generally rejected. In the second place, the author may have retained the "we" because he was a mere compiler, copying out his sources with mechanical accuracy, and so unable to make the simple editorial change of "we" to "they." This hypothesis is excluded by the striking similarity of language and style between the we-sections and the rest of Luke-Acts, which shows that if the author of the completed double work is in the we-sections making use of a source written by some one else, he has revised the source so as to make it conform to his own style. But if he revised the source, he was no mere compiler, and therefore could not have retained the first person plural which in the completed book produced nonsense. The whole hypothesis therefore breaks down.