As I write an article on why historians no longer trust Acts, and now categorize it as mythography rather than history (though it emulates a history), I realize one question needs to be settled separately, because otherwise it’s just too tedious a digression: why the canonical book of Acts at only three brief sections in its second half flips into a first person plural narration. Christian apologists often cite this as evidence the author of Acts was one of these people and therefore “was really there” and thus a reliable source. None of that follows—liars can pretend to have been there; and people who were there can lie about everything anyway; so if we accumulate evidence that the author of Acts (traditionally said to be Luke) was a habitual liar and fabricator, the whole notion that he is reliable merely because he occasionally uses a first person narrative falls apart anyway. But apart from that fallacy, is the premise even true?

There is a long and complex literature on this question, with a plethora of theories advanced and defended as to why these “we” passages exist, ranging from “the author was there” or “the author at these points copied from a first-person source he found, maybe a ship’s log” to “this was a common feature of fiction” or “the author is pretending to have been a witness to these events.” And every theory has its problems. See William Sanger Campbell’s thorough study in The We Passages in the Acts of the Apostles: The Narrator as Narrative Character (Society of Biblical Literature 2007). It must be noted first that a first person plural indicates a collective, not an individual perspective. It therefore isn’t directly linked to an author, but to someone identifying with a group or community. That could be the author—or a source, or a fiction, or a ruse. So…what do I think?

Ancient Fiction

There are indeed three (and only three) small sections of Acts where the text inexplicably jumps into the first person plural, and then out again (each one longer than the next: 16:9–18; 20:4–21:19; and 26:32–28:17, which is interrupted by a brief switch-back, in 27:29-28:1, but still from the same collective perspective—“we” are still there the whole time—so though some scholars count four sections of first person narrative, it is really three). The fact that the author never explains why they do this indicates it had no relation to their actually being present. If that is what they meant, that is what they’d have said—like, “that is where I joined Paul” or “then we parted,” or some other kind of introduction explaining why the personal voice is changing precisely there and not elsewhere, or indeed at all. In all other ancient literature, a first person voice is explained when that is what it meant. If the author means they were there, they say so (either there or contextually elsewhere in the work so the reader knows). Indeed, they especially did this, to establish authority for their account. Which is also why authors often named their sources, something Luke never does, not even in these passages, which is weird. However, ancient fiction did on occasion incorporate first person plurals (as did histories taking only a narrative pose), indeed even unannounced like this, to heighten excitement or drama, or reader participation.

If someone were to say “such switches are too subtle and too infrequent to constitute a fictitious apologetic ploy,” they’d have it exactly backwards. To the contrary, they are too subtle and too infrequent to constitute an author identifying their personal involvement. This odd abruptness would more resemble fiction—a stylistic choice, rather than an epistemic or rhetorical one. Of course there could be other causes of these odd transitions, like layers of redaction in the composing of Luke (e.g. “maybe he is copying a source here,” though that wouldn’t mean the source itself was any less fiction), but while apologists will denounce those as being speculation sans evidence, fact is, so is their theory. There is no evidence the author means they were there either. They never say so, even though they surely would have. So the apologist’s theory of this indicating the author was there is also challenged by the facts. The facts are simply weird no matter what you propose to explain them.

So in a field of theories none of which have any clear evidence to back them, we are left to play the odds on prior trends. Even when some historians (like Polybius and Josephus) switch in and out of the first person plural, it is made clear (by one device or another) who is meant, and that they are the ones there, and why the switch has occurred. This isn’t happening in Acts. Was it usual for authors to signal their presence in events they are relating with a sudden and unexplained introduction of the first person plural in an otherwise anonymous narrative? No. In fact, there are zero examples of that across ancient literature. Are there any examples of this happening in fiction? Yes. The device was used in the book of Jubilees and the Odyssey, for example. And Acts employed both as models.

Indeed, a sandwich of Jubilees allusions conspicuously frames the first two “we” sections in Acts (Zachary K. Dawson, “The Books of Acts and Jubilees in Dialogue: A Literary-Intertextual Analysis of the Noahide Laws in Acts 15 and 21,” Journal of Greco-Roman Christianity and Judaism 13 (2017): 9-40). But more significant is that all three of the “we” sections in Acts span three extended sea narratives, the only extended sea narratives in Acts; and the first person plural is deployed nowhere else in Acts. So unless the source or the author was a sea captain, it is not plausible to take this as indicating the author was there on only those occasions. And again, if he was the ship’s captain (composing the log, say), if that’s why he only appears when a ship is around, he’d say so. Yet, instead, the author keeps distinguishing himself (and hence the “we” and “us”) from the crews (and captains) of the ships they’re on.

So this odd conjunction is a highly unlikely coincidence—unless it is indicative. As it happens, it was a popular literary style to narrate sea adventures in the first person plural, even in the midst of narratives otherwise given in the third person, or even first person singular. This was famously documented, with extensive examples, in Vernon K. Robbins, “By Land and By Sea: The We-Passages and Ancient Sea Voyages,” Perspectives on Luke-Acts (1978), pp. 215-242. As Robbins points out, lots of pre-Homeric texts used first person singular narration for sea voyages, yet “Homer’s Odyssey, in contrast, contains the earliest [known] example among Mediterranean literature of a sea voyage that employs first person plural narration.” It is thus in fact a prominent and distinctive feature of that Homeric epic, which inspired emulation in many subsequent authors, including the most preeminent emulator of the Odyssey, Virgil’s Aeneid, which by the time Luke wrote was the national epic of the founding of Rome, both of which Acts is particularly attempting to rival in producing a parallel epic of the founding of the Church (as numerous scholars have documented, as recently summarized by Michael Kochenash in “Reconsidering Luke-Acts and Virgil’s Aeneid: Negotiating Ethnic Legacies,” in Christian Origins and the New Testament in the Greco-Roman Context: Essays in Honor of Dennis R. MacDonald (2016)). It’s therefore of no small importance that Luke-Acts repeatedly employs both epics as literary models.

That many authors emulated this narrative style, even producing the abrupt unexplained transition from third person to first person plural, establishes that there isn’t anything unusual about the author of Acts picking up the same device. For example, in the pre-Christian text Voyage of Hanno, from its introduction the text cuts inexplicably from third to first person and stays in that mode thereafter:

It pleased the Carthaginians that Hanno should voyage outside the Pillars of Hercules, and found cities of the Libyphoenicians. And he set forth with sixty ships of fifty oars, and a multitude of men and women, to the number of thirty thousand, and with wheat and other provisions. After passing through the Pillars we went on and sailed for two days’ journey beyond, where we founded the first city…

The same happens in sea transitions in the Odyssey. For example, book 9 starts, “Odysseus, of many wiles, answered him … ‘let me tell you also of my woeful home-coming, which Zeus laid upon me as I came from Troy … to Ismarus, where I sacked the city and slew its men, and from it we took their wives…’,” proceeding thereafter in first person plural, with only occasional and otherwise never-explained reverts to a first person singular. Robins documents how the sea voyages in Acts emulate numerous features common to sea voyage narratives, particularly first-person plural narratives, both historical and fictional (and pseudo-historical), and indeed these features are peculiar to these sections of Acts (missing from the remainder), demonstrating deliberate emulation of the entire literary genre, bolstering the explanation of also incorporating the feature of first person plural narration also only here.

Luke’s employment of this same device is therefore plausibly an emulation of literary fashion, even if his deployment of it looks awkward owing to his book’s peculiar maintenance of anonymity. That clashes with the usual glorifying of narrators in fiction (it is, for example, obvious to a hearer of the Odyssey that every “we” and “us” means Odysseus and his crew). But that fits Luke’s pattern of reversing exactly this theme in popular mythmaking. For example, in his Emmaus narrative Luke has Jesus emulate the legend of the resurrected Romulus (see Richard Miller’s study Resurrection and Reception in Early Christianity, and my discussion in On the Historicity of Jesus, Ch. 4.1), only in reverse: appearing in humble disguise, rather than glorious visage, and concordantly preaching humility and submission rather than conquest. So this explains all the oddities of the text (why peculiarly only here, why peculiarly in this way), something no other theory does, while it requires the fewest ad hoc assumptions, since it rests on established precedents. It is therefore the most likely explanation.

This answers the objections of Susan Marie Praeder in “The Problem of First Person Narration in Acts,” Novum Testamentum 29.3 (1987). She agrees the “we” passages cannot indicate the actual author’s presence (because they don’t follow the pattern of such indications elsewhere in ancient literature, nor any authorial logic at all), but she is also skeptical of the “sea voyage” thesis because not all sea narratives use this device, nor any combine it with anonymity. But our argument is not that “all” such narratives use this device so Luke “had” to use it; our argument is that many did, and so Luke chose to emulate that (particularly as it was prominent in his favorite models, Homer and Virgil). Praeder advances no objection to our actual argument. And we all acknowledge Luke’s strict anonymity is simply weird, so that it isn’t common elsewhere is not any objection to our argument either. The device appears in Jubilees, but not to the same purpose, though it could have served as an inspiration (Luke was aware of it). And in any case, our argument is that combining these two features is Luke’s own innovation.

Literary Devices

This has further support from Robyn Faith Walsh’s study of the literary methods and intentions of the authors of the Gospels and Acts (see Robyn Faith Walsh and the Gospels as Literature). That they would deliberately craft narratives in fashionable and artful ways like this is what authors like them would typically do. So it has prior probability. And we have ample demonstrations of these authors doing this a lot (see my extensive survey of examples in On the Historicity of Jesus, Chs. 9 and 10). Which ups that prior here even more. When we look at Luke’s inaugural use of “us” to draw the reader in as sharing an identity with these adventurers and experiencers of pinnacle moments, and his frequent use of Homeric models to craft his stories, even down to choices of vocabulary and plot elements and messaging (see Dennis MacDonald, Does the New Testament Imitate Homer? Four Cases from the Acts of the Apostles), including especially the Odyssey and Aeneid (itself imitating the Odyssey) and their sea voyage material (see Dennis MacDonald, Luke and the Politics of Homeric Imitation: Luke–Acts as Rival to the Aeneid, esp. pp. 12-29), there is nothing ad hoc left about this proposal.

Indeed, you might be interested to know that this entire three-voyage sequence of “we” passages begins with Paul setting sail from the region of Troy (and the second “we” passage then coincidentally takes place in that same city of Troy, after weirdly having gone back). The actual city in these accounts, Troas, was a more recent namesake city in the vicinity of the old (and more famous) Troy, then called Ilium; so the coincidence of name and location had obvious literary value to Luke. Lest you forgot, both the Odyssey and the Aeneid narrate sporadic sea journeys originating at old Troy (Odysseus attempting to get home after his sack of Troy; Aeneas fleeing to find a new home after that very same sack). Coincidence? Probably not.

The author of Acts is deliberately playing on this motif, and signaling it by having Paul leave (what was then the more modern port city of) Troy the moment he begins inserting the dramatic “we” constructions, then revives the motif upon his return to that Troy, and then revives it again with a dramatic parallel to fleeing the original Troy unto eventual salvation, when Paul begins his final journey from Judea to Rome, after surviving his own de facto “war” (even dodging hundreds of troops under armed guard) over his own de facto “bride” (the Church; in the Odyssey, this is Helen in rearview and Penelope in foreview), complete with a shipwreck interrupting his voyage, also emulating Aeneas’s journey to his future kingdom (creating an analogy between the founding myth of Rome and the founding myth of the Church). And indeed, all three of these sea narratives borrow elements and motifs extensively from the Odyssey (documented in MacDonald, Luke and Politics) and Aeneid (documented in Marianne Palmer Bonz, The Past as Legacy: Luke-Acts and Ancient Epic; see also Neil Godfrey, “Acts and Virgil’s Aeneid: Comparison and Influence” at Vridar). It is unlikely these conjunctions are accidental.

In Dennis MacDonald’s ever-useful Synopses of Epic, Tragedy, and the Gospels he provides a detailed survey of all the emulations of Homer and Virgil in the book of Acts (pp. 359ff.). In our sections of interest, the emulations run:

Odysseus’s we-voyage from Troy (Odyssey 9.1-42)	Aeneas’s we-voyage from Troy (Aeneid 3.4-17)	Paul’s we-voyage from Troas (Acts 16:9-12)
Death of Elpenor (Odyssey 10.552-574, 11:51-80, 12:7-12)	Deaths of Palinurus & Misenus (Aeneid 5.587-871, 6.156-371)	Death (and resurrection) of Eutychus (Acts 20:7-12)
Hector’s farewell to Andromache & Astyanax (Iliad 6.360-502)	Aeneas’s farewells to Creusa & Ascanius (Aeneid 2.671-789, 12.430-443)	Paul’s farewell to the Ephesians (Acts 20:13-38)
Odysseus’s shipwreck (Odyssey 5.269-454)	Aeneas’s shipwreck (Aeneid 1.34-119)	Paul’s shipwreck (Acts 27:9-44)
Odysseus’s welcome by the Phaeaceans (Odyssey 5.467-6.244)	Aeneas’s welcome by the Carthaginians (Aeneid 1.173-613)	Paul’s welcome by the Maltese (Acts 28:1-10)

Within each emulation, features of vocabulary, plot, order, and messaging are shared. Of greatest interest, though, are aspects of plot—as that relates to what is supposed to have actually happened, and not just in what language it is described. If that ends up being fiction, so is the rest. For example:

Odyssey 10–12	Acts 20:5–12
Odysseus and crew are leaving Troy en route to Achaea	Paul and crew arrive at Troas en route from Achaea (Corinth: Acts 18)
Narration in first-person plural begins	Narration in first-person plural begins
After a sojourn, they have a meal	After a sojourn, they have a meal
Disaster comes at night	Disaster comes at night
The crew sleeps in Circe’s “darkened halls”	“There were plenty of lamps in the upper room”
Narration switches to third person with “There was one, Elpenor, the youngest”	Narration switches to third person with “A certain young man named Eutychus”
Elpenor (whose name means “Hopeful” but he is declared to be “unlucky”) falls into “sweet sleep”	Eutychus (whose name means “Lucky”) falls into “a deep sleep”
Elpenor falls off the roof	Eutychus falls from a third-story window
Elpenor’s soul goes to Hades	Eutychus’s soul remains in him
Associates fetch his dead body	Associates fetch his dead body
… but only after they meet him in Hades	… but Paul declares he isn’t really dead
Elpenor is then buried at dawn	Eutychus then comes back to life at dawn

Even more parallels structure the shipwrecks of the Odyssey and Acts, complete with angelic help and washing ashore at an island, and being at first feared but then mistaken for a god.

Needless to say, all this literary art and emulation casts doubt on any of this being historical. Is it likely that Paul would get into three sea adventures paralleling Odysseus and Aeneas in numerous odd details? Not really. Which means the “we” claiming to have been there through all of those adventures (but never elsewhere) are no more real—just as they are not real people in Jubilees, the Odyssey, or the Aeneid. I think it likely that these voyages represent in Luke’s design the entire Christian mission, through the central hero of Paul. By switching only then to the first person plural, and only at sea, and only in the final climactic sequence of Acts (chapters 16 to 28), I think Luke intends his readers (who are meant to be Christians, each a “Lover of God” or “Beloved of God,” the literal meaning of his thus-probably-fictional addressee Theophilus) to feel that they are on that ship with Paul.

Conclusion

Luke is thus aiming at an affective fallacy, to seduce his readers into believing the gospel is true because they feel they have shared in Paul’s adventures, travails, and triumphs. They have sailed the dangerous seas and been carried to safety on the ship of faith. The “we” thus means all Christians, inclusive of the reader. This explains also why this happens only in the second half of Acts: precisely when Paul begins to spread the faith to the whole world. For the first time Paul crosses a sea (a symbol of mortal danger; and a ship, a symbol of common purpose in facing it) to spread the Gospel. And this is at the behest of a vision of a “man of Macedonia” telling him to come save “us,” leaving it no coincidence that the first “we” narrative begins literally in the next sentence. So we have a literary clue here who the “we” really are: all of us. Hence as Robins puts it, “the author uses the we-sections to create a special role for mission by sea.” Indeed, I suspect Luke invented “Macedonia” as the catalyst for it because it was the homeland of Alexander the Great, the first to conquer the Eastern Mediterranean and Hellenize it, after which his empire was subsumed by Rome: Paul is passing through history, first going to conquer the Greek Empire on his way to conquer the Roman. Paul is going to conquer the world, subsuming all empires. There are also three journeys, representing the three days of Jesus’s victory, from death to resurrection, so Paul can accomplish one journey to conquer Greece, one journey to conquer Jerusalem, and one journey to conquer Rome.

I suspect this also explains the brief interlude in the final sea journey, when the text even more inexplicably switches back to the third person and back into first (27:29-28:1), even though it is clear the “we” are still continually there, so their absence cannot explain the change in narrative person. This occurs at the depths of the shipwreck tale. I believe this represents, by literary device, a symbolic death and resurrection for the whole Christian body. We are to emotionally feel that “we” have figuratively died (leaving only the third person agents to sort everything out while we are dead), after which “we” have figuratively been raised back from the dead. Exactly in the middle of this interlude is a single first-person line, “Altogether there were 276 of us on board,” as if to advertise the souls that could be lost if not saved. Before this line, everyone assumes they are as good as dead, but Paul gives them hope by providing them a makeshift Eucharist meal and words of faith and encouragement. Then after this line, everyone finds salvation ashore. The moment they step onto land, the first person narrative resumes. Salvation achieved; the “tomb” of the sea escaped; our figurative resurrection complete.

There are many clever literary allusions to this sense being intended. For example, they are shipwrecked only after setting out for the town of “Phoenix,” the famous resurrecting bird. And the entire ordeal begins (it is implied) on the day after the Day of Atonement (marking the sacrifice of universal atonement that Jesus’s death replaced) and ends fourteen days later, which counting inclusively would be the day after the festival of Sukkot ends and the Day of Blessing and Assembly concludes that cycle of ceremonies (using the count of days applied to the Diaspora and not within Judea), which cycle is also a theme in Jubilees. This all makes another death-and-resurrection metaphor. This sacred period had also famously been shortened to account for bad weather endangering pilgrims, triggering the entire bad-weather-endangerment theme of the entire scene the author constructs. And the switch from first to third person narrative when all hope is lost (the figurative death) turns precisely on the revelation that this was their third day at sea, evoking (and thus anticipating) the fates of Jesus and Jonah. Those heroes rise on the third day. Paul and his crew figuratively die on the third day. But hope is not lost. They will be saved on the Day of Blessing and Assembly, the concluding event of the Atonement Day cycle.

This take on the text (as well as the rest I have explained here) would be supported if we find Acts is full of fictionalizing and mythmaking and literary art like this, leaving this even more plausible. And lo and behold, we will. In my next article I will summarize that case, which will better contextualize this one.