This blog, by Richard Fellows, discusses historical questions concerning Paul's letters, his co-workers, Acts, and chronology.

Monday, September 13, 2010

"we passages", Thucydides, Polybius, Josephus, & W.S. Campbell

Here I discuss a paper by William Campbell, "The Narrator as "He", "ME," and "We": Grammatical Person in Ancient Histories and in the Acts of the Apostles", JBL 129, no. 2 (2010): 385-407. I will use Campbell's observations to show that the author of Acts was present during the events of the "we passages" and at other times too.

A writer can refer to himself/herself in the first person singular ("I", "me"), the first person plural ("we", "us"), or using the third person, as Richard does now. Campbell explores the styles of self-reference by the ancient historians, Thucydides, Polybius, and Josephus, and compares them with Acts. He distinguishes two types of self-designation: narrator level and event level.

Narrator level
Narrative-level self-reference is when the author refers to himself in his capacity as author/narrator. Campbell gives examples from Thucydides' History of the Peloponnesian War: "Thucydides, an Athenian, wrote the history of the war..."; "the earliest sea battle of those we know". Campbell tells us that Polybius referred to himself frequently (using first person singular) at the narrator level. E.g. "I was persuaded to write"; "I do not think it seemly...". Examples from Josephus include, "as we said previously"; "I would not hesitate to say"; "we Jews".

Concerning Acts, Campbell writes,
Another difference between Acts and the four texts analyzed is how infrequently first person (singular or plural) appears at the narrator level. The narrator rarely intrudes in the story to offer commentary, which compels the narrative to stand on its own merits. Only two instances each of first person singular and plural narrator-level comments occur in Luke and Acts, and these are located in the prefaces (Luke 1:1-3 and Acts 1:1).
My interpretation:
Luke's reluctance to intrude with narrator level self-reference should come as no surprise since his works, like the other gospels, are anonymous. Baum (Nov. Test. 50 (2008)) argues that the purpose of this anonymity was to give priority to the subject matter. Excessive self-reference would distract attention from the narrated events.

Event level
Event level self-reference is when the author refers to his own participation in the historical events. Campbell shows that the third person is generally preferred:
Large sections of War are devoted to Josephus's involvement in the conflict, and once again, as was the case in Thucydides and Polybius, event-level passages with Josephus as a character - indeed the main character and protagonist - are narrated in the third person. One brief example is the account of his appointment as commander of Galilee: "John, son of Ananias, was appointed commander of Gophna and Acrabetta, and Josephus, son of Matthias, of each of the two Galilees" (J.W. 2.568).
Here is an example from Thucydides:
4.104.4 The opponents of the betrayers .... sent to the other commander of the areas in Thrace, Thucydides, son of Olorus .... 
Campbell gives the following passage from Polybius, which has striking parallels with Acts.

36.11.1 When instructions arrived in the Peloponnese from Manilius for the Achaeans that they would do well to send Polybius the Megalopolitan with Haste to Lilybaeum, as there was need of him for affairs of state, the Achaeans resolved to send him in accordance with the petition of the consul. 36.11.2 We, thinking it our duty for many reasons to obey the Romans, putting aside all other matters, set sail when summer began. 36.11.3 Arriving in Corcyra and finding there a letter from the consuls that had been sent to the Corcyraeans in which they made quite clear that the Carthaginians had already handed over the hostages to them and were prepared in every way to obey them, 36.11.4 thinking that the war had been brought to end and there was no further any need of uswe sailed back again to the Peloponnese.

My interpretation:
Since Luke refers to himself very sparingly at the narrator level we should expect him to do the same at the event level. Therefore he may well have been present for more of the events than he indicates. If Luke was present only for the 'we passages', as is commonly assumed, he mentioned his presence at every available opportunity, and this would not be consistent with his tendency to minimize self-reference. Indeed, I have argued before, here, that the author of Acts was present with Paul on his journey to Troas (Acts 16:1-9) before the "we passage" of Acts 16:10ff, and was again present in Achaia (Acts 20:3-4) before the "we passage" of Acts 20:6ff). 

Since these historians referred to themselves by name at the event level, we should be open to the possibility that Luke did too (albeit more sparingly than the other historians). The convention of anonymity followed by Luke and the other gospel writers required that the author not be identified (obviously). However, it did not require that no reference should be made to the author by name, because otherwise no-one would have accepted the view that the fourth gospel was written by John, or that the first was written by Matthew. I have argued that Luke did indeed refer to himself by name once (Lucius of Acts 13:1).

Of the three forms of self-reference available to Luke (first person singular, first person plural, and third person), first person plural was the least obtrusive. It was the form that drew least attention to the author as an individual and so we should not be surprised that Luke preferred it.

The passage from Polybius above is particularly interesting because the author switches to the first person plural at precisely the point where he sets sail, just as Luke does (Acts 16:10; 20:6; 27:1). This supports the view of Vernon Robbins that first person narrative was preferred for sea voyages (but not his view that the author need not have been present on the voyage).

It seems to me, therefore, that Acts combines the conventions of the historians about styles of self-reference with the convention of the gospels about limiting self-reference. He limits narrator level self-reference to his prefaces, and gives his name (Lucius/Luke) only once (Acts 13:1). He chose to make explicit reference to his participation only for sea voyages and their aftermath, where convention allowed him to use the relatively unobtrusive first person plural. For land events he omitted reference to himself completely, not because he was absent, but because convention required the use of the third person, "Lucius", which he did not want to overdo.

Campbell's evidence shows that when historians used the first person they were indeed claiming to have been participants of the events described. Therefore, Luke was (claiming to be) present. This is, in any case, common sense.

Campbell's deductions about Acts
It is therefore baffling that Campbell uses the same data to conclude that:
The analysis of grammatical practice by Thucydides, Polybius, and Josephus raises questions about the traditional argument that the first person plural undeniably documents historical eyewitnessing.
 His reasoning seems to be this:
If the author of Acts wished to lift up his historical presence at and involvement in events, the grammatical guidance offered by Thucydides, Polybius, and Josephus would seem to be that he identify himself by name, associate himself unmistakably with the narrator character, and report events in which he claims participation primarily in the third person.
 He assumes that "the author of Acts never appears as a third person actor in the narrative".

Campbell has failed to make allowance for Luke's desire to avoid excessive self-reference that would detract from the centrality of his subject matter. I have shown above that Luke follows the grammatical guidance of the other historians as much as was possible without obtrusive self-reference.

Moreover, Campbell does not explain how the "we passages" could have been understood by the original readers as implying anything other than the actual presence of the author.

Campbell's paper is valuable because he has made some important observations about the styles of self-designation employed by ancient historians. It is a real pity that his analysis of Acts misses the point completely.


  1. Hi Richard:

    Excellent, as always. But, who are "we" in 20:6?

    "[Paul] decided to return through Macedonia. Sopater he Berean, son of Pyrrhus, accompanied HIM; and of the Thessalonians, Aristarchus and Secundus; and Gaius of Derbe, and Timothy; and the Asians, Tychicus and Trophimus. THESE went on ahead and were waiting for US at Troas, but WE sailed away from Philippi after the days of Unleavened Bread, and in five days we came to THEM at Troas, where we stayed for seven days"

    Who was with Luke?

  2. The "we" of Acts 20:6 was comprised of Lucius-Luke, and Paul, I think. Both were Jews (see Rom 16:21) and this would explain why they wanted to celebrate the festival.

    Also, I suspect that the travelers split into two groups for security reasons. The larger group would have carried the money. Lucius-Luke was from the east and, assuming he was Lucius of Cyrene, had probably spent time in Judea. Therefore people would naturally assume that he was part of the plan to deliver money to Judea. So it would be safer for him to travel separately from the money, in case he, like Paul, was under surveillance. See my earlier post on the plot against Paul here.