Luke’s account of Paul in Acts is best understood as drawing on multiple eyewitness sources within the Pauline circle rather than on Paul’s letters or Luke’s constant companionship with Paul. Barnabas, Silas, and possibly Timothy supplied information about different stages of Paul’s ministry, while Luke himself became an eyewitness only during Paul’s later years.
See also The Puzzle of Paul in Acts: A Historical and Linguistic Interpretation (Routledge, 2026).
By Nathan Nadeau
Toronto Baptist Seminary
Professor of Hellenistic Greek and New Testament
McMaster Divinity College
Research Fellow
August 2026
Introduction and Comments on Acts as (and in) History
It seems to me a good case can be made that the Lukan writings were central to two monumental controversies in the history of the interpretation of the New Testament. The first was the controversy surrounding Marcion of Sinope in the second century. This was precipitated by Marcion’s expulsion from the Christian community at Pontus by his father, his later rejection by churches in Rome, and the collection of his so-called canon, which included ten letters of Paul and a form of Luke’s Gospel (on Marcion’s canon see BeDuhn, 2013). The “Acts of the Apostles” (a second work by the author of Luke’s Gospel) was used—indeed, emerged onto the historical record—in the arguments of the proto-orthodox writers (especially Tertullian and Irenaeus) against Marcion. I think Ernst Haenchen was obviously right to see that this first appearance of Acts does not indicate its construction at that time: it was an earlier text (much earlier, by my lights) that now found a Sitz im Leben in the second century church (Haenchen 1971, 9).
The second controversy comes much closer to our own time. It happened in the course of the emergence of the “higher criticism” of the New Testament on the Continent in the eighteenth and nineteenth centuries. Jesus and the Gospels had been deconstructed early on in this movement, but the study of the historical Paul became ground zero in the efforts to found a liberal New Testament Christianity around ostensibly historical persons and reliable texts. The movement reached its apex in the middle of the 19th century, and, emerging out of the “most important theological event in the whole history of theology from the Reformation to the present day” (Harris 1975, v)—the founding of the Tübingen School—came a work by Ferdinand Christian Baur that will forever be remembered as both the master product of that school and that work which set a tone for the study of Paul up until the present. It may seem a curious thing, however, that Baur’s book (Paulus der Apostel Jesu Christi) has had its lasting impact in the study of Paul in Acts. But Acts was Baur’s first concern, and this set a tone for a critical discourse begun in that time which in many ways has continued unabated, circling around the historical Paul, his letters, and Acts. Given the crumbling of Baur’s critical paradigm (his Hegelianism), it is surprising that the dogmas that emerged from its foundations remained alive in the history of interpretation vis-à-vis Paul and Acts and in some sense live still today. Regarding these, Baur insisted that Acts is “scarcely worth attention” unless it “altogether agrees” with the letters (Baur 2021, 4).
Less surprising but no less interesting is the phenomenon begun in that time of scholars like Baur who took up the very same texts (the Lukan writings; the letters of Paul) for the same effort (a fresh inquiry into the historical nature of earliest Christianity), but came to starkly different conclusions about the coherence and reliability of Luke’s story of Paul. William M. Ramsay stands out as a fountainhead of the retreat from Tübingenism, but others are notable even if they admired the German school and in one sense represented its culture, while denying its import for Acts (Adolf Harnack).
My partiality to one side may seem clear in what follows, but without detailing the whole history of the modern interpretation of Acts I will simply say that Luke’s story of Paul remains an area of controversy. The usual areas of dispute are well known, including the genre of Acts (and the Gospel of Luke), the date of Acts, Paul’s chronology in Acts (especially vis-à-vis Galatians 1–2), the “we” material in Acts and the question of authorship, the speeches in Acts, etc. While all are in agreement that the author of Acts has constructed a sweeping and engaging narrative that at least appears to be a redoubtable work of history, some argue that “appearance” is all it is; “[it] ain’t [history], [it] just looks that way.”
I and others think differently. I think Acts is written by the Luke named in the tradition who was probably a Syrian from Antioch; I think the best explanation for the appearance of earnest history-writing in Acts is that the author believed he was a responsible writer of apostolic history with respect to his sources (which he reasonably deemed reliable) and interests. This is true even if we can discover pervading qualities of a literary nature in the narrative, since to be literary is the nature of historiography and it is a category error to draw the inference from the latter (its literary nature) against the former (its historical quality). This point should be obvious, but judging from some of the discourse on Acts (scholarly and otherwise), it yet needs to be said.
Yet what would follow if Acts was an accurate historical narrative? Luke might pass his geography and civics test, but does it follow that he has Paul right? Not strictly, no. He could have gotten things wrong; he could have shaped the monumental figure into his preferred image with nary a regard for the “facts of the matter” at this level; he could have been so sophisticated as to see that literary artifice and historical record can coexist in his work, and that he could take a heavy preference for the former as concerns the hero of his tale and the latter as concerns his setting. Some have more or less argued all of these, founding their theories in part on the fundament of seemingly clear and incontrovertible discrepancies between Paul, his letters, and Acts. But it is precisely these (sometimes merely assumed) discrepancies which I would first question.
For example, I think there is a pervasive agreement between Acts 9–15 and Galatians 1–2, and that old immovable contradictions vanish when the texts are studied closely from a linguistic and literary perspective. A linguistic analysis of Gal 1:13–2:10 shows beyond reasonable doubt that Paul is narrating a tight and strict chronology and omits no journey to Jerusalem: the repetition of time language, associated verbs of movement, and callback to the foundational event of the so-called conversion experience make this almost certain. This account easily fits within Acts 9–15, with the trip of Galatians 2 to Jerusalem corresponding to Acts 11:30/12:25. Partly in this light, the obstacle of the infamous “many days” of Acts 9:23 (apparently rendering Acts and Galatians irreconcilable for the former’s truncated chronology) must now be taken for what it is, a relatively long period of time wherein Paul was mainly in Damascus, the fuller length of which is (inclusively) counted by Paul in Galatians as “three years.” Ἱκανός and related terms, when collocated with a numeral, always signify relative substantiality and/or largeness. Thus, not only can the same phrase “many days” indicate months of time elsewhere (a possible inference from Acts 18:18; cf. Bruce 1990, 394–97; Murphy-O’Connor 2002, 164–67), it must indicate, in Acts 9, a long enough tenure for Paul to go from being unknown to the disciples there (Acts 9:19b–21) to having highly devoted disciples of his own (9:25).
I also think that the “we” passages in Acts stand out as a source with all the marks of reliability, with pervading mundane testimony (increasing and explicit time references throughout), coherence with Paul’s own autobiographical statements (especially as found in 2 Cor 11:22–28), use of place names, and, in the last “we” passage, details which have consistently been used by historians as the single best example of the historical realia it narrates (see especially Reiser 2001).
I further think that Paul’s most famous speech—that in Athens as narrated in Acts 17:16–34—has heretofore unrecognized, subtle, yet unmistakable Pauline fingerprints at the linguistic level, including marks of Paul’s rhetorical style. This opens up the live possibility of some fuller speech originating in Paul, now worked by Luke’s editorial hand (not unlike James’s speech in Acts 15 and its resonances with the epistle of James, whatever one makes of these). Long-held contentions of the speech being at odds with Romans 1 (see, classically, Vielhauer) should fade into the umbra of a failure to disambiguate kinds of arguments based in natural theology (natural law on the one hand, a subset of natural theology, on the other).
The rub in all of the above is the many details, puzzle pieces so to speak, that then need to be fit together without undue force. In much fuller treatments of the general subjects (genre and history; Acts and Galatians; “we” passages; Paul’s Athenian address), I have done so (see Nadeau 2026). But even if one is not convinced by new and detailed argumentation, we can at least all still come together around this one abiding question concerning the composition of Luke’s narrative: what were his sources?
If this remains a more engaging question across the board, then it is another way into the inviting question of historical possibilities: plausible identification of sources can shed light on Luke’s social position vis-à-vis Paul and the earliest Christian communities (Acts 11:26), his authorial interests, and more. While I think the preface to Luke’s Gospel is enough on its own to answer much of this, below I present a precis of a fuller case that Luke did of course use sources, but not the ones sometimes (and seemingly now more frequently) thought—Paul’s letters.
I do not think Luke used Paul’s letters, and I do not think Luke was a constant companion of Paul. I do think, however, that Luke had reliable sources for his story of Paul (and more) that can be named. Let me make my case in brief.
Did Luke Care about Eyewitnesses?
It’s not that “seeing is believing,” but, in one sense for the ancient Greeks, seeing was knowing (consider εἴδω, οἶδα, etc.; cf. Byrskog 2003, 281). It seems clear that when Luke mentions “eyewitnesses” (αὐτόπται) handing on information, in the context of a carefully crafted sentence affecting a historiographical mode more than any other kind of literature (Luke 1:1–4), he means what other historians meant when they used similar terms in similar contexts: cf. Herodotus, Hist. 2.29.1, 3.115.2, 4.16.1, 8.79.4, and 8.80.1; Polybius Hist. 3.4.13; Josephus C. Ap. 1.1, 1.55, 2.1. That is, he meant those who had directly looked at the events or were present in them. He also seems to be establishing himself as one who followed the contemporary strictures of good history writing, where, by his day, a historian “must seek out reliable people, including those who played a leading role in the events being considered” (Marincola 2017, xxxix). Was Luke himself such an eyewitness? The tradition has assumed “yes,” and argued that the whole story of Paul in Acts can be accounted for on this basis. That is, that Luke was a constant companion of Paul (Irenaeus is a case in point [Adv. Haer. 3.14.1]; see also Keener 2012–2016, 3:2361–62). But I am not so sure.
The simple reason is this: Paul is fairly consistent at naming his various and sundry co-workers, generally locating them in the times and places they worked with him. He does so eighty-three times by my count across his letters in the New Testament (I’m a maximalist with respect to Pauline authorship but won’t argue for it here). The tradition would have us believe that Luke, alongside Silas and Timothy, accompanied Paul on the Macedonian mission (according to the “we” material in Acts 16), but where Silas and Timothy enjoy feature positions in the Thessalonian correspondence that came out of the mission, Luke is nowhere to be found. Hardly consistent with Paul’s practice elsewhere. And it is not that Paul never names Luke; he does, and does so incidentally across three letters (Colossians, 2 Timothy, and Philemon) which all locate Luke as being with Paul in only the very latter part of his ministry. I think this settles the matter that Luke, who wrote the Gospel and Acts, was only with Paul at this later time.
This brings some precision to the question of Luke as an eyewitness and his eyewitness sources which can now be taken up with a reappraisal of the “we” material.
Luke and the “We” Material
If, as argued above, Luke was an eyewitness late in Paul’s missionary tenure, he would be located in his own narrative right around the time of a passage that stands apart in the entire Lukan corpus: Acts 27. “One of the most literary sections of Acts” (Dibelius 1956, 205), the sea voyage and shipwreck in Acts 27:1–28:16 is unique for being (1) a “we” passage; (2) a famously historically detailed account; and (3) perhaps the most literarily elevated section of the work. Points 2 and 3 could be part of a larger case for a participant’s own memory. The redoubtable nature of the seafaring narrative from a historical perspective is certainly expected on the view that it comes from an eyewitness (see Reiser 2001), especially when one considers how some details read as if they are recorded by a non-specialist. And in this passage we find one of the few direct examples of Homeric allusion, where on the lips of Paul—whether by Luke’s memory, summary, elevation, or all three—we read in perfect meter and position, the following: οὐδενὸς . . . ὑμῶν θρὶξ ἀπὸ τῆς κεφαλῆς ἀπολεῖται (“the hair of not one of you shall be lost from your head”) (see Reece 2022, 11–12). The convergence of extensive historical detail with dramatic expression befits an eyewitness memory. A survey of the other “we” material makes the case stronger.
Acts 27–28 is taken to contain the last of the “we” material in Acts, in relation to similar material in Acts 16, 20, and 21. These are notoriously difficult passages to square with the rest of the work; the first three anyway. Unlike the narratively smooth material in Acts 27–28, the former “we” material tends to end where one textual unit begins or to begin where such a unit ends. The former material cuts against the structural grain of the narrative and comes on with no rhyme and little direct reason; it resides mostly in verbal endings and when in pronouns, more often than not in oblique grammatical cases (accusative, genitive, and dative). Furthermore, the first three “we” passages, though separated by non-first-person narratives, are tightly joined in geography: where one ends, the next begins. The first ends in Philippi (Acts 16:10–17), the next picks up there (20:6); that one ends in Miletus (20:15), where the next picks up (21:1); that one ends in Jerusalem (21:17). If three makes a pattern, it is broken by the aforementioned fourth (Acts 27:1–28:16)—which begins in a different locale (Caesarea Maritima) and is separated by Paul’s lengthy solo imprisonment in Caesarea (Acts 24) and the commencement of his Roman trial. As stated, this latter material has no loose ends or beginnings with respect to the narrative structure, and the fact that Luke is mentioned at just this time in Paul’s letters (alongside other companions named in Paul’s letters and named in the sea voyage in Acts like Aristarchus) seems to demonstrate further that Luke was an eyewitness for Paul here. It is at least the simplest explanation, accounting for the varied data, and one which enjoys the support of the tradition without taking on its more contentious claim that Luke was Paul’s constant companion.
If this is true, what of the rest of Luke’s story of Paul? Did Paul tell him everything, or is there reason to think Luke had other, prior, sources? There is reason to say “yes” to the latter, and to contend that these sources can be named.
Paul’s Early Period
Paul’s early chronology is probably a more technical nut to crack than the “we” material, and one I certainly cannot run through the analytical machine presently. As stated above, I think the story of Paul between Acts 9 and 15 actually lines up quite nicely with a close reading of Galatians 1 and 2 and other Pauline autobiographical texts. While perhaps controversial to some, one need only acknowledge that the similarities between Acts and Galatians are so significant as to cause many to propose that Luke is cribbing from Paul’s letter here. An interesting theory, but it is too ad hoc an explanation to be plausible (but some cases are stronger than others, such as that made by Schellenberg 2015). Important differences concern the depiction of Paul in Jerusalem, the depiction of time, and Luke’s apparent ignorance of, or dismissal of, things not only mentioned by Paul, but emphasized by him. We need an explanation that does justice to the apparent earnestness of Luke as an author and his account’s similarities with Paul’s letters on the one hand, and to the real differences between Luke and Paul on the other. Such an explanation can be found in positing a source other than Luke’s artifice, Paul the man, or Paul’s letters.
Acts 9:27 portrays Barnabas as an official narrator of Paul’s earliest period after his conversion: “Barnabas brought [Paul] to the apostles and narrated to them how on the road he saw the Lord and that he spoke to him and how in Damascus he spoke openly in the name of Jesus.” The term for “narrate” here is διηγέομαι, the noun form of which occurs in the preface to Luke–Acts in reference to a “[historical] narrative” (διήγησις) of the “many” with respect to the story of Jesus. The noun is a hapax legomenon in the New Testament, while five of the eight New Testament uses of the verb are found in Luke–Acts. In every instance besides the citation of LXX Isa 53:7–8 in Acts 8:33, the term refers to a narration of past, usually recent, events. In Luke 8:39 and 9:10, these comprise “as much as . . . [was] done” (ὅσα σοι ἐποίησεν in 8:39; ὅσα ἐποίησαν in 9:10).
Likewise in Acts 14:27, Barnabas and Paul are said to “relate” or “report” (ἀναγγέλλω) “as much as God had done with them” (ἀνήγγελλον ὅσα ἐποίησεν ὁ θεὸς μετ’ αὐτῶν). When Paul and Barnabas are at the Jerusalem council narrating the same account, Luke uses the same language: “and they reported as much as God had done with them” (ἀνήγγειλάν τε ὅσα ὁ θεὸς ἐποίησεν μετ’ αὐτῶν). It is thus clear that one aspect of Barnabas’s significance in Luke’s story of Paul is his status as a teller of that story.
With this in mind, much sense comes to the fuller portrait of Barnabas and his potential role for Luke. Barnabas’s identity as a Cyprian Jew (Acts 4:36), for example, helps to answer the question of Luke’s source material for the first mission on which Barnabas (and John Mark) accompanied Paul. Details within the first leg of the first mission on Cyprus stand out in this regard, with unimportant seaports like that of Salamis in Cyprus on the one end and the identification of the coastal destination from Paphos as “Perga” on the other; these befit a local’s information more than a distant historian’s sketch (on “Perga,” see Hemer 1989, 109). The narrative of Sergius Paullus also reads with a historical verisimilitude, as his family is otherwise known from inscriptions in the south of the Roman province of Galatia and on Cyprus (Breytenbach 1996, 38–45, 180–88). Paul of course could be a source for this information, but it is more expected on the view that Barnabas was. Acts 9:27 and 14:27 thus look like veritable bookends that establish the plausibility of Barnabas as a source for Luke. When compared with the major alternatives often advocated for in the literature (an ad-hoc “itinerary source”; an unprovenanced “Antiochene source”; theories of reliance on Paul’s letters, or assumption of sole reliance on Paul), the solution of Barnabas as a major participant in the events seems the best of the lot. Where did Luke meet Barnabas? Likely at Antioch, Luke’s traditional home, and the place where Barnabas presided over the conversion of many, perhaps Luke himself (Acts 11:19–30).
Yet Barnabas leaves the scene after the Jerusalem Council, and so Luke would be left without his key witness. What comes next seems to support a larger view which I can now state plainly: Luke put together his story of Paul on the basis of multiple eyewitness participants who were part of the Pauline circle. Indeed, as soon as Barnabas departs, another arrives, and we encounter our second major source for Paul in his second missionary partner: Silas.
Back to the “We” Material and Beyond
I have already carved out a special place for the latter “we” material in Acts 27–28 by considering Luke’s role as an eyewitness there. But what of that other, somewhat separate, and stranger “we” material? Dibelius called it strange, not me, but I have already elucidated that it simply does not fit neatly within the textual structure of its context. I speak of Acts 16:10–17, Acts 20:5–15, and Acts 21:1–18. I will focus here on Acts 16, but draw some conclusions about the rest of the “we” material as well. As mentioned above, Paul features Timothy, and especially Silas, in the Thessalonian letters which emerged out of the adventures in Macedonia.
It is striking that Paul actually singles out Silas as an author over against Timothy in this regard. The “we” of 1 Thessalonians does not always include Timothy but presumably does Silas: consider 1 Thess 3:2, 6, where “us” includes Paul and Silas, but not Timothy. The same effect of identifying the “we” by use of “us” is at work in Acts 16:17. “Us” separates a plural group besides Paul, who, together with the “us,” made up the “we” of 16:10. Hard to follow? It goes like this: there are three and only three named persons in the “we” of Acts 16:10 if we follow what is explicit in the narrative: Paul, Silas, and Timothy. When we later read “Paul and us” in v. 17, we are left with Silas and Timothy. But this means we can identify the author/eyewitness in this first-person narrative as Silas and/or Timothy. Silas has already been established by Luke as a special partner with Paul in the wake of the split with Barnabas; Silas is otherwise established as an authorial hand in the Thessalonian correspondence; and important linguistic parallels between those letters and Acts 16:10–17 can be shown, involving the notion of the “gospel,” its proclamation, and first-person plural language. When we see also in Acts 16 an emphasis on Philippi as a Roman colony and consider in turn Silas’s status as a Roman citizen, when we consider a theme of suffering across Acts 16:16–40 that parallels not only the same theme in 1 Thess, but also in 1 Peter—where Silas is very likely identified as Peter’s scribe (1 Pet 5:12)—it seems clear that Silas, like Barnabas, stands as a live option and source for this leg of the mission. While Silas is not mentioned after Acts 18:5, Timothy is, and he is part of the travelling band in the next “we” passage (Acts 20:4) and thus may have carried on the task of writing a sea-travel diary.
I would say more about these two, since a puzzle can even be pieced together which places them clearly with Paul in Athens at some point, and which may explain newly discovered parallels between Paul’s speech in Athens and Paul’s own linguistic style (see Baum 2006; Nadeau 2026).
I can only suppose where Luke collaborated with these latter members of the Pauline circle. The lengthy Caesarean imprisonment is the best candidate, as Paul was given a degree of freedom and “his own” were allowed to care for him (Acts 24:23–27). Since this went on for some time (v. 27), it is possible that companions attended to Paul there and to his writing needs (with an analogy in Paul’s later request of Timothy, according to 2 Tim 4:13 [an analogy which retains significance whatever one thinks of Pauline authorship of the Pastorals]). Luke must have ventured to meet and eventually join Paul in his ongoing trial only at this point. Luke thus may have met Silas again there (having first met him at Antioch following the Jerusalem Council) and may have met Timothy there as well. He would have received their own notes from them, potentially put together or edited by Silas and which formed a sea-travel diary that Luke worked into his narrative, careful to leave the “we” intact wherever he found it, for its historiographical significance.
Too many “may haves”? Consider the alternatives: unprovenanced diaries of unnamed individuals; a fictive “we” so problematically intermixed against the grain of its narrative as to be dead on arrival for its stated purpose of creating historical credibility; a similarly inexplicable sporadic use of the first-person by Luke who, the traditional view would have us believe, was nonetheless always there. What I have supposed here is a supposition, but it is one made only after more firmly securing the borders of the puzzle to the extent that it enjoys a stronger rationale than the others.
Conclusion
Arguments like those surveyed above, when brought together, suggest that Luke’s story of Paul, whatever else it is, is an earnest account based on personal involvement within the narrative, and with the small circle of Paul’s major companions. Instead of an impossibly abstract world of literary artifice or purely hypothetical sources, we encounter a much more realistic scenario of a handful of capable eyewitnesses who worked together to pass down the living tradition of the story of Paul. Who else would we expect?
Bibliography
Baum, A. D. “Paulinismen in den Missionsreden des lukanischen Paulus: Zur inhaltlichen Authentizität der oratio recta in der Apostelgeschichte.” ETL 82 (2006), 405–36.
BeDuhn, Jason D. The First New Testament: Marcion’s Scriptural Canon. Salem, OR: Polebridge, 2013.
Breytenbach, Cilliers. Paulus und Barnabas in der Provinz Galatien. AGJU 37. Leiden: Brill, 1996.
Bruce, F. F. The Acts of the Apostles: The Greek Text with Introduction and Commentary. 3rd ed. Grand Rapids: Eerdmans; Leicester, UK: Apollos, 1990.
Byrskog, Samuel. “History or Story in Acts—a Middle Way? The ‘We’ Passages, Historical Intertexture, and Oral History.” In Contextualizing Acts: Lukan Narrative and Greco-Roman Discourse, ed. Todd Penner and Caroline Vander Stichele, 257–81. SBLSymS 20. Atlanta: Society of Biblical Literature, 2003.
Dibelius, Martin. Studies in the Acts of the Apostles. Ed. Heinrich Greeven. London: SCM, 1956.
Hemer, Colin J. The Book of Acts in the Setting of Hellenistic History. Ed. Conrad H. Gempf. WUNT 49. Tübingen: Mohr Siebeck, 1989.
Murphy-O’Connor, Jerome. St. Paul’s Corinth: Texts and Archaeology. 3rd rev. and exp. ed. Collegeville, MN: Liturgical, 2002.
Nadeau, Nathan. The Puzzle of Paul in Acts: A Historical and Linguistic Interpretation. London: Routledge, 2026.
Reece, Steve. The Formal Education of the Author of Luke–Acts. LNTS 669. London: T&T Clark, 2022.
Reiser, Marius. “Von Caesarea nach Malta: Literarischer Charakter und historische Glaubwürdigkeit von Act 27.” In Das Ende des Paulus: Historische, theologische und literaturgeschichtliche Aspekte, ed. Friedrich Wilhelm Horn, 49–74. BZNW 106. Berlin: De Gruyter, 2001.
Schellenberg, Ryan S. “The First Pauline Chronologist? Paul’s Itinerary in the Letters and in Acts.” JBL 134 (2015), 193–213.