f. Conclusions regarding Historical Tradition

f. Conclusions regarding Historical Tradition

That Luke both preserves and varies some of his direct-discourse material when he repeats it indicates that Luke himself made no secret that he was not reporting speeches verbatim (compare Acts 9:5–6 with 22:8, 10; 9:15–16 and 22:10 with 26:16–18; 10:5–6 with 11:13–14; compare also 9:4 with 26:14; and especially 1:4–8 with Luke 24:47–49). (He makes clear that he used summaries [Acts 2:40; Luke 3:18].) That ancient history is a genre that differs in this respect from modern history should also lay to rest any attempt of modern readers to anachronistically demand of Luke a kind of writing that he does not claim to offer. Following the practice of the best ancient historians, Luke could employ (and would probably have needed to employ) at least some more freedom in his speeches than in his narrative (likely the opposite of the situation in his Gospel, given the parallel material in Matthew).

Further, the speeches in Acts display the sort of literary unity that characterized the speeches of ancient historians in general. Yet historians also sought to provide speeches appropriate to the speakers, and elements distinctive to particular speakers in Acts may represent the ancient rhetorical practice of borrowing elements of speakers’ style or language deemed appropriate to them when writing speeches for them.

At the same time, it is reasonable to argue that although Luke authored the speeches in their current form, he also preserved their gist when he had this information available, like many other historians. At the very least, he seeks to preserve the authentic theological message and traits of the speakers, albeit selectively and adjusted to fit his larger work; one cannot consciously borrow distinctive elements of a speaker’s style or message unless one has some familiarity with that style or message. Further, I have argued above that Luke has oral eyewitness sources for most of the events recorded in Acts (sometimes, I will argue, Luke himself). The sources that provided the events in his narrative could also have supplied him with the gist of speeches given on significant occasions, which were themselves “events” (especially in view of his work’s focus precisely on the spreading of the gospel message). Although clues are not pervasive, enough elements of early tradition may appear to suggest that Luke was familiar with the theology, argumentation, and sometimes even style of some early Christian preachers, the strongest case among them being for Paul.

Long ago C. H. Dodd suggested that the author “apparently used to some extent the liberty which all ancient historians claimed (after the example of Thucydides), of composing speeches which are put into the mouth of the personages of the story.” But he further concluded (though not always on the most objective grounds) that Luke apparently used “his historian’s privilege with considerable restraint”; he argued that the speeches reflected pre-Lukan Christian tradition of the apostolic kerygma. It may be relevant that, for all the first-century debates about circumcision and the role of Gentiles (cf. Gal 2:2–5), there seems to have been consensus between Paul and the Jerusalem apostles on the basic apostolic message of the resurrection (1 Cor 15:1–11; cf. Gal 2:7–8). Unity of a core message is not Luke’s invention, though it may serve his literary and theological purposes and receive more emphasis than some other elements that he could have included.

Such conclusions accord well with a centrist position adopted by many NT scholars regarding the speeches in Acts. From a literary standpoint, those who claim that the speeches are Lukan compositions are correct. From a historical standpoint, however, Luke could well have generally preserved the substance of what was said on the various occasions (and other “primitive” material in apostolic tradition plausible for the speakers), when this information was available to him. Virtually no mainstream scholars have argued that the speeches, especially those early in Acts, are verbatim compositions. But many argue that these speeches do preserve the gist of what was said on the occasions. At minimum, many of the speeches preserve genuine reminiscences and reflect traditions about earlier Christian preaching.

Of course, scholars continue to debate how much of the speeches as we have them represents the original gist and which speeches might reflect prior traditions. Beyond the secure information that Luke had, he could have employed the freedom of the genre of ancient history to present what could plausibly be inferred from what was known. He could have developed the sort of argument expected for particular speakers, filling in with details like those that the speakers were reported to have sometimes used or at least were of the kind that they would have probably used. Thus Dunn argues that Acts’ speeches reflect Luke’s summary of speakers’ views as best he can reconstruct them; ancient readers would not expect more than this.

In the final analysis, I have provided only a basic generic context for the discussion here; those who look for form-critical signs of underlying tradition must examine one speech after another, and even such detailed analysis can yield at most probabilities in historical terms. Given our limited historical data, the historical method is limited in its results, restricting the potential conclusions we can articulate on historical grounds here. Although I will pause at some points to interact with previous scholarship and to note Pauline parallels where these prove relevant for particular speeches, my focus in the commentary proper will be on matters on which historical study can speak with fuller probable results—namely, the speeches in their context in Acts and their function for a late first-century Greco-Roman audience.

With, e.g., Laird, “Rhetoric,” 204.

Walbank, Speeches, 1 (excepting Cratippus and Pompeius Trogus). This recognition is widely shared, e.g., Feldherr, “Introduction,” 2.

The observation of Damon, “Rhetoric,” 444, regarding the examples in Wiseman, Historiography. Even A. J. Woodman’s broader critique of narratives based on inferences is limited in its reach (Damon, “Rhetoric,” 445–46).

For a systematic and historical review of scholarly approaches to Acts’ speeches, see esp. Soards, Speeches, 1–11.

Often noted, e.g., Judge, First Christians, 380. Historiographic speech-writing customs vary in different periods; thus, e.g., it was customary in the nineteenth century to reword verb tenses throughout the entire speech (Tomkins, Wilberforce, 225).

This practice regarding speeches differs from Jesus’s sayings in the Synoptics, the wording of which tends to be preserved more carefully than that of the narratives (see discussion below concerning the transmission of sayings, as opposed to entire speeches).

Even where they had existing speeches available, failing to rewrite them made historians appear rhetorically deficient, at least those writing for elite audiences.

The exception would be speeches for which there were no surviving witnesses (clearly invented) or for which the writer’s sources remain extant.

To put the matter differently, what does it mean for Luke to be a good ancient historian when we study the question of speeches? To try to make him a good modern historian is to force his work into a genre that, in its current form, could not yet exist. Nevertheless, the most nuanced approach allows us to identify a range of speech-composition practices in ancient historiography and to ask where in this range Luke falls (or where particular speeches fall; e.g., most scholars quickly conclude that Luke 6:20–49 hews close to its source [cf. esp. Allison, Jesus, 309–51] but that the conversation in Acts 25:14–27 represents free composition).

For studies of Luke’s speeches, see Grässer, Forschungen, 99–115, 254–70; before 1965, see also Mattill and Mattill, Bibliography, 165–73, §§2220–2338.

See Shiell, Reading Acts, 102–36 (esp. 133). Literary cues would signal proper changes in vocal inflection (see in detail 83–89, addressing such cues as indignation, irony, ridicule, flattery, or other emotionally evocative tones and postures).

Kennedy, Classical Rhetoric, 129. See more specific parallels in Ellis, “Midrashic Features”; Bowker, “Proem and Yelammedenu Form.” I do recognize that conventional Jewish rhetoric predominates in some early Christian material (emphasized in Keener, “Study of Rhetoric”).

See Wills, “Form of Sermon”; and esp. the response in Black, “Form of Sermon”; in Christian preaching, see Bauckham, “Summaries,” 191–204 (Ascension of Isaiah) and 204–9 (Ignatius).

As is widely recognized (Lieberman, Hellenism, 47–82; Hengel, Judaism and Hellenism, 1:80ff.; Levine, Hellenism, 113–16); for Hillel’s seven rules, cf. Cic. Inv. 2.40.116 (Stambaugh and Balch, Environment, 103; Hengel, Judaism and Hellenism, 1:81). E.g., with qal vaomer, compare Arist. Rhet. 2.23.4–5, 1397b; gematria has Greek roots (Sambursky, “Gematria”); notarikon may have Babylonian origins (Cavigneaux, “Herméneutique”); even the later pilpulistic hair-splitting of the rabbis resembles sophistic discussion (Diog. Laert. 2.30). Naturally, this was especially obvious regarding allegory (with, e.g., Sandmel, Judaism, 113; Mickelsen, Interpreting, 27).

See Keener, “Rhetorical Techniques”; Neyrey, “Forensic Defense Speech”; and other sources cited loc. cit.

Barrett, Acts, 2:lxxx; Porter, Paul in Acts, 124.

Cic. Brut. 93.324.

Kennedy, “Survey of Rhetoric,” 15. Porciani, “Enigma,” 333–34, suggests that Thucydides may have actually condensed some available materials for his speeches.

Not, of course, several hours in length. They could be condensed (Pliny Ep. 1.16.4), but Luke’s are unusually brief.

E.g., Soards, Speeches, 141.

See, e.g., Neyrey, “Forensic Defense Speech”; Keener, “Rhetorical Techniques.”

Kennedy, New Testament Interpretation, 115. I agree in Paul’s case (though not, perhaps, John’s); most of Paul’s letters are carefully crafted (see, e.g., Harvey, Listening, passim).

Dion. Hal. Isoc. 2, 3, 12 critiques Isocrates for piling up ornate rhetorical devices generating needless words. For his Atticist purity, see, e.g., Dion. Hal. Lysias 2; also Rebenich, “Historical Prose,” 292–93.

Dion. Hal. Lysias 3. When simply illustrating a point, a rhetorician could articulate simply for clarity (Hermog. Inv. 1.1.94–95), a point relevant to Luke’s mere speech summaries.

Dion. Hal. Lysias 4.

On the sometimes obscure style of Paul’s letters, see Anderson, Rhetorical Theory, 280–81.

E.g., Borgman, Way, 11, counts 19 major speeches; Dibelius, Studies in Acts, 150, counts 24 speeches; Eckey, Apostelgeschichte, 30–31, counts 26; Aune, Environment, 124–25, counts 32.

Smith, Studies, 9, 40.

See Schubert, “Final Cycle”; he believes that this does not fit Thucydides’s approach. Cf. Martinson, “Ending,” 316, viewing Acts as proclamation (though incorrectly counting this as Acts’ “unique” genre rather than distinctive focus); Penner, “Discourse,” 84 (though exaggerating in seeing the speeches as so central that Luke lacks interest in historical narrative). Pervo, “Direct Speech,” 302, acknowledges the widespread attribution of Luke’s focus on speeches to his proclamation theme, but himself attributes it to the popular genre instead.

Soards, Speeches, 183n3, estimates higher than Horsley because Horsley counts “only . . .  the major speeches”; but presumably he counts only those in other historians as well. It may well be because Pervo counts all direct speech and not simply orations that his figures for novels run much higher than for history (note the concession in “Direct Speech,” 302).

Horsley, “Speeches,” 613. Laird, “Speech presentation,” rightly emphasizes the prominence of direct-discourse monologues in ancient narratives but curiously plays them down in ancient historiography; Pervo, “Direct Speech,” does the same, taking the highest count for Acts and the lowest for ancient historians generally.

Horsley, “Speeches.”

Aune, Environment, 125.

Ibid., 91, offering other examples as well. Also Aune, Dictionary of Rhetoric, 139; although Homer’s poetic genre differs, it became the model for subsequent narratives (447). Ancients understood that poets composed speech material freely (Vell. Paterc. 1.3.2–3).

Aune, Environment, 124–25. Ehrman, Introduction, 133, suggests “nearly a quarter” for the average history. On indirect and direct speech in Polybius (the former often giving way to the latter), see Usher, “Oratio Recta.”

Aune, Environment, 125.

It does not fit drama at all (with, e.g., its staging, choruses, poetic lines, and barely any narrative) whereas most other genres have less speech material.

With others in addition to Schubert, e.g., Witherington, Acts, 118, 665.

Johnson, Acts, 10 (citing Acts 1:4–8, 16–22; 2:14–40; 3:12–26; 4:8–12, 24–30; 5:35–39; 7:2–53; apart from 1:4–8, these are evident).

Horsley, “Speeches and Dialogues”; cf. Cadbury, Making, 184.

Biographers regularly wove sayings and stories together; e.g., Philost. Vit. soph. 2.5.574; see more fully Keener, Matthew, 33–36; Burridge, Gospels, 203; cf. the large amount of speech material in Iambl. V.P. 8–11. In Roman histories after Sallust, following Thucydides’s model, see, e.g., Cary and Haarhoff, Life, 263. For rhetorical instructions concerning how to insert sayings into narratives, see Theon Progymn. 4.73–79; cf. 5.388–441.

For one brief survey of the discussion, see van Unnik, “Storm Center,” 26; still more briefly, Powell, Acts, 30–31. Contrast here Witherington, “Addendum,” 24, 28–29; idem, Acts, 48, 117, 455.

Soards, Speeches, 9–11; Porter, Paul in Acts, 127–28.

Cf. Tannehill, Acts, 329; Bauckham, “Summaries.” Cf. Bruce, “Apostolic Succession,” who applies the apostolic preaching in Acts as a model for modern ministers.

On John, cf. Keener, John, 53–80, esp. 69.

E.g., Dion. Hal. Isoc. 20; Theon Progymn. 1.74–76; Diog. Laert. 7.1.42 (on Stoics); Hermog. Issues 34.21–35.2; Men. Rhet. 1.1.331.4–9; Nicolaus Progymn. 1.pref. 3; discussion in Anderson, Rhetorical Theory, 97; Calboli, “Genera.” In greater detail, see Kennedy, Art of Rhetoric, 7–18, for forensic; 18–21, for deliberative; and 21–23, for epideictic oratory.

See, e.g., Aune, Environment, 91 (classifying speeches in the Iliad).

Garver, “Aristotle,” noting that they did not all fit these categories even in Aristotle’s time.

Ancients recognized the diversity of style among artists (Fronto Ad verum imp. 1.1.1), poets (1.1.2), philosophers (1.1.3), and others.

With Watson, “Speech,” 184. In my view less convincingly, he places Acts 20:17–38 in the epideictic category because it is a farewell speech (190–91), but though it may praise Paul for Luke’s readers (191; but farewell speeches normally praised the place being left, not the speaker), its function in the story world appears more deliberative. Not surprisingly, “the vast majority of historiographical speeches” were deliberative (Marincola, “Speeches,” 127; for Livy, also Laistner, Historians, 96), with many forensic cases as well (Marincola, “Speeches,” 128).

True epideictic was typically elaborate and sometimes rhythmic, however (Hermog. Inv. 1.5.106–8; 4.1.171; Fronto Ad M. Caes. 3.16.1), so it fits better some christological passages in the Pauline corpus (often viewed, perhaps inappropriately, as hymns) rather than speeches in Acts.

Mixed rhetorical genres were common (Aune, Dictionary of Rhetoric, 419).

Heath, Hermogenes, 9–10; Rhet. Her. 1 (for judicial speeches); 3.7–9 (deliberative); 3.11–15 (epideictic); Diog. Laert. 7.1.43 (among Stoics); Nicolaus Progymn. 1.pref. 4; 11, “On Ecphrasis,” 69–70; for examples, see Dion. Hal. Lysias 9; Tac. Dial. 19.

See, e.g., Laird, “Rhetoric,” 204–6.

See, e.g., Rodríguez Ruiz, “Discurso misionero.” Naturally, there is the risk of imposing too much of the taxonomy of classical ideals on genuine texts (Black, Rhetoric of Gospel, 21); flexibility was intrinsic to good rhetoric (Cic. Quinct. 10.35).

Horsley, “Speeches,” 609.

Quint. Inst. 10.7.21.

E.g., Pliny Ep. 2.3.1–3. Hermog. Method 17.433, thus advises feigning this approach for jurors; commenting, Kennedy adds Quint. Inst. 10.6.6.

See Winter, Philo and Paul, 205–6.

On the importance of delivery, see, e.g., Cic. Brut. 17.55; 37.141; 43.158; 66.234; 82.283; 91.316; Rhet. Her. 3.11.19; Plut. Demosth. 7.2–3; 11.2–3; Pliny Ep. 2.3.9; 2.19.2–4, 6; 5.20.3; 6.11.2; Suet. Claud. 30; 41.2.

E.g., Isoc. Panath. 3; Cic. Quinct. 1.1–4; 24.77; 26.80–27.85; Sall. Jug. 85.31; Sp. G. Cotta 4; Quint. Inst. 4.1.8–9, 11; cf. Ovid Metam. 13.137; Pliny Ep. 8.3.3; Suet. Tit. 7.1; Tac. Hist. 4.73. It is used often enough by Dio Chrysostom (Or. 1.9; 32.39; 46.7; 47.1, 8).

See Winter, Philo and Paul, 205–6.

Laird, “Rhetoric,” 206–7.

On ancient distinctions among purpose, theme, speaker, and form of speeches in terms of types of causes in Aristotle’s Physica, see Poster, “Affections,” 35n20.

See, e.g., Dunn, Acts, xvii–xviii. Varying genres within a work aided its reception (Pliny Ep. 6.33.7–9).

Laird, “Rhetoric,” 207.

Laird, “Rhetoric,” 208–9. One may compare how preachers apply biblical texts to their audiences, although some ancient historians may have been more creative.

See Plümacher, “Missionsreden.” For a comparison of the function of Acts’ mission speeches with the function of the speeches in Dionysius and Livy, see idem, Lukas, 33–38.

Rosner, “Biblical History,” 76. For the interpretive function of OT speeches, see also Konkel, Kings, 23.

Most scholars think that Luke uses speeches to advance his purpose (e.g., Strelan, Artemis, 23). We examine Luke’s theology in both speeches and narrative more extensively in below.

Lindner, “Geschichtsauffassung”; Attridge, “Historiography,” 326; Marincola, “Speeches,” 119. Israelite historical works, like Acts, often used speeches to summarize a unit or move the narrative forward to the following unit (Rosner, “Biblical History,” 76).

Cadbury, Making, 184.

Despite brief hints that some of the speeches should be understood as more rhetorically proficient, as in, e.g., Acts 17:22–31; 24:10–21.

Thucyd. 1.22.1; Satterthwaite, “Acts,” 355–56. Speeches in Thucydides provide his estimate but not primarily his personal views; rather, they expound “what Thucydides thought would have seemed to him the factors in a given situation, had he stood in the place of his speakers” (Finley, Thucydides, 96); they function in connection with the rest of the work (100).

Plümacher, Geschichte, 109–26, comparing Luke.

Aune, Environment, 91. On Herodotus’s special liberties, see Kennedy, Classical Rhetoric, 110; he was “the first [known] historian to use direct speech for dramatization” (Aune, Dictionary of Rhetoric, 448).

Cadbury, Making, 185.

Endres, Interpretation, 198–99.

Cf. Schubert, “Final Cycle”; Martinson, “Ending,” 316.

See, e.g., Ferreira, “Plan.”

Gempf, “Speaking,” 261. This can remain true in political history (cf. the prominence of speeches’ historic role in the British Parliament in, e.g., Metaxas, Grace, passim).

See, e.g., Bruce, Commentary, 24. Besides speeches with a deliberate apologetic function in the narrative world, Luke seems to employ most outsiders’ speeches apologetically, in a manner comparable to other Jewish narrative (including historiographic) works (see esp. Padilla, Speeches). Emphasizing apologetic elements pervasive in the speeches, Mauck, Trial, 62–64, argues that most speeches reveal that the Christian movement is genuinely Jewish or that the Christians are evangelists rather than subversives.

See esp. Downing, “Pagan Theism and Speeches,” comparing common arguments in Josephus and Dionysius.

See those in the later apocryphal acts in Pervo, “Rhetoric in Apocrypha.”

See, e.g., Watson, “Speech”; Veltman, “Defense Speeches”; Neyrey, “Forensic Defense Speech”; Keener, “Rhetorical Techniques.”

So Parsons, “Progymnasmata,” 43–44 (arguing further [44–63] that he probably had learned from the sorts of rhetorical exercises in the handbooks); Asso, “Raconter” (though some would critique his emphasis on Aristotle for this period).

E.g., probably occasionally prose rhythm (Aune, Dictionary of Rhetoric, 382).

Thus, e.g., the Italian Sallust composes a letter for Mithridates that is more anti-Roman than what Pompeius Trogus, a Gaul, composes (Adler, “Anti-Roman”).

Soards, Speeches, 142–43. Penner, “Discourse,” 84, thinks that Luke as historian is so interested in speeches that much of the narrative simply sets the stage for them and that hence Luke is not much interested in history; I would argue, however, that Luke’s emphasis on speeches fits the normal generic expectations for history, and especially histories the focus of which included substantial speaking roles (see discussion in – above).

As noted in , a historical monograph would involve a specific focus. Historians desired compositional unity, including in the speeches (Polyb. 3.32.2; Diod. Sic. 20.1.5–9). Nevertheless, most included more presentation of varied perspectives than Luke does (though cf. Acts 5:34–39; 19:24–27, 35–40).

Burridge, “Gospels and Acts,” 520.

Satterthwaite, “Acts,” 355–56, citing many supporting references.

History must remain unified like a living being (Diod. Sic. 20.1.5–9); one should use suitable words in speeches (20.2.2).

Plümacher, Lukas, 33–38; for English summary, see Soards, Speeches, 137–38, but note his disagreement, 142–43.

Plümacher, “Mission Speeches”; idem, “Missionsreden.”

Aune, Environment, 127, citing Lucian Hist. 47–48 and Jos. Ag. Ap. 1.49–50 on rewriting, and the absence of speeches in “unfinished” parts of Thucydides (Thucyd. 5.10–83; 8) and Herodian (Hdn. 5–8). The lack of speeches in part of Thucydides, however, may stem from an experiment, as Cratippus suggested (Wade-Gery, “Thucydides,” 1518).

Cadbury, Foakes-Jackson, and Lake, “Writing History,” 13–14.

E.g., Theon Progymn. 1.93–171; Libanius Anecdote 1.4; 2.3; Maxim 1.2–5; 2.3; 3.2; Hermog. Method 24.440; cf., e.g., Epict. Diatr. 1.9.23–25 with the Loeb note referring to Plato Apol. 29C, 28E (Oldfather, LCL, 1:70–71). On paraphrase in secondary education, see also Hock, “Paul and Education,” 202–3.

Contrast tedious repetition in some earlier literature, e.g., Hom. Il. 8.402–8, 416–22 (except the change from first to third person). Some Palestinian Jewish sources also employed paraphrase (e.g., 4Q422, a homiletic paraphrase of Genesis; Elgvin, “Section”).

Cf., e.g., the redactional structure of m. ʾAbot 2:9, where Johanan ben Zakkai asks five disciples a question in positive form, commending the answer of the fifth; when he repeats this in negative form, he again receives mainly the same answers in negative form and again commends the fifth.

See Dibelius, Studies in Acts, 150–51; Pervo, “Direct Speech,” 291; Padilla, “Παιδεία,” 432–34. For speeches in John, including dialogues, see Keener, John, 53–80. Opposing speeches offered historians an opportunity to display rhetorical skill (Protagoras may have offered the philosophic foundation for that approach; cf. Aune, Dictionary of Rhetoric, 36), which may be one significant reason that Luke barely includes them.

Cf., e.g., Lévy, “Conversation”; Pernot, “Rendez-vous”; though for the dialogue form, see Keener, John, 61–62, 65–66.

Aune, Environment, 127. Aune, 91, follows W. C. West in counting 52 direct speeches in Thucydides in addition to 85 in indirect discourse. Pervo, “Direct Speech,” 303, attributes Luke’s preference for a high proportion of direct discourse to his work’s popular genre; for other suggestions, not necessarily incompatible with this one, see above.

Horsley, “Speeches and Dialogues,” thinks that it occurs in eight of ten speeches to conceal the speeches’ brevity.

See ibid. This is not to mention the more common literary forms of dialogue and imaginary interlocutors.

E.g., Cic. Prov. cons. 8.18; Caesar C.W. 2.33 (positively); Mus. Ruf. 3, p. 38.25–26; Plut. Demosth. 6.3; 8.5; Lect. 11, Mor. 43BC; Dio Chrys. Or. 15.26–32; Pliny Ep. 9.13.19; Lucian Dem. 14; Aul. Gel. 8.10; 18.13.7–8; 20.10.1–6; Diog. Laert. 7.1.19; ʾAbot R. Nat. 6 A; t. Sanh. 7:10. Often in court settings, e.g., Plut. Cic. 16.3; Pliny Ep. 3.9.25; 3.20.3 (in the senate); Philost. Vit. soph. 2.30.623; in church, perhaps 1 Cor 14:34–35. Crowd interaction was normal (Lucian Z. Rants 41; contrast Aeschines Tim. 35); lack of interaction was rude (Pliny Ep. 6.17.1–4; cf. Eunapius Lives 460), but so was excessive interruption (Plut. Lect. 4, Mor. 39CD; 18, Mor. 48AB; cf. Sen. E. Controv. 3.pref. 4–5). For instructions how to handle interruptions, see Rhet. Alex. 18, 1432b.35–40; 1433a.14–18; Cic. Or. Brut. 40.138 (less seriously, Lucian Prof. P.S. 20).

For imaginary interlocutors, see, e.g., Demosth. Philip. 3.15; Cic. Rosc. com. 3.8–9; Verr. 2.3.20.50–51; Tusc. 3.23.55; Sen. Y. Dial. 3.6.1; Ep. Lucil. 42.2; 66.40; 92.14; 94.32; 102.8–10, 18; Mus. Ruf. 4, p. 46.13; 12, p. 86.20; 15, p. 98.18, 26; Jos. Life 340; Epict. Diatr. 1.1.23–25; Dio Chrys. Or. 50.9; Pliny Ep. 7.17.5; Fronto Ad M. Caes. 3.12; Max. Tyre 3.5; 25.4; 27.8; y. Sanh. 6:1, §1; Pesiq. Rab. 13:7. On rhetorical conventions regarding this and related figures, see Rhet. Her. 4.23.33 (cf. 4.16.23–24); Cic. Or. Brut. 40.137; Anderson, Glossary, 14, 18, 33, 51–52, 58, 108, 124; Stowers, Diatribe, 128–29. For examples in Epictetus, see Tobin, Rhetoric in Contexts, 93.

Soards, Speeches, 141–42; Witherington, Acts, 46.

Cf., e.g., Polyb. 9.28.1–31.6; 9.33.3–39.7.

Rather, “cameos” (Dunn, Acts, xviii), “abstracts” (Hemer, Acts in History, 427), or “synopses” (Gasque, “Speeches,” 249).

See Witherington, Acts, 46. Others did sometimes abbreviate speeches found in their sources (Aune, Environment, 82, cites here Tac. Ann. 11.23–25; Jos. Ant. 12.417–18), and Rhet. Alex. 22, 1434b.11–18 provides directions for how to abbreviate speeches (circumstances also could dictate abridging some sections, e.g., Men. Rhet. 2.3, 379.2–4); but the speeches in Acts are often no more than précis.

Witherington, Acts, 46. Sometimes they help explain (e.g., Acts 7), but this is far less common than in even the Fourth Gospel (see Keener, John, 69–70). Perhaps fitting Luke’s popular audience, the speeches themselves include action, the narrative element predominating. Helping translate somewhat analogously for modern audiences, a dynamic equivalent video rendering of Acts could keep the action moving by including clips of the actions that the speeches narrate.

Soards, Speeches, 143. This is not to deny Luke’s interest in settings; Witherington, Acts, 46, contends that Luke gives more space to the speeches’ narrative settings than appears in Thucydides.

Soards, Speeches, 142, cites Stephen’s antitemple language and justification in Acts 13:39 (also, e.g., Munck, Acts, xliv); Plümacher, by contrast, cites Peter’s “Pauline” language in 15:7–11.

As one must expect if one allows even that the speeches are not verbatim transcripts but rendered in Luke’s own style, and in view of his likely sources having usually preserved, at most, a summary or random details of most speeches.

Even the reports in other histories do not represent full speeches, considering that most orations in the empire were limited to two hours (earlier speeches were not limited; Tac. Dial. 38; See comment on Acts 24:4). Counterspeeches do appear in Acts 19:25–27, 35–40; 24:2–21.

See Judge, “Sources,” 278–79, though observing on 280 the biblical origins of the valuing of authenticity that informs modern historiographic practice. Modern historians’ generalizing summaries offer a functional equivalent to the more detailed speeches of ancient historians (279).

Dibelius, Studies in Acts, 3, 138–85; idem, Tradition, 16; Luke could develop materials as a historian writing literature, but he was an evangelist as well as a historian (Studies in Acts, 185). Dibelius did recognize, like Dodd, that the sermons’ structure followed the ancient kerygma (Tradition, 16–18). Nock, Essays, 828–29, appreciated Dibelius’s recognition of the difference between the character of narrative and speeches in ancient historiography.

Cf., e.g., the critique of C. Williams, Acts, 37–38.

Cadbury, Making, 185.

Ibid., 192–93. Ancient historical speech-writing, though different from modern practice, is not bad; we simply must understand its function (see Laistner, Historians, 16; cf. ibid., 42).

So Walbank, Speeches, 19.

Cadbury, Making, 186–87.

With, e.g., Walbank, Speeches, 19; Woodman, Rhetoric, 13; Hadas, “Introduction,” xvi; Judge, First Christians, 379–80. Laistner, Historians, 129, notes that he retains the substance of Claudius’s speech and even some elements of his style, but the speech is largely in Tacitus’s own style, deletes material that Claudius borrowed from an earlier speech, rearranges material, and even adds some observations.

Tacitus includes many speeches for which we cannot expect exact words to have been preserved (e.g., Ann. 2.71–72, 76, 77; 6.48; 11.7; 12.48; 13.21; 16.22), including private conversation (4.7, 52, 54, 68–69; 12.65); indirect speech also appears at events where we cannot be certain that scribes would have kept records (e.g., 12.2; 14.53–54, 55–56; 15.51). Even public speeches to the emperor or senate (e.g., 2.33, 37, 38; 3.33–36, 58, 69; 4.34–35, 37–38; 12.6, 37; 14.43–44; 16.28, 31; perhaps 13.42) will not have been preserved verbatim (though the substance would, presumably, appear in official records). Some speeches are even summaries of what people more generally (rather than an attributed individual) were saying (indirect but at length, in 3.17, 40; 14.20; 15.59; 16.25; more briefly, 4.64; 16.4, 29). Sometimes, however, sources close to the event may report conversation we would otherwise disregard (15.63, which Tacitus regards as too well known to repeat; 15.67, for which Tacitus seems to have exact words though they were not so widely published; cf. possibly 13.20). Probably, Tacitus often reports the substance of his sources, even when we suspect that the claims could have arisen originally from gossip or rumor (e.g., 14.9, 59). Tacitus elaborates speeches where contemporary biographers omit them (Hist. 1.29–30) or merely mention them (Hist. 1.15–16 with Suet. Galba 18.3; Plut. Galba 23.2), but sometimes includes the same point as in their summary (Tac. Hist. 1.83–84 with Plut. Otho 3.8; Tac. Hist. 2.47–48 with Suet. Otho 10.2; Plut. Otho 15.3–17.2; cf. Tac. Hist. 1.37–38 with Suet. Otho 6.3).

Talbert, Mediterranean Milieu, 211, cites Plut. Otho 15 and Tac. Hist. 2.47; but note comment above. For agreements in the narrative cf. Keener, “Otho.”

Talbert, Mediterranean Milieu, 211, cites Caesar Gall. W. 1.40 vs. Dio Cass. 38.36–46.

Jos. Ant. 1.279–84 vs. 1 Macc 2:50–68 (Talbert, Mediterranean Milieu, 211).

Herod’s speech to soldiers in Jos. Ant. 15.127–46 vs. War 1.373–79 (Talbert, Mediterranean Milieu, 211).

For rhetorically trained historians, reproducing speech sources verbatim may have smacked of something of the odium we attach to blatant plagiarism (except that the issue was one’s reputation for rhetorical creativity, not one’s integrity in crediting sources).

Cadbury, Foakes-Jackson, and Lake, “Writing History,” 13–14, comparing Plut. Otho 15 with Tac. Hist. 2.47; Jos. War 1.373–79 with Ant. 15.126–46; Dio Cass. 38.36–46 with Caesar Gall. W. 1.40; and for omissions of speeches, Sall. Catil. 31; Tac. Ann. 15.63. Cf. Toynbee, Thought, 179–80; Whittaker, “Introduction,” lix. Plutarch, Livy, and Josephus are notoriously less careful with speeches; perhaps not all representatives would fare equally poorly.

E.g., Dio Chrys. Or. 15, between two highly literate Athenians. Dio uses dialogues in Or. 21, 23, 25, and 26 (though in Or. 25 Dio takes over speaking after initial interchanges); Or. 23 is in the form of a dialogue with one of his students (J. W. Cohoon in LCL, 2:301, suggests that Or. 23 might recall Dio’s genuine pedagogic interactions).

Although the “dialogue with Jews” form developed a long history (Hulen, “Dialogues”; Bokser, “Justin and Jews,” 98) and even Justin’s conversion account seems stylized (Wilken, “Social Interpretation of Apologetics,” 443; Osborn, Justin, 7; Kaye, Apology, 13–14; Chadwick, “Defence,” 280), it might reflect genuine dialogue. Thus, for example, Trypho does not convert (Justin Dial. 142; see Schneider, “Reflections,” 173, 175); some of his objections seem stronger than Justin’s own case (e.g., Dial. 27, 55–56); and some of Justin’s arguments (e.g., Dial. 46) may support a genuine debate in the mid-130s, though later elaborated with every subsequent argument he could find (see Barnard, Justin Martyr, 23–24, 39–40; Williams, Justin Martyr, xi–xix; cf. Osborn, Justin, 7; Kaye, Apology, 13–14; Chadwick, “Defence,” 280; perhaps drawing on debates with various Jewish disputants: cf. Dial. 50; Aune, “Use,” 184). Though we cannot isolate the source, Justin betrays some knowledge of postbiblical Jewish beliefs (Barnard, “Old Testament,” 401, 406; idem, Justin Martyr, 45–46; Williams, Justin Martyr, xxxii–xxxiii, 80n4; Shotwell, Exegesis, 80, 89; Schneider, “Reflections,” 168–69), though he also clearly misunderstands or misrepresents Judaism at points (see, e.g., Schneider, “Reflections,” 169; Higgins, “Belief,” 304–5; Bokser, “Justin and Jews,” 112). Traditions like the hidden Messiah (Higgins, “Belief,” 300; Williams, Justin Martyr, 18n2; Barnard, Justin Martyr, 46–47; Shotwell, Exegesis, 72) and apparent use of “Man” as a divine circumlocution (Dial. 59.1; Eccl. Rab. 2:21, §1; 8:1, §1; Pesiq. Rab. 14:10; Marmorstein, Names, 65–67) parallel rabbinic sources; by contrast, his dialogue offers some unsupportable claims, such as the one that the rabbis do not challenge belief in the Messiah’s deity (cf. Dial. 63; Higgins, “Belief,” 305; see b. Men. 110a; Sanh. 38ab; Pesiq. Rab 21:6).

In Brutus, Cicero employs real characters who knew him, and he may include some of their actual ideas; but this is hardly a verbatim presentation. (For a higher estimate of some of Cicero’s and others’ sources for dialogues, cf. Kennedy, “Source Criticism,” 136).

Poets, of course, felt free to compose words for their characters, even if their story renders it impossible for surviving human witnesses to have recounted the words (e.g., Prop. Eleg. 3.7.57–64). Though citing epic and dialogue, Hermog. Method 20.436 freely attributes oaths contained therein to the composers (Homer Od. 20.339; Plato Gorgias 489E). Xenophon countered Plato’s Symposium and Apology with more realistic versions (though Plato knew Socrates better and they share some common, probably historical, tradition, e.g., Xen. Symp. 8.12; for a helpful evaluation of historical tradition, see Marchant, “Introduction,” ix–xv, esp. x–xiv); Xenophon composes narrative better, but Plato dialogue.

This exercise apparently continued even into the Islamic period (Cook, Dogma, 51).

Brin, “Uses,” thinks that the author of the Temple Scroll (esp. col. XLVIII) sought to imply for it Mosaic authorship; cf. elsewhere, e.g., Abegg, “Pseudo-prophets.” Cf. Egyptian attribution of new works to Thoth (Frankfurter, Religion in Egypt, 240).

Aul. Gel. 3.3.

E.g., Polyb. 5.42.7; 5.50.11; Livy 40.55.1; Jos. Life 356; Pliny Ep. 6.22.3–4; 10.58.3; 10.65.2; Arrian Alex. 6.12.3 (though this work turned out to be genuine); Apul. Metam. 4.16; Philost. Vit. Apoll. 7.35; Iambl. V.P. 1.2; cf. Suet. Vesp. 6.4; for Roman laws regarding forgery, see Robinson, Criminal Law, 36–39 (cf. Quint. Inst. 5.5.1; sometimes addressing forged wills, e.g., Suet. Aug. 33.2; Claud. 9.2; 15.2); on ancient understanding of forgery, see, e.g., Judge, Athens, 75–76. A minority take this approach to documents believed pseudepigraphic in the NT (see Verhoef, “Pseudepigraphy”; idem, “Paulines”; idem, “Impact”); certainly, the patristic era rejected documents that were thought pseudepigraphic (Carson, “Pseudonymity,” 860–63), though such works (e.g., the Cynic Epistles) were known among contemporaries. Cf. the distinctions between forgeries and propagandistic pseudepigraphy in Görgemanns, “Epistolography,” 1146. Common though pseudepigraphy was, it was often viewed negatively (see Aune, Dictionary of Rhetoric, 388; Ehrman, Forged, views pseudepigraphy even more negatively).

Maclean and Aitken, Heroikos, xlix. Of less relevance here, some created fictitious attributions for speeches to give them authority while protecting the real authors from harsh retribution (Tac. Ann. 5.4).

Supposedly, deserters reported the substance to Cyrus (Xen. Cyr. 3.3.48), but this is a plausibility measure in a narrative that integrates fiction with history in a way rarely followed by successors. Although Xenophon may have woven in earlier anecdotes, he reports in direct speech the lengthy boyhood conversations of Cyrus (e.g., 1.3.2–18). For skepticism about generals’ speeches, see Marincola, “Speeches,” 128.

Though this does relate to the following narrative (Xen. Cyr. 5.1.18), it clearly ends with the wisdom of the more reliable character (5.1.12). As noted in , Xenophon’s Cyropedia is probably largely fictitious.

Penner, Praise, 208–11, compares with historians’ speech composition the practice in rhetorical handbooks and the tragic conventions in Aristotle’s Poetics.

E.g., Eckey, Apostelgeschichte, 32–34. Witherington, “Addendum,” 24, 28–29, argues that this practice stemmed from particular historians and was never a “convention.” He may define “convention” more narrowly than most do.

With, e.g., Horsley, “Speeches,” 609; Ehrman, Introduction, 133–34. Some disciples sought to preserve as much as possible of the style and content of their teacher (Epict. Diatr. 1.pref., perhaps in contrast to Plato), but this would be more difficult for individual speeches.

Pliny E. N.H. pref. 12 complains that his Natural History is not a highly literary work, for it lacks “digressions, . . .  speeches or dialogues” as well as striking incidents “interesting to relate or entertaining to read” (LCL, 1:9). For its readers, however, see, e.g., Symm. Ep. 1.24.

E.g., Dibelius, Studies in Acts, 138–85; Plümacher, “Luke as Historian,” 398–99; Johnson, “Luke-Acts,” 409; Talbert, Acts, 28–30.

Lüdemann, Acts, 24, avers that, at the very least, Luke freely composed a key speech of Jesus (Luke 4:18–27) from Mark 6:2. Historians could, in fact, expand on such a nugget as Mark 6:2, but we cannot prove that Luke did so in this case or regularly. First, our lack of a direct source for the speech in Luke 4 does not mean that Luke must have lacked such a source. But second and, in my opinion, more important, Luke’s inaugural scene is a special one; this is one of only two significant places where he diverges from Mark’s chronology, and it does not reflect a typical habit of his.

Lüdemann, Acts, 24.

He cites only Thucyd. 1.22.1 and Fornara, Nature of History, 154–55.

They might well depend for this on the known claims of earlier historians (whose own sources might ultimately be accurate or themselves freely composed). Laistner, Historians, 17, thinks that the basic picture the speeches conveyed was true, even if the words were invented.

Kennedy, Classical Rhetoric, 110, referring to Hdt. 3.80–83. In contrast to some later historians, Herodotus seems to have used an epic model (Pohlenz, according to C. Williams, Acts, 36–37).

Aune, Dictionary of Rhetoric, 448.

Quint. Curt. 4.14.9–26 (a prebattle speech of Darius); 5.8.6–17; note also the advice speech of one of Darius’s officials, 5.9.3–8; note esp. 3.2.11–16, where the speaker (3.2.17) and primary auditor die soon afterward. Naturally, Quintus Curtius also supplies speeches for the Greeks (4.14.1–6; 6.3.1–18; 6.8.15–6.9.24), including an impromptu defense speech (6.10.1–37).

Whittaker, “Introduction,” lix (noting the requirement for plausibility in Thucyd. 1.22 and appropriateness in Lucian Hist. 58).

Aune, Environment, 82.

Ibid.

Judge, “Rhetoric of Inscriptions,” 819.

Cf. also Tac. Ann. 15.63, where he plainly says that he does not need to reproduce Seneca’s words exactly (Talbert, Mediterranean Milieu, 210). Laistner, Historians, 128, contends that as Tacitus matured his speeches grew more concise, with greater “individual characterization and more contrast between the speakers.”

Compare Arrian’s summary of his teacher Epictetus’s teachings in Encheiridion with the more rambling Diatribai, which preserves even Epictetus’s style through Arrian’s more detailed notes.

Gempf, “Speaking,” 284. A core could prove, not surprisingly, much shorter than rhetorical historians’ (as opposed to Luke’s) elaboration of speeches; even in recent history, extant notes for a genuine, three-and-a-half-hour speech may be read aloud in three minutes (Tomkins, Wilberforce, 80, 84).

Cf. Tacitus’s qualifications regarding Galba’s speech: “according to the report,” he “spoke to this effect” (Hist. 1.15 [LCL, 1:27]); he “spoke further to the same effect” (1.16 [1: 33]); and “people report” the soldiers’ response (1.17). On historians sometimes explicitly disavowing their reports being verbatim, see further Marincola, “Speeches,” 120, citing, e.g., Polyb. 18.11; Sall. Cat. 50.5; Livy 37.45.11; Arrian Anab. 5.27.1; Tac. Hist. 1.15–16; Agr. 29.4, and noting that emphasis on verbatim quotes was rare and dealt with short lines (e.g., Tac. Ann. 14.59.4; 15.67.4). The only historian who offered verbatim speeches was Cato, and the speeches were his own (Woodman, Rhetoric, 13).

Balch, “Ἀκριβῶς,” 229–39.

Ibid., 244.

Aune, Environment, 93, and sources he cites. Idem, Dictionary of Rhetoric, 448, seems to exempt only Thucydides and Polybius from sourceless speech composition (cf. Plümacher, “Luke as Historian,” 398–99, perhaps not even exempting Thucydides), but (given our limited data) this verdict seems too narrow; for Arrian, see Hammond, “Speeches,” and for even Livy, see discussion below.

As often noted, e.g., deSilva, Introduction, 351.

Thus Dion. Hal. Thuc. 41 would examine speeches only for appropriateness when he could infer from the narrative that Thucydides was neither present nor dependent on those who were (see Damon, “Rhetoric,” 446).

Aune, Environment, 125.

See Small, “Memory.” In ancient schools, see, e.g., Alexander, “Memory,” 143.

So Bauckham, Eyewitnesses, 333–34 (citing memory studies), noting also variation in detail among the different Gospels, which nevertheless preserve the essential substance of events (such as Peter’s denials, 344–45). See also Allison, Jesus, 11–13; Eddy and Boyd, Legend, 275–85, and the many memory studies they cite.

C. Williams, Acts, 37, contrasting this common and mediating view with the extremes; cf. Witherington, “Addendum,” 24–25.

See Zimmern, Commonwealth, 199 (as cited in Pelikan, Acts, 255).

Kennedy, “Survey of Rhetoric,” 15. Marguerat, Histoire, 32–34, contends that Thucyd. 1.22.1 requires verisimilitude and appropriateness.

Thucydides probably did not use the speeches as an opportunity to show off his rhetorical skill; see the critique of this defect by the rhetorical historian Dionysius (Dion. Hal. Pomp. 3, cited by Fitzmyer, Acts, 105; Witherington, Acts, 48). Dionysius appears more interested in Thucydides’s rhetorical inappropriateness than in his fictionalizing speeches (Marguerat, Histoire, 34).

A view noted in Walbank, Speeches, 3, as accepted by many German scholars in the mid-twentieth century. Woodman, Rhetoric, 11–12, argues (probably rightly) that Thucydides claims to recount only the gist, but on 12–13 argues (based esp. on memory studies, addressed below) that even the gist may have often been unavailable; he questions the degree of Thucydides’s reliability on 9–47 (e.g., questioning his personal experience for the geographic digressions [15–16] and his claims to exclude myth and legend [16, 23]).

See Walbank, Speeches, 4. For this balance between fidelity and invention, see also Marincola, “Speeches,” 121–22, noting further that (122) “his inability on occasion to learn the actual words spoken will have necessitated a certain amount of imaginative reconstruction.” Porciani, “Enigma,” 333–34, allows for an interpretive framework but argues against mere creation of passages based on imagination.

False information came into many ancient histories through their dependence on funeral eulogies, which often included falsified information (Woodman, Rhetoric, 91, citing Cic. Brut. 62; Livy 8.40.4–5).

See Porter, Paul in Acts, 110–12; idem, “Thucydidean View?”

Lest anyone question the use of historical imagination: even in Thucydides, the speeches are long and fleshed out. Woodman, Rhetoric, 13, suggests that Thucydides formed long speeches to rival Homer.

Walbank, Speeches, 4. Yet Marincola, “Speeches,” 123–27, shows that Thucydides’s model remained an ideal; Croke, “Historiography,” 567–68, shows that he was most read and probably the prime model.

Fornara, Nature of History, 167.

Horsley, “Speeches,” 609.

Fornara, Nature of History, 167–68.

Ibid., 168. Even Roman reports of speeches (less relevant to a discussion of Luke-Acts) apparently are “substantially trustworthy from the time of the Second Punic war to the end of the fourth century A.D.” (168). Of course, even where writers may have adapted a speech from earlier sources, the earliest source in the chain could have composed the speech more freely; subsequent preservation and the ideal of historical fidelity do not by themselves guarantee historical veracity in the modern sense. I am thus less optimistic than Fornara, though I acknowledge that the gist may have been preserved in many cases.

Ibid., 142. Many modern readers, indeed, would prefer at most a summary of the speech’s thrust and effect to its plausible reconstruction, just as some of my students, accustomed to fast-action television, prove bored with the long speech of Stephen in Acts 7:2–53.

Cadbury, Making, 186–87; Hanson, Acts, 36; Gempf, “Speaking,” 264, 299. See especially the balance in Marincola, “Speeches,” 121: “in almost every historiographical speech there will have been a mixture of what was actually known and what could be surmised.”

Cf. Gempf, “Speaking,” 264: “Just as a writer was expected to represent faithfully the strategies, tactics and results of a battle . . .  so a writer was expected to represent faithfully the strategies, tactics and results of a speech, without necessarily recording the exact words used on the day.”

That is, from our perspective and their own; their detractors, naturally, would have differed in such estimates.

Gempf, “Speaking”; cf. also his dissertation (idem, “Appropriateness”); also Witherington, “Addendum,” 24, 28–29; idem, Acts, 48, 117, 455; Hemer, Acts in History, 75–78.

Gempf, “Speaking,” 272. Polybius may exaggerate Timaeus’s failure here; cf. Marincola, “Speeches,” 125.

E.g., Polyb. 3.63.3–13; 3.64.5–8; 3.108.6–109.12; 3.111.3–10. For other speeches, see, e.g., 9.28.1–31.6; 9.33.2–39.7; 11.4.1–7.8; 11.28.3–29.13. Walbank, Speeches, 12, argues that Polybius does have military speeches in his sources; although this may well be the case, Polybius undoubtedly develops them. Other historians also use speeches of opposing leaders, not likely preserved in significant detail (e.g., Tac. Ann. 14.35; 15.1, 2).

That writers could choose to omit some passages from speeches (Philost. Vit. soph. 2.29.621) assumes that they did have copies of the speeches (by this period, often published).

Fornara, Nature of History, 143–54. Cf. Gempf, “Speaking,” 264: “Woe betide the historian if the speech is not faithful to the alleged situation and speaker.”

Esp. Hammond, “Speeches.”

Fornara, Nature of History, 154–68.

Ibid., 160–61; also Witherington, Acts, 40. For Livy’s selection and adaptation of earlier sources, see Luce, Livy; Forsythe, Livy (cited in Forsythe, “Quadrigarius,” 396). For Livy’s composition of speeches to provide characterization for the speakers, see Laistner, Historians, 96. Though unwilling to extrapolate to other speeches, Laistner, Historians, 97, notes that Livy is less rhetorically attentive when reproducing Polybius’s speeches closely.

Gempf, “Speaking,” 283 (based on comparison of such speech reports as Polyb. 3.62ff.; 11.28ff.; 15.6.4ff.; 21.19ff. with Livy 21.42–43; 28.27ff.; 30.30; 37.53–54; noting Walsh, Livy, 231). Historians would adapt their predecessors’ speeches when they reported the same speech occasions (rather than composing speeches for different occasions, as they also did), but they felt bound to their basic content (Marincola, “Speeches,” 129, giving as examples Livy’s reworking of Polybius and Tacitus’s of Claudius).

Fornara, Nature of History, 161. Regarding excesses, Laistner, Historians, 16, notes that modern historiography also has pockets of excess (citing “economic determinists”).

Fornara, Nature of History, 166–67.

In Dio Chrys. Or. 13.14–15, Dio refers to a speech of Socrates. This may be spurious; but it is earlier than Dio, and he believes it to be such a speech. In 18.18–19, he refers to written speeches in Xenophon.

Cf. the examples (e.g., Quint. Curt. 9.6.13–15 with Arrian Alex. 6.13.4–5) in Quintus Curtius (who invents most of his speech material) in Baynham, “Reception,” 297–99.

Fitzmyer, Acts, 105.

See, e.g., Dudley, “Speeches,” who also ranks Luke closer to Thucydides than to Livy or Josephus.

Talbert, Mediterranean Milieu, 210–11.

Ibid., 211–12; cf. Marshall, Acts, 41. Talbert, Mediterranean Milieu, 211, also notes Polybius’s complaint about Timaeus (Polyb. 12.24–25).

E.g., Aul. Gel. 1.4; Cic. Or. Brut. 46.156–57; Fam. 13.27.1; Max. Tyre 21.4; cf. Cic. Brut. 91.316; Anderson, Glossary, 53–54, 114; Trapp, Maximus, 182n9; Nock, “Vocabulary,” 137; Lee, “Translations of OT,” 776–77; Keener, John, 324–25. One should merely take care to avoid “improper” synonyms (Rowe, “Style,” 123–24); for ancient discussion of synonyms, see, e.g., Porph. Ar. Cat. 68.5–27. In some writers, a more consistent sense obtained, but this was unusual (Aul. Gel. 2.5.1).

That Luke uses speeches as theological commentary on his narrative is likely; some scholars have also suggested that the narratives illustrate the themes articulated in the speeches (Hinkle, “Preaching,” 100).

With T. Smith, “Sources,” 75.

See , ; also, e.g., Keener, Historical Jesus, 102–5 and notes 95–150 on 445–48; Fitzmyer, Acts, 105.

As noted by Bruce, Acts³, 35; cf. Aune, Dictionary of Rhetoric, 326. There is also debate as to whether Luke expands Jesus’s inaugural speech (we cannot discount the possibility that Luke possessed sources in addition to Mark 6:1–4; see Luke 1:1), but this case may be a special one, since it is Luke’s programmatic scene and one of the very rare occasions where he diverges from Mark’s sequence.

Cf. Marshall, Acts, 41, noting that, in general, the evidence for tradition is “insufficient to demonstrate that all the speeches were actually delivered on the occasions specified—a point which probably lies beyond historical proof in any case.” Hengel, Peter, 85, suggests that, given Peter’s prominence in the gospel tradition and even more as leader of the early apostolic movement, Luke could have had genuine Petrine tradition available for his Petrine speeches.

With, e.g., Witherington, Acts, 119–20. In epic genres, one could invoke inspiration to make up for one’s lack of knowledge (Keener, John, 116; cf. Philost. Hrk. 43.6; Nagy, “Prologue,” xxx–xxxiv; Maclean and Aitken, Heroikos, xliii, xlix-l); not all expected such inspiration to prevent all error or bias (cf. Philost. Hrk. 24.1–2; 25.2–17; also Ramelli, Hierocles, 98, on Hierocles Gods criticizing Hom. Il. 9.497 in Stob. Ecl. 1.3.53; less relevantly, cf. Lucian True Story 2.32; Libanius Refutation 1.1–3, 12; 2.1; Invective 7.2), but it certainly was often thought to help (cf., e.g., Libanius Confirmation 2.1–2). (Gentiles did approach Homer differently than the way that Jewish people approached the Torah; see Rives, Religion, 28.) But while Luke might have allowed biblical prophet-historians this possibility (cf., e.g., 2 Kgs 6:12), it is not part of the Polybian or Thucydidean historiographic tradition; then again, Josephus’s view of his prophetic ability shows that it need not have been incompatible with Jewish historiographies.

C. Williams, Acts, 37–38. For a similarly nuanced perspective, written long before Dibelius, on ancient historians regarding both gist and attempted accuracy, see Abbott, Acts, 15, from 1876.

E.g., Knox, Acts, 75; Zehnle, Pentecost Discourse, 56–60; Robinson, “Primitive Christology.”

Some have suggested Luke’s skill in this exercise (Kennedy, New Testament Interpretation, 114–40; Burridge, “Gospels and Acts,” 520; Aune, Environment, 125). It also appears in diatribe (Stowers, Diatribe, 90–91; Tobin, Rhetoric in Contexts, 93, citing Epict. Diatr. 1.4.28–29; 1.9.12–16; 1.26.5–7, 10, 20–22, 26–49; 3.24.68–70, 97–102); see further Anderson, Glossary, 106–7; Rowe, “Style,” 144; Tobin, Rhetoric in Contexts, 227 (citing, e.g., Theon Progymn. 2.117.6–32; Quint. Inst. 4.1.28; 6.1.25–27; 9.2.29–30). (It is probably implied also in Lucian Dance 65.) At a more advanced level, declamations integrated speeches-in-character exercises with other basic rhetorical exercises (Bloomer, “Declamation,” 299).

Many rhetoricians technically limited “prosopopoeia” to personified things, using “ethopoeia” for real persons (e.g., Hermog. Progymn. 9, “On Ethopoeia,” 20; Aphth. Progymn. 11, “On Ethopoeia,” 44–46S, 34–36R, esp. 44–45S, 34R; Nicolaus Progymn. 10, “On Ethopoeia,” 64–65). Cf. also the application of “prosopopoeia” to nonhuman things and sermocinatio to humans in Rowe, “Style,” 144 (citing, for the former, Demosth. Fals. leg. 119; Cic. Cat. 1.7.18 and, for the latter, Demosth. Fals. leg. 235; Aug. Ep. 108.17); for sermocinatio, see also Porter, “Paul and Letters,” 582 (citing 1 Cor 1:12); see further Anderson, Glossary, 33 (citing Dion. Hal. Thuc. 37; Rhet. Her. 4.65; Quint. Inst. 9.2.31); an example in Tzounakas, “Peroration.” Ethopoeia was a form of what was more generally called prosopopoeia (Anderson, Glossary, 60–61).

E.g., Demet. Style 5.265 (he cites [5.266] an example in Plato).

Aune, Dictionary of Rhetoric, 383, citing Quint. Inst. 9.2.37. Naturally, this would always be the case in historical narrative.

Cf. Aune, Environment, 93, 125; Johnson, Acts, 53; Penner, Praise, 211–12.

Johnson, Acts, 53, citing Lucian Hist. 58 (also cited in Talbert, Mediterranean Milieu, 211).

Hengel, Acts and History, 61. Some rhetorical exercises included “impersonation” (Lucian Dance 65).

See Butts, Progymnasmata of Theon, 459n; cf. Maclean and Aitken, Heroikos, xlix; for its importance in later antiquity, see Gibson, “Notes,” xxi. Greek students also learned to imitate the style of famous authors (Oldfather, “Introduction to Epictetus,” xiv; cf. Aune, Environment, 31).

Later in Aphth. Progymn. 11, “On Ethopoeia,” 45–46S, 35–36R; Libanius Speech in Character 8. Niobe is a mythical example; both mythical and historical examples were used (Gibson, “Notes,” 355).

Lucian Phal. passim (though most rhetorical exercises of writing in character were not tongue in cheek, as noted by A. M. Harmon in Lucian, LCL, 1:1).

Hermog. Progymn. 9, “On Ethopoeia,” 21, associating ἦθος (ēthos) with the speaker and πάθος (pathos) with the emotional situation involved. Some characterizations focused on emotion and others on character (Aphth. Progymn. 11, “On Ethopoeia,” 44–45S, 35R; Nicolaus Progymn. 10, “On Ethopoeia,” 64).

Dio notes that Diogenes was an exile (Dio Chrys. Or. 4.1), and compares himself with him (4.3). Many of Dio’s speeches involve dialogues of Diogenes (Or. 4; 6 [partly describing, partly quoting]; 8; 9; 10) or Alexander (with Philip, Or. 2; with his superior Diogenes, Or. 4); cf. also “Pheidias’s” speech about graven images (taking up much of Or. 12).

See, e.g., Alciphron Letters of Fishermen; Letters of Courtesans; Aelian Letters of Farmers; cf. Maclean and Aitken, Heroikos, xlix.

Tobin, Rhetoric in Contexts, 93 (citing Epict. Diatr. 1.4.28–29; 1.9.12–16; 1.26.5–7, 10, 20–22, 26–49; 3.24.68–70, 97–102).

Cicero is rhetorically superior to Plato on this point, presupposing a more careful study of various alternative systems of thought.

An example in Longin. Subl. 3.2 is Gorgias of Leontini calling Xerxes the “Persian Zeus.”

See Black, “Oration at Olivet,” 88 (who cites esp. Arist. Rhet. 3.2, 1404b–1450b; Cic. Or. Brut. 75–121; De Or. 3.9.37–39; 3.52.199; Quint. Inst. 1.5.1; 8.1–11). A rhetorical speech in the mouth of a non-rhetorician’s character could even denounce rhetoricians (Quint. Decl. 268.16–20).

Thus rhetoricians criticized dramatists who used bombast in character’s speeches, because it failed to resemble genuine speech (Longin. Subl. 3.1–2). One might praise a rhetor who used a more diverse array of arguments, ideas, and presentation, though many failed in this (Dion. Hal. Lysias 17; Isaeus 3).

Soards, Speeches, 140, following Plümacher, Lukas, 72–78. Romans also used deliberate archaisms “to impart solemnity” (Holford-Strevens, “Archaism”).

Likewise, although Luke’s Gospel often calls Jesus “Lord,” raising the suspicions of some (e.g., Sandmel, Judaism, 369), until the resurrection, the vast majority of nonvocative uses of this title by mortals are from the narrator himself, not participants in the story (Moule, “Christology of Acts,” 160).

The parallels probably stem not from imitating competing writers but from the historical Socrates (see Marchant, “Introduction,” ix–xv).

Jervell, Theology, 9.

For a treatment of Johannine speeches, see Keener, John, 53–80; these might resemble the use of speeches in ancient historians (Keener, John, 68–76; Bauckham, “Historiographical Characteristics”).

Such occasions could potentially include some of Luke’s speech settings, such as the Olivet discourse, or Pentecost or the healing in Acts 3. For Luke 6, see Allison, Jesus, 309–51.

On this last point, see Crosby, “Introduction.”

E.g., Cic. Brut. 24.91 (remarking on the laziness of those who fail to write them out); Pliny Ep. 2.5.1–2; 5.8.6; 5.20.3 (noting that the reception of written speeches was less subject to chance); 7.17.5.

So Pliny Ep. 1.20.7–8 (countering a contrary view mentioned in 1.20.6).

Pliny Ep. 3.18.1 (referring to his Panegyricus).

E.g., Pliny Ep. 3.13.1; 4.5.3.

E.g., Pliny Ep. 4.26.1–3; Pliny expects to hear from another when that person sees Pliny’s published version (4.5.3).

Rhet. Alex. 36, 1444a.16–28; Sen. E. Controv. 1.pref. 17. Probably also implied in Eunapius Lives 494. Notes of declamations could include instructions to provide vivid descriptions of events (Sen. E. Suas. 3.2).

As we have noted, even in recent history, extant notes for a genuine, three-and-a-half-hour speech may be read aloud in three minutes (Tomkins, Wilberforce, 80, 84).

Winter, “Official Proceedings,” 307; idem, “Captatio benevolentiae,” 526–28.

See Rüpke, “Commentarii,” 628. Associations, like cities, could have secretaries keeping minutes of business meetings.

Even had verbatim recording been possible, elite audiences would have dismissed such reports as rhetorically inept, because they expected historians to be more creative, just as we expect writers to compose in their own words and not plagiarize their sources.

Maehler, “Tachygraphy,” 1468. Richards, Letter Writing, 67–68, points out that Latin shorthand precedes Paul by at least a century (Plut. Cat. Min. 23.3–5; Euseb. Chron. 156); Richards’s argument (68–69) that the use of Greek terms in Cicero, the alleged Latin originator, suggests an earlier Greek practice, is less certain (since Plutarch, who preserves the tradition, writes in Greek).

Richards, Letter Writing, 65–72. That shorthand appears even in the desert in P.Murabbaʿât 164 suggests that it was widespread by the time of the Second Judean Revolt (pp. 71–72).

Richards, Letter Writing, 73 (who concludes that it is unlikely that Paul had access to shorthand secretaries, 73–74).

From a modern historical standpoint, we should also be clear that even firsthand notes may contain mistakes at times. Anyone who has been part of a group reviewing minutes from a previous meeting will recognize that the group’s memory may sometimes correct notes from a month or season before.

Gempf, “Speaking,” 299, citing Haenchen, Acts, 590; and Quint. Inst. 11.2.2, 25.

Lutz, “Musonius,” 7, 10.

Sen. Y. Ep. Lucil. 108.6. Cf. evidence in Millard, Reading and Writing, esp. 199–208 (in Jewish circles, see 204–8). For personal notebooks as diaries, see Rüpke, “Ephemeris.”

See Gamble, “Literacy,” 646; cf. Millard, Reading and Writing, esp. 184–209. Although some have argued against overplaying the orality/literacy divide (Malina, “Criticism,” 98, as cited in Botha, “Cognition,” 37), some others warn against playing it down too much (Botha, “Cognition”).

Cf. Sen. Y. Ep. Lucil. 108.6; Arius Did. Epit. 2.7.11k, pp. 80.36–82.1; Lucian Hermot. 2; see also Kennedy, “Source Criticism,” 129–31 (with favorable evaluations in Meeks, “Hypomnēmata,” 159–60; Fuller, “Classics,” 178–79); Stowers, “Diatribe,” 74, on Arrian’s notes on Epictetus; Lutz, “Musonius,” 7, 10, on notes from Musonius’s pupils; cf. Iambl. V.P. 23.104; Hippol. Ref. 1.15. Cf. the brief discussion of Plutarch’s notebooks in Cherniss, “Introduction,” 398–99. Notes could derive from other sources as well; an intellectual might in several days fill five notebooks with extracts from sixty “books” (including brief speeches; Fronto Ad M. Caes. 2.10.2). I am reusing and augmenting here some material from Keener, John, 55–56.

Kennedy, Classical Rhetoric, 19.

Oldfather, “Introduction to Epictetus,” xii–xiii. Even in the Encheiridion, where Arrian organizes and summarizes his master’s teaching, Epictetus’s character dominates.

Epict. Diatr. 1.pref. (LCL, 1:4–5). He might employ γράφω in contrast to συγγράφω, “compose” (the latter perhaps recalling the treatment of Socrates in Plato and Xenophon [LCL, 1:4–5n1]); one may be tempted to compare Luke 1:3, but Luke uses γράφω in the sense of “write to” rather than “write down.”

Quint. Inst. 1.pref. 7–8 (LCL, 1:8–9). Other teachers also had problems with people pirating their books and publishing them before they could nuance them properly (Diod. Sic. 40.8.1). When recounting such exploits to my students, I often add facetiously, “Don’t do that to me!”

Gerhardsson, Memory, 160–62; cf. Safrai, “Education,” 966.

See Millard, “Literacy”; Head, “Note on Reading.”

Gempf, “Speaking,” 299, citing esp. Quint. Inst. 11.2.2.

Zeno in Diog. Laert. 7.1.20.

Although we cannot prove that Luke planned to write a historical work when he traveled with Paul, the detailed travel itinerary strongly suggests that the narrator did take some notes.

Cf. Montanari, “Hypomnema,” on hypomnemata as notes or memory aids (though after the Hellenistic period they often represent a running commentary on texts, 72); Kennedy, “Source Criticism,” 136–37.

Winter, “Official Proceedings,” 307n7. Forensic speeches typically employed a plain, unadorned style compared to the grand style of much epideictic (Fronto Ad M. Caes. 3.16.1); yet they could be impassioned (so, e.g., Pliny Ep. 7.9.7), as opposed to the more reasoned style of deliberative rhetoric.

Winter, “Captatio benevolentiae,” 506.

For the coexistence of literacy and orality, see the example in Aune, Dictionary of Rhetoric, 325: “Greek schoolchildren who used to memorize from written texts of the poets in order to recite them without the written text” (Plato Prot. 325E). Byrskog, History, 111–12, stresses Plato’s preference for memory over texts.

Byrskog, History, 110–11; cf. examples in Allison, Jesus, 29, 374. Oral history approaches memory more optimistically than the dominant contemporary emphasis in psychology and sociology (Eddy and Boyd, Legend, 280). For one discussion of the issues in psychological memory studies, see Bauckham, Eyewitnesses, 325–41. For one useful survey of orality studies with respect to the Gospels, see Iverson, “Orality.”

Lewis, History, 43; Vansina, “Afterthoughts,” 110; cf. Yamauchi, “Homer,” on very general historical events behind Homer; also Redman, “Eyewitnesses,” 191, though specifying (as we normally should also for Greco-Roman antiquity) content rather than verbatim wording.

Lord, Singer, 138.

Redman, “Eyewitnesses,” 191–92.

See Harvey, Listening, 41; the Hindu child in Noll and Nystrom, Clouds, 129. On the retention of “gist” in eyewitness memory even when details are inaccurate, see Bauckham, Eyewitnesses, 333–34.

Dunn, Perspective, 112 (with 110), 122; cf. Kirk, “Memory,” 170–72. Most Roman literature in this period was orally performed, affecting its shaping (Schmidt, “Public recital”); nevertheless, our knowledge of how gospel tradition was orally performed in the eastern Mediterranean remains quite limited (cf. Gerhardsson, “Performance Criticism”). On the frequent lack of verbatim recall, and often re-creation, see, e.g., Moniot, “Historiography,” 56–57; performances are adapted to new audiences (Bazin, “Past,” 70–71). The variation is typically greater in prose than in verse (Kennedy, “Source Criticism,” 143).

Cf., e.g., Hoeree and Hoogbergen, “History”; Aron-Schnapper and Hanet, “Archives”; on rote memorization in traditional Quranic education, cf. Wagner and Lotfi, “Learning.” Limitations do, however, exist, especially over time (e.g., Iglesias, “Reflexoes”; Harms, “Tradition”; Raphael, “Travail”). For a recent and well-documented case for the potential reliability of oral transmission of gospel materials in light of more general knowledge about oral transmission, see Eddy and Boyd, Legend, 239–68 (esp. 252–59).

Used in academic African historiography since the 1950s (Moniot, “Historiography,” 50), needed to reconstruct practices before modern influences (Horton, “Possession,” 14). Westerners’ prior neglect often reflects ethnocentric prejudices (Chrétien, “Exchange,” 77).

With, e.g., Eddy and Boyd, Legend, 395 (on events no older than 80–150 years), citing various sources.

Henige, “History,” 103, rightly warns of the fanciful excesses caused by using theory to interpret evidence in reconstructing history behind oral tradition. Eddy and Boyd, Legend, 260–64, suggest that even folklore traditions about historical characters may often preserve more of the substance than Western critics have often assumed, but their focus is more recent history.

Ancient theories on how memory worked varied (see Arist. Concerning Memory and Recollection; Plato Meno 81CD; Philost. Vit. soph. 1.22.523); Cicero expounded on mnemonics (De or. 2.351; Olbricht, “Delivery and Memory,” 163). Philo (Boccaccini, Judaism, 192–94) and Ps.-Aristeas (Boccaccini, Judaism, 194–98) stress memory, blending Greek language with Jewish memorial traditions concerning God’s historic acts.

Cf. West, “Rhapsodes.” The frequency of literary fragments scattered throughout the papyri (cf. Avi-Yonah, Hellenism, 248) attests to the commonness of their usage.

Although we cannot confirm this report, the culture that Dio reports apparently did exist (cf. CIG 2.2077 in J. W. Cohoon, LCL, 3:447n2).

See Finkelberg, “Cypria” (doubting that texts always develop from fluid to rigid); cf. Yamauchi, “Homer.”

E.g., Paus. 1.23.2; cf. also Maclean and Aitken, Heroikos, xc–xci. Cf., e.g., Xen. Cyr. 1.2.1 for an example of long informal traditioning by storytelling and song. Scholars often have misgivings about Pausanias’s oldest oral information (cf. Pretzler, “Pausanias and Tradition”).

See Charlesworth, Pseudepigrapha and New Testament, 1–3; cf. Bailey, “Tradition.” For stories passed on in Mediterranean households, see Pizzuto-Pomaco, “Shame,” 38.

Eunapius Lives 453 (writing it down fixed it and prevented further changes). Even first-century writers recognized that centuries of oral transmission could produce variations in ancient documents (Jos. Ag. Ap. 1.12).

For example, other factors being equal, we would normally prefer Velleius Paterculus’s contemporary account to that of Tacitus, writing roughly a century later (Woodman, Rhetoric, 204).

Redman, “Eyewitnesses,” 192–93 (cf. also the concession on 179).

Moniot, “Historiography,” 56–57, notes that “formal instruction” in traditions is rare today.

See, e.g., Quint. Inst. 1.3.1; Plut. Educ. 13, Mor. 9E; Mus. Ruf. frg. 51, p. 144.3–7; Diog. Laert. 6.2.31; Eunapius Lives 481; Watson, “Education,” 310, 312; Heath, Hermogenes, 11; further Carr, Writing (esp. 111–73).

Mus. Ruf. frg. 51, p. 144.3–7 (though it is misattributed or, more likely, Musonius recycled an earlier saying of Cato [144.10–19]).

Theon Progymn. 2.5–8.

On their learning, see Quint. Inst. 2.4.15; Jeffers, World, 256; memorizing poets in Aune, Dictionary of Rhetoric, 143.

Jeffers, World, 256. Historians frequently quoted other historians from memory (Marincola, “Introduction,” 2; cf. Laistner, Historians, 77), a practice that, in biographic form, might also help explain some variants in the gospel tradition.

Jeffers, World, 256.

Antisthenes in Diog. Laert. 6.1.5 (LCL).

See, e.g., Val. Max. 8.15.ext. 1; Sen. Y. Ep. Lucil. 108.17, 20, 22; 110.14, 20; Mus. Ruf. 1, 36.6–7; Philost. Vit. Apoll. 7.22. Especially among rabbis, correcting a teacher was rare (ʾAbot R. Nat. 1 A); against striving with one’s teacher, see Num. Rab. 18:20; against offering legal decisions in front of one’s teacher, b. Šeb. 6:1, §8 (it merited death, Sipra Sh. M. d. 99.5.6; b. ʿErub. 63a; Tem. 16a; Lev. Rab. 20:6–7).

E.g., ʾAbot R. Nat. 25 A; y. Bik. 3:3, §9; Gen. Rab. 22:2; for rank for esteemed Jewish sages, see, e.g., b. Moʾed Qaṭ. 2:17; Sanh. 7:8; b. Hor. 13b, bar.; y. Sanh. 1:2, §13; Taʿan. 4:2, §§8–9. If a pupil needed to point something out to the teacher, even this should be done respectfully (e.g., Fronto Ad verum imp. 2.3); a pupil who contradicted a teacher might need to be reconciled to him (Philost. Vit. Apoll. 5.38). Later, one who despised a rabbi despised one’s help (b. Šabb. 119b) or risked losing eternal life (Pesiq. Rab. 11:2); by stinginess toward rabbis (b. B. Bat. 75a) or speaking ill of them after their death (b. Ber. 19a), one might inherit hellfire; offending rabbis might yield deadly judgment (cf. b. B. Qam. 117a; b. Meṣ. 84a; ʿErub. 63a; Tem. 16a).

Diod. Sic. 10.5.1; Iambl. V.P. 29.165; on their memories, see further 20.94; 29.164; 35.256. On memorization techniques, cf. Rhet. Her. 3.22.35; see further Kennedy, Classical Rhetoric, 98.

Iambl. V.P. 31.188.

Cf., e.g., Thompson, Wenger, and Bartling, “Recall,” 210 (noting that it becomes unnecessary beyond a certain point; this source was supplied to me by M. Bradley, then a student at Duke).

Iambl. V.P. 35.256; Philost. Vit. Apoll. 3.16. Certainly, their handing teachings down secretly only by memory (Iambl. V.P. 32.226) is portrayed as unusual.

Culpepper, School, 193; Alexander, “Memory,” 141; Aul. Gel. 7.10.1; Socrates Ep. 20. Some schools emphasized memorizing texts, others, memorizing the teacher’s words (Culpepper, School, 177).

Diog. Laert. 10.1.12, on Epicurus; on followers of Pythagoras, cf. Culpepper, School, 50. All schools emphasized memory, though Epicurus more than Plato (Alexander, “Memory,” 133, 138).

Aul. Gel. 8.3.

See, e.g., Philost. Vit. Apoll. 5.21; Jos. Life 11; Xen. Mem. 1.2.3; Sen. Y. Ep. Lucil. 108.4; b. Ber. 62a; Liefeld, “Preacher,” 223; Robbins, Jesus, 64; the apparently contrary statement of Eunapius Lives intro. 452–53 refers in context to casual activities only (cf. Xen. Symp. 1.1). For rabbinic behavior establishing legal precedents, see t. Pisha 2:15–16; Sipre Deut. 221.1.1; y. b. Meṣiʿa 2:11, §1; Demai 1:4 [22b]; Nid. 1:4, §2; Sanh. 7:2, §4; Yebam. 4:11, §8.

Watson, “Education,” 312.

Traditional Israelite learning had emphasized the transmission of wisdom while also allowing individual learning (cf. Shupak, “Learning Methods”).

Jos. Life 8; Ag. Ap. 1.60; 2.171–73, 204. Josephus’s statements on Jewish literacy, like that in m. ʾAbot 5:21, may reflect the literate elite, with much of the population learning Torah orally (Horsley, Galilee, 246–47); but there were, undoubtedly, reasons others considered Judeans a “nation of philosophers” (Stern, Authors, 1:8–11, 46–50; Gager, Anti-Semitism, 39), and “the synagogue was a comparatively intellectual milieu” (Riesner, “Synagogues in Jerusalem,” 209). The illiterate learned Torah orally (Kirk, “Memory,” 157–58).

See Riesner, “Education élémentaire”; idem, Jesus.

See Gerhardsson, Memory, 124–25. For Gerhardsson’s right point, see Kelber, “Work,” esp. 191–94.

Cf. Keener, “Assumptions,” 40–48; “Reply,” 117. Given the scarcity of surviving evidence from antiquity, historians of antiquity routinely draw on and compare a range of evidence, even when attestations outside the document one is analyzing may be later or from a different location; note Levinskaya, Diaspora Setting, ix–x, 16–17; Donaldson, Paul and Gentiles, 51; and discussion in this commentary’s prolegomenon.

Sipre Deut. 48.1.1–4; Goodman, State, 79; cf. Sipre Deut. 4.2.1, 306.19.1–3; b. Ber. 38b; y. Meg. 4:1, §4; Gerhardsson, Memory, 113–21, 127–29, 168–70; Zlotnick, “Memory.”

E.g., t. Yebam. 3:1; Mek. Pisha 1.135–36; Sipre Deut. 48.2.6; ʾAbot R. Nat. 24 A; Pesiq. Rab Kah. 21:5; b. Sukkah 28a; y. Šeq. 2:5; cf. m. ʿEd. 1:4–6; Sipra Behuq. pq. 13.277.1.12; see further Moore, Judaism, 1:99; Urbach, Sages, 1:68; Gerhardsson, Memory, 122–70; idem, Origins, 19–24; Riesenfeld, Tradition, 14–17. When the proper attribution was unknown, this was sometimes stated (y. Ter. 8:5).

Gerhardsson, Memory, 136–48, 173; Goulder, Midrash, 64–65. Similar sayings thus could appear in different words (m. Šabb. 9:1; ʿAbod. Zar. 3:6).

Simeon ben Azzai in Sipra VDDen. pq. 2.2.3.1, 3.

b. Soṭah 5:6, §1; cf. y. Ketub. 3:1, §4. Of course, the rabbi may have issued several different opinions on a subject in his lifetime; cf. y. B. Qam. 2:6, §3. Sometimes rabbis also seem to have told stories as fictitious homiletic illustrations instead of wishing to be understood as drawing on previous traditions (cf., e.g., Sipre Deut. 40.7.1).

Davies, “Aboth,” 156. Cf. communal memory’s selectivity for relevance (Kirk, “Memory,” 168).

See, more fully, my comment in Keener, John, 62–65; esp. idem, Matthew, 25–31, where the topic is far more relevant than in Acts.

For this verb in passing on information in early Christian circles, cf. Luke 1:2; Acts 16:4; 1 Cor 15:3.

Cf. Hermog. Progymn. 4, “On Maxim,” 8–10. (By the strictest definition, “maxims” might be unattributed; cf. the later Nicolaus Progymn. 5, “On Maxim,” 25.) On boys learning these and chreiai, see also Anderson, Glossary, 126–27 (citing Sen. Y. Ep. Lucil. 33.7). While historians did not reproduce entire speeches verbatim, they could reproduce brief sayings in this manner (Marincola, “Speeches,” 120, cites Tac. Ann. 14.59.4; 15.67.4; and compares a saying in Arrian Anab. 4.12.5; Plut. Alex. 54.6).

See documentation in Keener, Matthew, 25–26; more fully Stein, Method and Message; Tannehill, Sword.

See documentation in Keener, Matthew, 25–29. Greek and Roman philosophers also could do the same (Philost. Vit. soph. 1.22.523), even using poetry to reinforce their teaching for early students (Sen. Y. Ep. Lucil. 108.9–10), though not advanced ones (108.12; poetry and song involved memorization, Apollod. Bib. 1.3.1; Sen. E. Controv. 1.pref. 2, 19).

Though letters are indeed different from lectures, Paul’s letters, if anything, may point to a diatribe lecture style.

See also Bruce, Acts³, 35.

See, e.g., Heath, Hermogenes, 7; Olbricht, “Delivery and Memory” (esp. 159, 163, citing Rhet. Her. 1.3; Cic. De or. 2.351); Russell, “Rhetoric, Greek,” 1313; Satterthwaite, “Acts,” 344; Walde, “Mnemonics” (citing Cic. Inv. 1.9; Rhet. Her. 3.28; Quint. Inst. 3.3.1); cf. Eunapius Lives 502. Ancient rhetoricians praised memory earlier (Aeschines Embassy 48, 112), and it became part of rhetorical and other pedagogy by Cicero’s day (see Gaines, “Handbooks,” 167, citing Cic. De or. 1.142).

For other praises of orators’ memory, see, e.g., Pliny Ep. 6.11.2.

Watson, “Education,” 310.

Quint. Inst. 11.2.1–51.

Among rhetorically trained hearers, the particularly adept might memorize lengthy speech material (e.g., Sen. E. Controv. passim, below), but most hearers would not fit this category.

Naturally, this would be easier for a speech heard recently; thus Lucian Peregr. 3 (LCL, 5:5, 7) promises to repeat verbatim, as best he can, the words of a Cynic he heard the previous day. Knowing Cynic style may have helped him with memorization. People (including myself) can often vividly recall scenes that stick in their memories though neglected in their notes at the time (e.g., Grindal, “Heart,” 67, supplementing his anthropological notes with vivid memories).

For other examples of exceptional mnemonic skills, see, e.g., Sen. E. Controv. 1.pref. 19. Despite some (rarer) modern mnemonic feats that render many accounts credible, not all the ancient claims appear plausible (e.g., those in Pliny E. N.H. 7.24.88).

For ancient memory techniques, see, e.g., Small, Wax Tablets; idem, “Orator” (including as a key element in rhetorical training, 196); Walde, “Mnemonics”; Byrskog, History, 163–65. Note Quintilian’s heavy emphasis on memory in Inst. 11.2 (see also Small, “Orator,” 196–97). Small, “Orator,” 202–3, notes that one would memorize a longer work piece by piece (Quint. Inst. 11.2.27); practice and then test oneself to ensure getting it correct (11.2.34–35); and repeat one’s lines aloud (11.2.33). Further, she notes, literate persons memorizing written material helped their memories by seeing it (11.2.32) and engaging it by marking it up (11.2.28–29). As modern studies also show, a night’s rest after memorizing helps retention (11.2.43).

Kennedy, “Source Criticism,” 143, argues that given the ancient emphasis on memory and the use of commonplaces in declamations, Seneca’s recollection of declamation pieces is more credible than some critics have allowed.

Tacitus also claims to remember long dialogues years later (Dial. 1), but to some degree, this is a literary device (cf. Justin Dialogue with Trypho, though some elements also suggest that some genuine dialogues may stand behind this work; see 273n133 above); although such a dialogue may have occurred, “Maternus” in the dialogue speaks for Tacitus’s own views. Contrasting modern weak memory, Eddy and Boyd, Legend, 283, cite the reported mnemonic feats of ancient and medieval cultures (e.g., Josephus and Thomas Aquinas). Exceptionally trained memories exist today but (esp. in the West) appear far more isolated.

Although some consider Seneca’s ability exaggerated, it is noteworthy that ancient mnemonics expert Small, “Orator,” 204, does accept the reports of Seneca and also the elder Pliny; on 205–6 she contrasts the mnemonic deficiencies of modern Western society. Some mnemonic claims from much earlier periods (Val. Max. 8.7.ext. 16: Cyrus’s knowledge of all his troops’ names, or Mithridates’s of the twenty-two languages of his subjects), does appear less credible (these and also Lucius Scipio’s memory of all the Roman people’s names appear also in Pliny E. N.H. 7.24.88).

See Dunn, Perspective, 43, 114–15. Traditional Middle Eastern culture values communal memory (45–46, citing Bailey’s decades of observations). Weeden, “Theory,” critiques Bailey, but Dunn provides a sound response in “Theory.” Cf. communal religious memory in Allison, Jesus, 29, 374.

Hillard, Nobbs, and Winter, “Corpus,” 212.

With, e.g., Robertson, Luke, 228.

Hanson, Acts, 21–27, 38.

Aune, Environment, 125.

Presumably, memory works in roughly similar ways in all cultures; my point in noting the cultural difference is simply that we would not expect an oral culture’s eyewitness memory to be less dependable than modern Western memory. Cultures that inculcate memory tend to develop it more strongly than in the West.

Bauckham, Eyewitnesses, 326, 333, 344. Thus estimates of numbers and related figures (such as height or weight) appear more easily forgotten, although timing is often remembered where it fits schematic expectations (Redman, “Eyewitnesses,” 182). Similar memories blend (Kirk, “Memory,” 166).

Redman, “Eyewitnesses,” 180–81.

Ibid., 185. Inaccuracies can arise even in the first few weeks after an event (196); Redman also notes that early Christians’ faith introduces bias, though I wonder why faith is considered more a bias than any other expectation.

Ibid., 186.

Ibid. Even among individuals, errors, once introduced into memory, tend to be preserved there (187–88).

Westerners tend to choose among conflicting individuals’ memories, but more group-oriented cultures tend to find ways to combine them (“compromise memories,” ibid., 187).

Bauckham, Eyewitnesses, 327, 333–34, 345. Eyewitnesses may differ on matter of detail, e.g., the recollections of John and Charles Wesley regarding the Holy Club (Tomkins, Wesley, 33–34), whether because at least one forgot earlier incidents or because they evaluated the relevance of incidents differently. Cf. also the problematic recollection in Tuttle, Wesley, 115n6.

Bauckham, Eyewitnesses, 330, 334–38, 350.

Ibid., 331. Reports of surprising details that violate observers’ expectations are unlikely to have arisen from the frequent process of memories adjusting details to fit the schema (Redman, “Eyewitnesses,” 181–82). Redman does warn on 182 that this observation does not guarantee their accuracy; one might add, however, that it improves the likelihood. Apart from particularly noteworthy, surprising details, in cases where some information fits the preexisting schema and other information does not, what fits the schema has the better chance of being recalled accurately (181).

Bauckham, Eyewitnesses, also Redman, “Eyewitnesses,” 183 (on events that one considers significant).

Bauckham, Eyewitnesses, 331–32; Elliott, Feelings, 44–45 (citing, e.g., Izard, “Relationships,” 22). Redman, “Eyewitnesses,” 184, emphasizes these emotionally intense experiences, which produce “flashbulb memories—memories that are especially vivid and appear to be frozen in time, as though in a photograph.” She considers eyewitness experiences of Jesus’s miracles and the like as fitting this category; while these are among the most accurate of memories, they may develop in the first week through consultation with other witnesses and deteriorate over time; for reconstructive errors there, see Allison, Jesus, 7n40. The partial exception to the greater accuracy of intense memories may be experiences so intense that they produce temporary sensory overload, as on battlefields (see Woodman, Rhetoric, 18–22).

Bauckham, Eyewitnesses, 332. But see here Allison, Jesus, 7n40.

Bauckham, “Eyewitnesses,” 334; Redman, “Eyewitnesses,” 189 (noting that also with stories, elements retold most often were quickly stereotyped and remembered). Given the degree of overlap in information between Bauckham and Redman, Redman’s conclusion (based on information similar to Bauckham’s) that Bauckham’s emphasis on eyewitnesses does not really increase the probability of accuracy significantly (193, 196–97) is surprising, and may reflect a misunderstanding of Bauckham’s primary argument. Although all remembering includes distortion (194), it surely retains more information than not remembering, and eyewitnesses are still generally more reliable than nonwitnesses. Redman recognizes that Bauckham was not really claiming verbatim transmission of Jesus’s words (do any scholars claim that?), yet supposes (195–96) that he wants to trust generally trustworthy witnesses fully, whereas in fact even they do not recall everything accurately. Bauckham was not, however, claiming to trust them on every point; rather, he regards eyewitness material as more likely accurate than other material, a thesis that, on average, should still hold true.

See Heath, “Invention,” 95; Anderson, Glossary, 26–29.

Diod. Sic. 20.1–2; Aune, Environment, 93; idem, Dictionary of Rhetoric, 449.

The idea of composite collections of speech material is not new; John Calvin, trained as a humanist, allowed that the Gospel writers might have arranged some of Jesus’s sermons from independent sayings (as cited in Blomberg, Reliability of John’s Gospel, 52). On anthologies, see, e.g., Schwindt, “Anthology.”

Horsley, “Speeches and Dialogues.” See comment on .

Soards, Speeches, 160.

Ibid., 156, noting also the heavy use of 2 Esd 19 in Acts 7; Kurichianil, “Speeches,” comparing the farewell speeches in Acts 20:18–35 and 1 Sam 12:1–25. Various prophetic discourses (e.g., Jer 14:2–15:9) function as entire speech units, though this does not mean they fit Greco-Roman patterns (Gitay, “Criticism,” 13–24).

Van der Horst, “Parallels to Acts,” 56; Soards, Speeches, 139–40, both following Plümacher, Lukas, 38–78.

Soards, Speeches, 156–57; idem, “Speeches in Relation to Literature.” The exception might be if one thinks of them answering accusations from Israel (cf. 1 Sam 12:3; but there does not seem to be actual controversy [cf. 12:4]).

In favor, cf., e.g., Turner, “Thoughts,” 46–47; Nock, “Vocabulary,” 138–39.

More common today, e.g., Horsley, Documents, 5:5–40; Porter, “Greek of New Testament,” 430.

Anderson, Rhetorical Theory, 281, commenting on Paul’s style.

See discussion on Luke’s sources in Acts in , esp. , above; see also discussion in this commentary’s introduction to Acts 1.

Bauckham, “Liber antiquitatum,” 68, though doubting that this practice would apply so “readily to historical figures who were remembered as authoritative teachers and whose teaching was preserved.”

Soards, Speeches, 157–60.

Ibid., 161; Soards allows that 2 Maccabees employs repetition, but contends that Acts uses it more (159–60).

On Josephus and rhetoric, see, e.g., Runnalls, “Rhetoric”; on the rhetoric in some of his forensic speeches (the same occasions but rewritten between the War and the Antiquities), see Saddington, “Rhetoric.”

His work is full of them, including speeches to encourage for battle (e.g., Jos. War 1.373–79), to dissuade from war (2.345–401), to reject Zealots (4.162–92), and to surrender (5.375–419). By contrast, Nicolaus’s defense of Archelaus (Jos. Ant. 17.240–47) may reflect a more reliable source, probably Nicolaus’s own writings. Josephus probably follows his sources more closely in the War than in the later Antiquities (Newell, “Suicide Accounts,” 285–86), where rhetorical interests (to produce a greater literary work) are stronger.

See, e.g., Mosley, “Reporting,” 24; Hemer, Acts in History, 76–78; Attridge, “Historiography,” 326; Mason, Josephus and New Testament, 192–94; Rajak, Josephus, 80; cf. Gempf, “Speaking,” 299, who regards Josephus as exceptional. With respect to compositional standards, however, Price, “Failure,” argues that Josephus’s careful composition meets Thucydidean standards.

Aune, Environment, 107. For his speeches’ deliberately revealing his distinctive theology and perspectives, see Lindner, “Geschichtsauffassung” (summarizing Lindner’s dissertation; addressing speeches in War 2, 5, and 7); Josephus uses his allies’ speeches to advance his own position (Rajak, Josephus, 180), these allies functioning as reliable characters. He does supply some opposing speeches.

Aune, Environment, 108; Bünker, “Disposition der Eleazarreden.”

See Rajak, Josephus, 81; for their function in denouncing rebels, see Rajak, Josephus, 81–82.

See Kelley, “Perspective.”

Gempf, “Speaking,” 289–90.

Ibid., 290. See also Jos. Ant. 10.7 (which echoes 2 Kgs 18:25); Josephus omits the more directly affirmed divine causation in 2 Chr 35:21–22 (Ant. 10.75–76; also missing in 2 Kgs 23:29).

Jos. War 1.373–79; Ant. 15.126–46; often noted, e.g., Gempf, “Speaking,” 290; Cadbury, Foakes-Jackson, and Lake, “Writing History,” 13–14; Dunn, Acts, xviii. (Woodman, Rhetoric, 176–79, suggests what could be an analogous change from Tacitus’s Histories to his Annals, though we should not presume that Tacitus lacked other sources, and hence details, no longer extant.) Luke varies wording in two versions of a speech in his two-volume work (Luke 24:47–49; Acts 1:4–8) but without changing the basic sense.

Mosley, “Reporting,” 24n2, illustrates by comparing 1 Macc 2:50–68 and Jos. Ant. 12.279–84; Aune, Environment, 82, by comparing 1 Macc 8:23–32 with Jos. Ant. 12.417–18 (where Josephus reduces the speech nearly by half). Still, Gafni, “Josephus and Maccabees” (esp. 126–27), emphasizes that even in the speeches, Josephus tends to adapt 1 Maccabees for his Hellenistic audience, not to create.

The only element of reproof sounded appears in Jos. Ant. 19.290, warning Jewish people not to look down on other nations’ practices, but even this is put in Jewish terms, filtered through Josephus’s perspective. It sounds much less explicit than Claudius’s warning that they should stop agitating for citizenship (P.Lond. 1912, col. 3, lines 88–89)!

Sanders, Judaism, 6.

Cohen, “What Happened at Masada?” But some scholars do suggest that there remain historical reminiscences beneath his hellenized reworking (cf. Bauernfeind and Michel, “Beiden Eleazarreden”).

Rajak, Josephus, 80–81. For the “noble barbarian” tradition, cf. this commentary’s introduction to Acts 8:26–40; briefly, Keener, Matthew, 105n96.

Unless Josephus invented them as literary informants; but given the conspicuousness of Josephus’s free composition, this is unlikely.

Granted, some women studied philosophy (Diog. Laert. 1.89–90; 2.72, 86; 4.1; 8.1.42–43; Iambl. V.P. 36.267; Lefkowitz and Fant, Life, 160, §168), including well-known examples such as Hipparchia (Diog. Laert. 6.7.96–98; Diogenes Ep. 3; Crates Ep. 28–29) and later Sosipatra (Eunapius Lives 466–69); for the Stoic ideal, see, e.g., Mus. Ruf. 4; Ward, “Musonius,” 288; among Jewish scholars, Beruriah (cf. Adler, “Virgin”). Others were also well educated (e.g., Sall. Catil. 25.2; Quint. Inst. 1.1.6; 4 Macc 18:11–19; Jos. Life 185; Sipre Deut. 307.4.1; Sop. 18:5; cf. Blomqvist, “Chryseïs”; Brooten, Women Leaders, 5–33; Blank, “Texts”). But the training was sometimes limited (Mus. Ruf. 4, p. 46.8–23; Ilan, Women, 190–204; Swidler, Women, 104–11) and was much rarer than with men (Gamble, “Literacy,” 645; Scholer, “Writing,” 1283; Safrai, “Education,” 955; Goodman, State, 74; Wegner, Chattel, 161–62; Swidler, Women, 97–104; Davies, Rhetoric, 252).

Sanders, Judaism, 7, thinks that Josephus portrays the Sicarii here more nobly than he dare elsewhere (and that he authentically portrays their spirit). Ladouceur, “Josephus and Masada,” argues (plausibly) that Josephus views the speech and suicide negatively rather than heroically or as justified by Stoic notions of suicide as a choice.

See more fully Luz, “Masada” (especially his comparisons with Greek philosophic rhetoric). Cf. Paetus’s poetic dying lamentations in Prop. Eleg. 3.7.57–64—to which there were obviously no witnesses.

But Josephus may have created even Agrippa II’s speech (in the form in which it appears in War 2.345–404) as commentary revealing his views (Roduit, “Discours”).

Rajak, Josephus, 80–82, 180.

Gempf, “Speaking,” 290.

So, e.g., Guthrie, Introduction, 381 (although he does believe [p. 382] that Luke preserves genuine reminiscences).

See extended comment at Acts 13:38–39, including the table comparing Pauline phraseology. One can also compare Paul’s preaching against idolatry and even (against some) the approach to natural revelation (despite differences; see comment on Acts 17:22–31).

Cf. also Robertson, Luke, 228.

See Pichler, “Anliegen,” 742–43. According to the count in Borgman, Way, 11, Peter and Paul each deliver seven speeches.

E.g., Knox, Acts, 16–39.

Schweizer, “Davidic ‘Son,’” 208–16. For structural similarities between Peter’s first two sermons, see Zehnle, Pentecost Discourse, 19–24.

Schweizer, “Davidic ‘Son,’” 214. For audience adaptation in good rhetoric, see, e.g., Dion. Hal. Lysias 9; Quint. Inst. 3.7.24.

One should also adapt one’s style for the different genres of speeches (Dion. Hal. Demosth. 45–46).

See Via, “Moses and Meaning”; followed by Puskas, “Conclusion: Investigation,” 42–43. Many of these components are, admittedly, self-evident elements of narrating a speech.

Dibelius and Kümmel, Paul, 11; Goulder, Type and History, 83. The classic statement of the non-Pauline character of Paul’s speeches in Acts is Vielhauer, “Paulinism,” but he clearly overstates his case (cf. Fitzmyer, Acts, 145–47; and above).

Cf. Barrett, “Acts and Corpus.”

Dodd and Dibelius found in these patterns “reproductions of the apostolic preaching,” but Wilckens thought otherwise (see Missionsreden); see van Unnik, “Storm Center,” 26.

Even Paul’s first-century opponents apparently “distorted” the “gospel” not by denying the resurrection but by undermining Paul (2 Cor 11:4–6) or requiring circumcision for salvation (cf. Gal 1:6–7; contrast the “pillars” in 2:2, 7, though Paul feels that Peter compromises the gospel’s implications in 2:14).

See Bruce, Commentary, 277, comparing Acts 2 with Acts 13; cf. also Peterson, “Extent”; Nanos, Mystery, 354, 358n45. Marshall, “Resurrection,” 92–107, argues that the resurrection preaching in Acts derives from the earliest Christian sources. For some of the common elements in Acts’ preaching, see Dodd, Preaching; and, e.g., in Petrine speeches, D’Souza, “Sermons of Peter.” Dibelius himself allowed that the sermons followed the ancient kerygma also found in 1 Cor 15 (Tradition, 16–19).

As often affirmed since Dodd (Preaching); e.g., D. Williams, Acts, 10–11; deSilva, Introduction, 352–53 (noting parallels in 1 Peter).

Given the paucity of first-century Christian sources overall, one must avoid arguing too much from this silence. See here Bruce, Acts³, 37, also arguing that the use of Scripture in these speeches (with, e.g., “it was fulfilled”) is not characteristic of Luke’s own narrative and hence may reflect an earlier “apostolic” method in speeches (following Harris, Testimonies, 2:80).

Bruce, Commentary, 277, comparing Acts 2 with Acts 13; for Acts and Paul, cf. Bruce, “Paul’s Use of Old Testament.”

Soards, Speeches, 182–83 (quote, 183), though some speeches are more to advance the narrative than to put forward a view. But on stylistic unity see also my pp. , .

Ibid., 203–4.

With, e.g., Munck, Acts, xliv; Soards, Speeches, 12–13. Rosner, “Biblical History,” 76, finds in Acts and the Deuteronomistic History analogous methods of periodization.

Witherington, Acts, 46; cf. idem, “Addendum,” 31; Acts 2:40.

Cf. Dibelius, Studies, 2, 184–85, 201; Dupont, Sources, 166.

Koester, Introduction, 1:108. Rhetoricians knew (cf. Anderson, Glossary, 39) but did not favor such paratactic style.

Witherington, Acts, 117.

Hanson, Acts, 37; Lüdemann, Christianity, 47–48; Hengel, Acts and History, 61; Aune, Environment, 126; Fitzmyer, Acts, 105–6; Witherington, Acts, 117–19; Porter, Paul in Acts, 130.

Bruce, Commentary, 354, claims that classicists tend to find it authentic (citing Cadbury, “Speeches,” 406n1); Bruce, Apostle, 244.

McCoy, “Shadow of Thucydides,” 15–16; Witherington, “Addendum,” 31; Laistner, Historians, 86 (uniform style in Thucydides and other historians); Porter, Paul in Acts, 130–31; Aune, Dictionary of Rhetoric, 448 (though allowing that some of these are inauthentic). This practice could, of course, occasion the complaint of a rhetorical historian (Dion. Hal. Pomp. 3.20, cited by Fitzmyer, Acts, 105); on memorizing or rewording speeches in earlier written material, see Dio Chrys. Or. 18.18–19. Thucydides’s speeches do not resemble extant speeches actually delivered in his period but follow rhetorical handbooks from the era (Walbank, Speeches, 3), further suggesting that he has composed them (though not demanding that they are bereft of accurate information; see Walbank, Speeches, 4).

Wharton, “Tacitus’ Tiberius” (noting that this conclusion does not resolve whether Tacitus used official sources).

So Walbank, Speeches, 13–15; see esp. 13: “If a point is valid, it may be made twice, and if all fourth-century speeches which mention the services of Athens against the Persians were to be branded as rhetorical forgeries, we should have few pro-Athenian apodeictic speeches left.” Polybius might have composed such points freely; the point here is that we need not assume that he did so on the basis of modern content criticism.

Porter, Paul in Acts, 130–31.

Aune, Environment, 91, giving as an example the savage Cyclops’s “‘inappropriately’ elegant speech” in Hom. Od. 9.447–60.

When it was available, sometimes they did reproduce the style, even when it was not rhetorically pleasant (e.g., Epict. Diatr. 1.pref. [see comments in Oldfather, “Introduction to Epictetus,” xii–xiii]; the claim in Philost. Vit. Apoll. 8.6). Probably for reasons suggested above, Luke seems to underline this unity in the speeches far more than his elite contemporaries. Because prosopopoeia invited composing in varied styles, Luke appears less rhetorically sophisticated than some of his contemporaries (despite Semitic coloring in early chapters, Paulinisms, etc.).

Hengel, Acts and History, 61, citing Lucian Hist. 58 and also Thucyd. 1.22.

Aune, Environment, 125.

Ibid., 126 (Aune does not believe that Thucydides succeeded in his goal of accuracy regarding speeches in every case [p. 92], but remains even more skeptical of Herodotus [91]).

Cf., e.g., the insistence of Marshall, Acts, 41–42, on genuine Lukan compositional freedom alongside tradition (also idem, “Acts,” 516). Despite the prejudicial stereotype among some of their detractors, the vast majority of conservative biblical scholars do not even claim verbatim reports (or something like exact translation) for Jesus’s sayings in the Gospels, which plainly vary in wording (cf. Stagg, Acts, 26; Wenham, Bible, 92–95; Feinberg, “Meaning,” 299–301 [the exact voice, but not words, of Jesus]; Bock, “Words,” 75–77); “minor variations” in Paul’s conversion stories likewise show lack of interest in verbatim retellings (Stagg, Acts, 26–27; Abbott, Acts, 111; Witherington, Acts, 311; Dunn, Acts, 117). Recognizing the literary genres involved, even the strictest models of inspiration would not ascribe to personages mentioned in, e.g., the book of Job the poetic discourses attributed to them (Metzger, “Forgeries,” 22). The use of speeches in the genre of ancient history may likewise be taken into account.

See Powell, Acts, 32.

Presumably, most of Luke’s ideal audience had never heard of Theudas, but perhaps some had.

This is one approach; other alternatives are composition without even attempting to imitate Paul (rhetorically improbable); insufficient information to imitate Paul well; and Luke’s (or his source’s) knowledge, from Paul, of the gist of what he said on that occasion. See this commentary’s comment ad loc.

See, e.g., Porter, Paul in Acts, 145. The approach in these texts is not, of course, exclusively Pauline; from Philo to Augustine, monotheists challenged pagan natural philosophy (Miller, “Idolatry”).

Others have also offered this comparison; e.g., Dodd, Preaching, 18; Ridderbos, “Speeches of Peter,” 9; Hemer, Acts in History, 78–79.

I treat the issue of Jesus’s sayings in more detail in Keener, Matthew, 24–32; idem, John, 29–34; idem, Historical Jesus, –.

E.g., Smalley, “Christology of Acts”; Munck, Acts, xlii; Longenecker, Exegesis, 82.

Witherington, “Editing,” 326–27 (quote, 326).

Correctly, e.g., Liefeld, Acts, 63.

Bruce, Speeches in Acts, passim; cf. idem, “Speeches Thirty Years After”; for OT use, idem, “Paul’s Use of Old Testament.”

E.g., Ridderbos, “Speeches of Peter,” 9; Munck, Acts, xliv; Longenecker, Exegesis, 80–82; Gasque, “Speeches,” 248–49; idem, “Acts and History,” 58–63; Fitzmyer, Acts, 105–6; Witherington, Acts, 117; Dudley, “Speeches in Acts.”

Of course, it is precarious to define “accuracy” and “gist” too narrowly, since what Luke had available may have varied from one speech to the next, and even in Luke’s Gospel, many scholars argue that he may take more liberty in constructing the Nazareth synagogue pericope than with most other scenes. At the very least, the gist would normally be the subject of the speech given on an occasion, along with whatever could be reconstructed of how it was developed on the basis of tradition regarding the speech, regarding the speaker, or (where the least would be available) regarding the era. Such a definition allows Luke literary flexibility but also allows him to draw on substantial tradition where he had it available. Given the brevity of the speeches, Luke probably does not need to flesh out his sources much.

Dibelius, Studies in Acts, 138–85; Townsend, “Speeches”; Wilckens, Missionsreden; Schweizer, “Concerning Speeches.”

E.g., Jones, “Christos in Luke-Acts”; for theological unity, Townsend, “Speeches.”

Zehnle, Pentecost Discourse, passim (e.g., p. 37 on the “masterpiece”; 43 on Acts 10:34–43). Lüdemann, Christianity, 48, thinks that elements of Acts 2 (such as 2:21, 34–35) are pre-Lukan.

Despite Luke’s portrayal of other aspects of the earliest church as ideal or pristine (cf. Acts 2:44–45). Granted, he can use the apostles as a foil for God’s work (as in the gospel tradition), but (as again in the gospel tradition) enemies might have made better foils than the church’s “heroes” if it did not happen much as Luke suggests.

Wilcox, “Foreword to Speeches.”

Dodd, Preaching, 19–20, following Torrey, Composition, who viewed most of the first half of Acts as Aramaic material; cf. Martin, “Evidence,” 59; Payne, “Semitisms”; Ehrhardt, Acts, 1. Torrey underestimated the extent to which Koine, Semitic (or aramaized Greek), and translation Greek overlap (cf. LXX; Jos. As.; cf. Turner, “Thoughts,” 46; Nock, “Vocabulary,” 138–39; though, for Rome, contrast Leon, Jews, 92), but it is true that apparent Semitisms predominate in the early chapters. Proposed Semitisms must be examined on a case-by-case basis.

Gasque, “Speeches,” 248–49; Bruce, Acts¹, 18.

At least—for the sake of verisimilitude—in Paul’s speeches. But that is, in fact, what we find for Paul’s speeches (see discussion below).

C. Williams, Acts, 47–48; Moule, “Christology of Acts”; Smalley, “Christology of Acts,” 358; idem, “Christology Again”; Robinson, Studies, 139–53 (regarding the Christology of Acts 3:12–26 as more primitive than that in 2:14–39).

Knox, Acts, 75–76.

Ibid., 77.

Ibid., 78. One may of course unintentionally echo one’s sources at points; thus, for example, some echoes of the voices of my interviewees remain in my reports of their words in Miracles (esp. in 264–599, 712–59, passim) despite editing by myself and other editors.

Doeve, Hermeneutics, 176. Cf. examples I note in comment on Acts 20:18–35.

C. Williams, Acts, 44; Soards, Speeches, 140. This may reflect deliberate septuagintalizing (Hengel, Acts and History, 62; De Zwaan, “Greek Language”) to fit the character of his speakers.

For christological language from Luke’s own day in Acts, see, e.g., Jones, “Christos in Luke-Acts.”

It does, after all, reflect the language of Ps 110:1 (Heb 1:13). The source of enthronement is less specific in Heb 1:3; 8:1; 10:12; 12:2; 1 Pet 3:22.

Hanson, Acts, 37–38.

E.g., Hengel, Acts and History, 61; Munck, Acts, xliv; Kistemaker, “Speeches in Acts”; Williamson, World of Josephus, 290, as quoted in Yamauchi, Stones, 97; D. Williams, Acts, 11.

E.g., Philost. Hrk. 11.5 (on Hom. Od. 18.359, using Il. 21.197); Dion. Hal. Lysias 11–12; Demosth. 50. But sometimes the same speaker employed different styles in different genres of writing (e.g., Cic. Fam. 9.21.1; Philost. Vit. Apoll. 7.35), and so even lack of similarities would not necessarily prove different authors.

E.g., Selwyn, Peter, 33–36; D. Williams, Acts, 11; Smalley, “Christology Again” (though I doubt that “Petrine” Christology is that distinctive). Some themes recur in Peter’s speeches, though these are adapted for different settings (cf. Tannehill, “Functions of Peter’s Speeches”). Kistemaker, Acts, 9–11, finds some convoluted Greek in Peter’s speeches (Acts 3:16; 10:36–37) comparable to 1 Pet 4:11; 2 Pet 3:5b–6, though convoluted Greek appears elsewhere. He also parallels appeals to God’s foreknowledge (Acts 2:23; 1 Pet 1:2) and references to the judge of the living and the dead (Acts 10:42; 1 Pet 4:5), but one might compare also non-Petrine texts such as Rom 8:29; 2 Tim 4:1. For prophecy in Petrine sources, see now Himes, “Peter.”

E.g., Baum, “Paulinismen,” esp. 414–35, on Paul’s mission speeches; see more detailed discussion in this commentary on relevant passages. Some appeal is also made to unusually distinctive local color in locations that Luke is not likely to have visited himself (e.g., Porter, Paul in Acts, 139, following esp. Breytenbach, “Zeus und Gott”). Some speeches to Jewish audiences also imply midrashic techniques in Luke’s sources that he does not develop and may not have recognized (see comments esp. on ).

Johnson, Acts, 367; cf. Hemer, “Ephesian Elders”; Stanton, Jesus of Nazareth, 111; Witherington, Acts, 171.

Walton, Leadership, 1, noting that it is “widely recognised as paralleling the language and ideas of the Pauline epistles.”

See, e.g., in Barnes, “Finishing,” 240; Gilliland, “Missionaries,” 260–70.

On which see, e.g., Malherbe, “Paul as Pastor” (esp. on 1 Thess, 126–28). Paul’s own ministry style was flexible, reflecting both the compassion and the sternness of a parent (e.g., 1 Cor 4:14–21; 2 Cor 12:14; 1 Thess 2:7, 11; cf. the thesis of Joubert, “Shifting Styles,” even if one might differ on the examples).

Malherbe, Philosophers, 152–54.

See ibid., passim (e.g., on 1 Thess, 49–66; but esp. 67–77, even if Malherbe’s conclusion [77] may be overstated).

From Kistemaker, Acts, 11–12. Parallels also appear in the speech summary in Acts 14:22, again addressed to believers (see comment ad loc.).

Walton, Leadership, 140–85; he also shows (203–12) that these parallels reflect dependence not on 1 Thessalonians but on authentic Pauline tradition. He observes (186–92) some parallels with Ephesians but notes that they are far less close than with 1 Thessalonians.

Hanson, Acts, 203, includes recalling the past (Phil 1:5; Col 1:6; cf. also, e.g., 1 Cor 2:1–5; Gal 4:13–14), tears (2 Cor 2:4; cf. also Phil 3:18), and protesting innocence (1 Thess 2:3–11; 2 Cor 7:2). Dunn, Acts, 271, adds others, most notably “profitable” (Acts 20:20; 1 Cor 6:12; 10:23; 12:7; 2 Cor 8:10; 12:1).

Porter, Paul in Acts, 117. Four of these also appear in Kistemaker, Acts, 11–12.

The verb δουλεύω appears elsewhere in Acts only at 7:7 (though in the Gospel at Luke 15:29; 16:13); it appears sixteen times in Pauline literature, in this sense in Rom 14:18; Eph 6:7; Col 3:24; 1 Thess 1:9; cf. Rom 7:25.

The exact match appears only in Eph 4:2 in the NT; the term ταίεινοφροσύνη is also mostly from Paul’s later years (Eph 4:2; Phil 2:3; Col 2:18, 23; 3:12; elsewhere in the NT only 1 Pet 5:5). Cognate evidence is more ambiguous; the cognate adjective appears three times in Paul (against once in Luke); the verb only once more in Paul than in Luke.

It is also Lukan (Acts 14:1; 18:4; 19:10, 17), though in the Diaspora, Pauline, section of Acts. As a Hellenist in Jerusalem, Paul probably naturally thought in these terms.

More general athletic imagery, including running, is common in Paul, as among other Diaspora teachers (1 Cor 9:24, 26; Gal 2:2; Phil 2:16; 3:12–14; see more fully Pfitzner, Agon Motif).

Eleven times in Pauline literature, most of them undisputed references; four other times in Luke-Acts (one in a summary of Pauline speech, Acts 13:43), with four other uses in the NT. “The grace of the Lord” or “our Lord” appears twice in Luke-Acts (Acts 15:11, 40) and eleven times in Paul (mostly undisputed; with only one other use in the NT).

Seven times in Paul (six undisputed).

Mainly Pauline in this sense, though see also Acts 9:31; perhaps 1 Pet 2:5, 7.

One of Paul’s letters written a year or two before this journey cites his own example, explaining why he works rather than accepts patronage (1 Cor 9:1–18).

He worked hard with his hands night and day (cf. Acts 20:31) so that he could preach to the Thessalonians freely (1 Thess 2:9). He did this, as here, as an example (2 Thess 3:8–9; cf. 1 Cor 4:12, 16); they too should work (1 Thess 4:11; 2 Thess 3:10–12; Eph 4:28), as here (Acts 20:25). Likewise, he emphasized to the Corinthians that he had worked (1 Cor 9:3–18) to avoid hindering the gospel (9:12).

Porter, Paul in Acts, 117–18.

With, e.g., Haacker, Theology, 143–44.

Although I will argue that they are not incompatible, the case for Acts’ reliability about Paul rests on correlations with reports of historical incidents more than on stylistic or linguistic parallels.

Aune, Environment, 125–26.

Hengel, Acts and History, 61–62 (suggesting that Luke treated Jesus’s sayings differently because he treated them like biblical quotes; cf. Acts 20:35).

Hengel, Acts and History, 61.

Gasque, “Speeches,” 248–49 (quote, 249); see similarly Bruce, Acts¹, 18.

Witherington, Acts, 44 (citing Aristides Ars Rhetorica 1.13.4 on the greater need for rhetorical skill in speeches than in narratives).

Gasque, “Speeches,” 249.

That these examples are relatively brief makes the verbatim reporting of longer speeches all the less plausible.

Dodd, Preaching, 17.

Ibid., 18–19.

With, e.g., Munck, Acts, xliv; Fitzmyer, Acts, 105–6; Bruce, Speeches in Acts; idem, Bruce, Acts¹, 21; Smalley, John, 164; Porter, Paul in Acts, 125; Hemer, Acts in History, 415–27; Dunn, Acts, xvii–xviii; Gasque, “Speeches,” 248–49; idem, “Acts and History,” 58–63; Dudley, “Speeches in Acts.”

Dunn, Acts, xvii–xviii; “Luke’s impression” of early Christian theology (based on probabilities available to him) as he has interpreted it (p. xviii); cf. Puskas and Crump, Introduction, 109; Allison, Jesus, 455. Witherington, Acts, 119–20, also admits that Acts 25:13–22 may be Luke’s best reconstruction.

↑ Back to top