² Pohlenz, 43, cited by I.J.F. de Jong, ‘Narrative unity and units’, in Brill's Companion (n. 1), 245–66, at 249, who translates: ‘He is a traveller who knows exactly where he wants to go, who has exactly plotted the main stages of his voyage and keeps to them, but who also allows himself the time to look at all beautiful and interesting things which the road offers, and needs not even be afraid to make long detours to this end, because he knows that he will eventually rejoin the main road at the right point.’

³ Dewald, 272–3 divides Herodotus’ authorial ‘I’ into two voices, ‘the narrator’ and ‘the histōr’. I generally use ‘Herodotus’ to refer to the narrator, but sometimes to denote the histōr as well. On Herodotus’ travels, see D. Asheri, ‘General introduction’, in id. et al., 1–56, at 6–7. Redfield, J.M., ‘Herodotus the tourist’, CPh 80 (1985), 97–118 Google Scholar, repr. most recently in R.V. Munson (ed.), Herodotus: Volume 2: Herodotus and the World (Oxford Readings in Classical Studies) (Oxford and New York, 2013), 267–91 calls Herodotus a tourist to contrast his methodology with that of modern anthropologists who travel as participant observers. Dewald, 276 discusses Herodotus as a guide on the road of logoi.

⁴ Paths of logoi: 1.117.2 and 2.20.1–2.22.1. Justification of ‘digressions’: e.g. 3.60.1 and 4. Return to earlier narratives: 4.82 and 5.62.1 = #13 and #14; preview: 7.77. The examples of ‘paths’ of logoi are noted by M.L. Lang, Herodotean Narrative and Discourse (Martin Classical Lectures 28) (Cambridge, MA and London, 1984), 151 n. 8. See also O. Becker, Das Bild des Weges und verwandte Vorstellungen im frühgriechischen Denken (Hermes Einzelschriften 4) (Berlin, 1937), 101–38, at 101–16; Létoublon, 185–6; and Dewald, C., ‘Narrative surface and authorial voice in Herodotus’ Histories ’, Arethusa 20 (1987), 147–70Google Scholar, at 165. De Jong (n. 2), 246–58 summarizes scholarship on the ‘digressions’ and argues that the term should be dropped. ‘Detour’ is perhaps a more useful concept, since the passages in question often feature ring composition, as noted by W.M. Bloomer, ‘The superlative nomoi of Herodotus's Histories’, ClAnt 12 (1993), 30–50, at 32, by which Herodotus returns to the earlier route, as Pohlenz, 43 describes. Purves, 118–58 argues that Herodotus’ style and narrative reflect a ‘hodological’ model.

⁵ 1.5.3 is the first place in which Herodotus calls his work a logos, a term by which he refers to both the Histories as a whole (e.g. 1.5.3 and 1.95.1) and narratives within it (e.g. 2.38.2), as O. Murray, ‘Herodotus and oral history’, in N. Luraghi (ed.), The Historian's Craft in the Age of Herodotus (Oxford and New York, 2001), 16–44, at 24–5 points out.

⁶ Dewald, 271 counts 1,086; in her 1987 article (n. 4), 150 n. 10 she counts 1,087.

⁷ Létoublon, 183–4 notes that Homer always uses ἔρχομαι with a future participle to express movement in space and purpose, e.g. Il. 1.12–13 and 2.801. Although Thielmann, 55 gives some examples in Homer that he thinks suggest a move towards an auxiliary usage, none of them is convincing. Without citing Thielmann, Becker (n. 4), 101 n. 1 and F.R. Adrados, Nueva sintaxis del griego antiguo (Madrid, 1992), 453 recognize the difference between Homer's usage in two of Thielmann's examples (Il. 13.256-7 and 14.205) and Herodotus’. Thielmann, 57 also identifies five allegedly periphrastic uses of ἔρχομαι or εἶμι with a future participle in tragedy from around the same time as Herodotus (ἔρχομαι at Eur. Alc. 503–4; εἶμι at Eur. Suppl. 326–7 and Hec. 579 and Soph. Trach. 83 and Phil. 1366), but in each case the verb seems to describe literal movement, and none of them is with verbs of saying. All but one of the commentaries I consulted agree that these verbs express physical motion: on Eur. Alc. 503–4, see W.S. Hadley, Euripides: The Alcestis (Cambridge, 1912), 94 and contra, H.W. Hayley, The Alcestis of Euripides (Boston, 1898), 118, who, however, acknowledges the ambiguity; on Suppl. 326–7, J. Morwood, Euripides Suppliant Women (Oxford, 2007), 65 and 169; on Hec. 579, A. Garzya, Euripide: Ecuba (Rome, 1955), 80 and K. Matthiessen, Euripides, “Hekabe” (Berlin and New York, 2010), 328; on Soph. Trach. 83, R.C. Jebb, Sophocles: Plays. Trachiniae (London, 2004; orig. Cambridge, 1892), 17 and P.E. Easterling, Sophocles Trachiniae (Cambridge, 1982), 83; and on Phil. 1366, R.C. Jebb, Philoctetes (London, 2004; orig. Cambridge, 1898), 211.

⁸ E.g. Thomas, 245 elides the future and the sense of movement by translating it as ‘I cannot say’; it is not mentioned by Fowler, 83–6 or Węcowski, M., ‘The hedgehog and the fox: form and meaning in the prologue of Herodotus’, JHS 124 (2004), 143–64CrossRefGoogle Scholar, nor by D. Fehling, Herodotus and His ‘Sources’: Citation, Invention and Narrative Art (trans. J.G. Howie, ARCA Classical and Medieval Texts, Papers and Monographs 21) (Leeds, 1989), 58–9, although he focusses on the sentence as a transitional statement.

⁹ Stein, 8 comments briefly on the construction's importance. Brock, 8 n. 13 observes that ‘the periphrastic constructions with ἔρχομαι tend to highlight the idea of movement in the narrative’ but does not go into further detail. Only one occurrence (2.11.1 = #3) is discussed in Asheri et al., at 250. It is treated as a periphrasis by Thielmann, 55–6; J.M. Stahl, Kritisch-historische Syntax des griechischen Verbums der klassischen Zeit (Heidelberg, 1907), 686; Dietrich, W., ‘Der periphrastische Verbalaspekt im Griechischen und Lateinischen’, Glotta 51 (1973), 188–228 Google Scholar, at 217–18; Coseriu, E., ‘Der periphrastische Verbalaspekt im Altgriechischen’, Glotta 53 (1975), 1–25 Google Scholar, at 6–7; and A. Rijksbaron, The Syntax and Semantics of the Verb in Classical Greek: An Introduction ³ (Chicago, 2006), 127.

The standard grammars are unhelpful: W.W. Goodwin, Syntax of the Moods and Tenses of the Greek Verb (Boston, 1897), §895 explains that, with some verbs of coming and going, the participle contains the main idea, but this undervalues the importance of ἔρχομαι. R. Kühner and B. Gerth, Ausführliche Grammatik der griechischen Sprache. Zweiter Teil: Satzlehre (Hannover, 1904), 60–1 attempt to explain it as a form of emphasis on the participle, but do not separate concrete from metaphorical usage, as Létoublon, 183 n. 11 points out. H.W. Smyth, Greek Grammar, rev. G.M. Messing (Cambridge, MA, 1984; orig. 1956), does not comment on any of the Herodotean examples.

¹⁰ Lucian uses it in the proem of Syr. D., written in a pseudo-Ionic style in imitation of Herodotus, and also cites an anonymous epigone of Herodotus who began his history of the Parthian War with ἔρχομαι ἐρέων περὶ ῾Ρωμαίων καὶ Περσέων (Hist. conscr. 18 = BNJ 203 F 1). As Létoublon, 187 describes, after Herodotus, Xenophon and Plato also use ἔρχομαι as an auxiliary with a metaphorical sense. Xenophon uses it in a similar way as Herodotus, but for the first time puts the participle before the verb (Ages. 2.7: οὐ τοῦτο λέξων ἔρχομαι, ‘I am not going to say this’). Plato extends the usage to other persons and moods of ἔρχομαι and to participles of verbs other than of saying (e.g. Tht. 198e4). The expression becomes popular again with Atticizing authors in the Roman period: besides the examples from Lucian, J.L. Lightfoot, Lucian: On the Syrian Goddess (Oxford and New York, 2003), 289 cites Dio Chrys. Or. 38.6, Aristid. Or. 26.42 Keil and Ael. NA 9.8. See also S. Ek, Herodotismen in der Archäologie des Dionys von Halikarnass: ein Betrag zur Beleuchtung des beginnenden Klassizismus (Lund, 1942), 45–7, for examples in Dionysius of Halicarnassus. A TLG search shows that the expression explodes in popularity in Late Antiquity.

¹¹ For my purposes it is irrelevant whether he actually visited the places he describes; what matters is that he says he did. In this regard I follow N. Luraghi, ‘Local knowledge in Herodotus’ Histories’, in Historian's Craft (n. 5), 138–60, at 140–1 on Herodotus’ source citations.

¹² This use of ἔρχομαι with a future participle appears neither in the poets mentioned above (n. 7) nor in the extant fragments of the twelve prose-writers mentioned by Dion. Hal. Thuc. 5, nor in the eight additional authors that Fowler, 68 thinks Herodotus could have known, nor in the ten known beginnings of fifth-century prose works by named authors (listed by Fowler, 69 n. 60). Since the remains of Herodotus’ predecessors and contemporaries in prose-writing are so slight and because expressions describing progress through the narrative are likely to occur only in texts of some length, one cannot say for sure that Herodotus was the first to use ἔρχομαι with a future participle of a verb of saying in this way, but one can say that he saw it as an important component of the Histories. For this view of Herodotus’ innovation, cf. Fowler, 69 and 74. A TLG search of the Hippocratic corpus gives only one instance of the first-person present ἔρχομαι, at Hippoc. [Ep.] 27 in É. Littré, Œuvres complètes d'Hippocrate, vol. 9 (Paris, 1861), 418, where it is used as an auxiliary with an infinitive. Incidentally, this pseudepigraphic letter evokes and even ‘corrects’ Herodotus’ narrative in several places, as W.D. Smith, Hippocrates: Pseudepigraphic Writings (Leiden, 1990), 2–4 describes. At 5–7 he dates [Ep.] 27 to between 350 and 250 b.c.e.

¹³ Cf. the discussion of ‘presentation by negation’ in S. Hornblower, ‘Narratology and narrative techniques in Thucydides’, in id. (ed.), Greek Historiography (Oxford and New York, 1994), 131–66, at 152–7. Many have read 1.5.3 as declaring a division between mythical and historical time, e.g. Pohlenz, 5–6, but Rood, T., ‘Herodotus’ proem: space, time, and the origins of international relations’, Ariadne 16 (2010), 43–74 Google Scholar, at 65–6 (with bibliography at 65 n. 58) points out that Herodotus does not rigidly separate them.

¹⁴ On his refusal to choose between different versions, see Dewald (n. 4), 162–3 and D. Lateiner, The Historical Method of Herodotus (Phoenix Supplementary Volume 23) (Toronto, Buffalo and London, 1989), 76–90. Fowler, 77 stresses that Herodotus is the first author to discuss alternative sources.

¹⁵ Stein, 8 and J. Marincola, ‘Odysseus and the historians’, in Munson (n. 3), 109–32, at 114–15 emphasize this reading of 1.5.3 in connection with the Homeric echo. G. Nagy, Pindar's Homer: The Lyric Possession of an Epic Past (Baltimore and London, 1990), 231 links the semantics of Herodotus’ οἶδα (‘I have seen: therefore I know’, his gloss and italics) with Homer's ἴδεν and ἔγνω. Pohlenz, 9 n. 1 also compares Parmenides’ description of the path ‘that carries a man who knows through all towns’ (DK 28 B 1.2–3: ὁδὸν … ἣ κατὰ πάντ’ ἄστη φέρει εἰδότα φῶτα).

¹⁶ On the polemical tone of 1.5.3, see Lateiner (n. 14), 35.

¹⁷ It has been so construed: e.g. H.M. Johnson, Herodoti Orientalia antiquiora; comprising mainly such portions of Herodotus as give a connected history of the East, to the fall of Babylon and the death of Cyrus the Great (New York, 1859), 83 translates: ‘I come not to say, i.e. with the intention of asserting’ (his italics). A further objection to this version is that it raises the question of where Herodotus is ‘coming’ from.

¹⁸ LSJ⁹ s.v. ἔρχομαι A.IV.2: ‘in Hdt. like an auxiliary Verb, … I am going to tell’. Stein, 8 translates it as ‘schicke mich an zu sagen, franz. je vais dire’, and Ph.-E. Legrand, Hérodote: Histoires, vol. 1 (Paris, 1932), 16 as ‘je ne vais pas prononcer’. K. Bentein, Review of L. Amenta, Perifrasi aspettuali in greco e in latino: origini e grammaticalizzazioni (Milan, 2003), Journal of Greek Linguistics 10 (2010), 127–41, at 128 notes the parallels in modern languages. But as Létoublon, 186–8 explains, if ἔρχομαι were lexicalized as an auxiliary verb like aller or ir in these examples from Romance languages, one would expect it to appear with participles other than ones of saying and subjects other than the first person, which does not happen until Plato. But even then, ἔρχομαι with a future participle never appears with an inanimate, abstract or concrete subject, which may indicate that it is not a full auxiliary, as G.C. Wakker, ‘Future auxiliaries or not?’, in E. Crespo, J. de la Villa and A.R. Revuelta (edd.), Word Classes and Related Topics in Ancient Greek: Proceedings of the Conference on ‘Greek Syntax and Word Classes’ held in Madrid on 18–21, June 2003 (Louvain, 2006), 237–55, at 253 n. 32 thinks. Finally, while in many languages auxiliaries cannot be separated from their complements (as Wakker, 243–4 describes), in three of the occurrences in Herodotus at least one word separates ἔρχομαι from its future participle (2.35.1 = #4, 2.99.1 = #6 and 7.102.2 = #12, where the intervening element is a particle).

¹⁹ Amenta (n. 18), 90–1 gives two translations for each occurrence of the construction in Herodotus: in brackets, a literal rendition with the sense of movement, ‘[io circa questo non vado a parlare]’, and in italics, a more idiomatic one without it: ‘ma non intendo parlare di questo’.

²⁰ Létoublon, 188 summarizes the development and describes the rhetorical function at 186: ‘Les locutions avec ἔρχομαι / ἤια + participe futur se rencontrent aux articulations du récit: elles fournissent des “transitions” rhétoriques, elles sont le signe du recul pris par Hérodote par rapport aux événements et à leur interprétation, et expriment la conscience qu'il a de ne pas faire de l'histoire “objective”: c'est le début d'un métalangage de l'auteur historien, inscrit dans l’œuvre aux points stratégiques.’

²¹ See P. Payen, Les îles nomades: conquérir et résister dans l’Enquête d'Hérodote (Recherches d'histoire et de sciences sociales 72) (Paris, 1997), 259 on how these verbs embrace historical time and narrative space.

²² σημήνας also participates in the metaphor of the way: at 2.20.1, Herodotus says he will ‘point out’ (σημῆναι) two of the three ‘ways’, or opinions (τριφασίας ὁδούς), that some clever Greeks hold about the Nile, but he himself takes a different path of logos. R.V. Munson, Telling Wonders: Ethnographic and Political Discourse in the Work of Herodotus (Ann Arbor, 2001), 33 n. 32 adduces this passage as an example of how ‘metaphorical and literal terms of the code of narration mix in a striking way’. A. Hollmann, The Master of Signs: Signs and the Interpretation of Signs in Herodotus’ Histories (Hellenic Studies 48) (Cambridge, MA and London, 2011), 27 n. 52 lists twelve occurrences of forms of σημαίνω describing the narrator's activity. Six indicate the direction of the logos (1.5.3, 1.75.1, 2.20.1, 6.39.1, 7.77 and 7.213.3), and three give the measurement of a physical route (2.9.2, 4.99.2 and 5.54.1).

²³ R. Friedman, ‘Location and dislocation in Herodotus’, in C. Dewald and J. Marincola (edd.), The Cambridge Companion to Herodotus (Cambridge and New York, 2006), 165–77, at 166 rightly comments on these verbs, ‘the language of his first statement of authorial method is the language of travel’, but does not discuss ἔρχομαι. Munson (n. 22), 33 (with n. 31) does not analyse the verbs in detail but succinctly explains their significance: ‘This metaphor of the journey creates the illusion of an overlap between the code of narration and the code of historie (where verbs of motion are in most cases used literally). Thus it reinforces our sense of the identity between the narrator and the researcher.’

²⁴ LSJ⁹ s.v. ἐπέξειμι A.III.2. J.E. Powell, A Lexicon to Herodotus (Cambridge, 1938), 129 lists it under ἐπεξέρχομαι. Stein, 8 compares Aesch. PV 870, where Prometheus, having recounted Io's wanderings and foretold her future, says: μακροῦ λόγου δεῖ ταῦτ’ ἐπεξελθεῖν τορῶϲ, ‘a long speech is needed to go through these things clearly’. Prometheus also uses the metaphor of the path of logos as a rhetorical signpost earlier in the same speech (PV 844–5: τὰ λοιπὰ δ’ ὑμῖν τῆιδέ τ’ ἐϲ κοινὸν φράϲω, | ἐϲ ταὐτὸν ἐλθὼν τῶν πάλαι λόγων ἴχνοϲ, ‘and I will declare the rest to you and to her in common, coming to the same track of my previous speech’). Given the context of Io, it is tempting to wonder if Herodotus may have had this passage in mind in 1.5.3. Regardless, Herodotus enhances the metaphor by using ἔρχομαι with a future participle, rather than the future indicative, as Prometheus does.

²⁵ προβαίνω also appears in this sense in a contemporary comedy, Cratinus’ Eumenides, fr. 69 K.-A.: ἐπίϲχεϲ αὐτοῦ· μὴ πέραι προβῆιϲ λόγου, ‘stop there: don't proceed further in your account’ (cited by LSJ⁹ s.v. προβαίνω A.I.3; PCG 4.156 ad loc. gives Hdt. 1.5.3 as a parallel). Artabanus, one of the four speakers in Herodotus’ narrative who employ ἔρχομαι with a future participle of a verb of saying, also uses προβαίνω in close proximity to it, but in a literal sense (7.49.3–4 = #11). Another speaker, Otanes, uses ἐπεξέρχομαι, but not as a metaphor of discourse, shortly before he uses the ἔρχομαι construction (3.80.2; cf. 3.80.5 = #8): εἴδετε μὲν γὰρ τὴν Καμβύσεω ὕβριν ἐπ’ ὅσον ἐπεξῆλθε … τὰ δὲ δὴ μέγιστα ἔρχομαι ἐρέων, ‘for you saw to how great an extent Cambyses’ hubris advanced … But indeed I am going to say the worst things’.

²⁶ Létoublon, 183–4 highlights this example.

²⁷ The image of the path of song (οἴμη) appears in Homer (Od. 8.74, 8.481 and 22.347) and Pi. Ol. 9.47. For a discussion of the term, see A.L. Ford, Homer: The Poetry of the Past (Ithaca and London, 1992), 41–4. The metaphor of the road of words (ὁδὸς λόγων) appears several times in Pindar: Ol. 1.110 and 6.23–4; Pyth. 4.247–8 and 11.38; and Nem. 7.51, on which see R. Nünlist, Poetologische Bildersprache in der frühgriechischen Dichtung (Stuttgart and Leipzig, 1998), 259 and 310 (on Ol. 1), 255 (Ol. 6), 241 (Pyth. 4 and 11) and 253 (Nem. 7). Thomas, 226 also notes the metaphor in Xenophanes, DK 21 B 7.1 and Hippoc. Nat. puer. 20.6 and VM 13.1 (at 243 n. 88). For a general account of the conceptual metaphor of argument as journey from the perspective of cognitive linguistics, see R. Martínez Vázquez and J.M. Jiménez Delgado, Metáfora conceptual y verbo griego antiguo (Zaragoza, 2008), 14–16.

²⁸ Becker (n. 4), 101–38 surveys the metaphor in Herodotus and discusses 1.5.3 at 101–2. Nünlist (n. 27), 239 n. 12 highlights 1.5.3 and Herodotus to show that the image of the way as a transitional formula appears in prose. Purves, 69 and 122–3 explains how Herodotus’ road of logoi is more complex than Homer's path of song.

²⁹ I do not include ἔρχομαι λέξων in 2.11.3, which is bracketed by Hude (n. 1), Legrand (n. 18) and Stein, following J. Schweighaeuser, Herodoti Musae sive Historiarum Libri IX (London, 1830), though retained in the Teubner edition of H.B. Rosén, Herodoti Historiae, vol. 1 (Leipzig, 1987). λέξων does not occur in the narrator's voice with ἔρχομαι, but does in both examples with ἤια (4.82 and 5.62.1 = #13 and #14). Herodotus uses both σημαίνω and φράζω as key verbs in his relationship with his audience, as Hollmann (n. 22), 24–7 and 29–30 shows. Although by Herodotus’ time both verbs were commonly used of speech, their connotations of visual display strengthen the metaphor of the path of logos.

³⁰ Wakker, 178.

³¹ Wakker, 180. First-person singular verbs of saying in the future indicative appear 27 times in the narrator's voice: ἐρέω (three times), φράσω (ten), σημανέω (seven), λέξω (one), δηλώσω (six). The narrator also uses the passive εἰρήσεται three times. Twenty-two of these instances indicate an immediate transition. φήσω, which the narrator uses twice (2.49.2–3), does not have the same function, since it simply introduces a dependent statement, as Wakker, 178 n. 11 notes.

³² Here I have partially adapted A. Lloyd's translation in Asheri et. al., 235.

³³ In the two other cases where a future indicative verb of saying appears in close conjunction with ἔρχομαι with a future participle, it also comes after and signals a smaller transition within the larger one already announced: 3.6.1–2 = #7 (see p. 10 below) and 4.99.2 = #9, where ἔρχομαι σημανέων is followed by ἐγὼ δὲ ἄλλως δηλώσω (4.99.5).

³⁴ I owe this point and the emphasis on the ‘detour’ to one of the referees. It is no coincidence that Herodotus remarks on the accretive tendency of his logos when he tells of a wonder (4.30.1): θωμάζω δέ (προσθήκας γὰρ δή μοι ὁ λόγος ἐξ ἀρχῆς ἐδίζητο) …, ‘But I am amazed—for indeed my logos sought additions from the beginning …’. On this passage, see F. Hartog, The Mirror of Herodotus: The Representation of the Other in the Writing of History, trans. J. Lloyd (Berkeley, Los Angeles and London, 1988), 234 and 344, and Munson (n. 22), 242. Hartog, 236 also comments on the link between the size of the wonder and the length of the narrative in 2.35.1.

³⁵ LSJ⁹ s.v. μηκύνω A.3 defines it thus and lists parallels in contemporary Attic writers.

³⁶ Strictly speaking, ἤια is the imperfect of εἶμι, but it serves as the imperfect of ἔρχομαι, which never appears in prose (Smyth, §774). F. Létoublon, Il allait, pareil à la nuit. Les verbes de mouvement en grec: supplétisme et aspect verbal (Paris, 1985), 59–109 analyses the relationship of ἔρχομαι and εἶμι.

³⁷ Létoublon, 185 cites other Herodotean examples of the metaphor with ἄνειμι/ἐπάνειμι δὲ ἐπὶ τὸν πρότερον λόγον and ἀναβαίνω/καταβαίνω.

³⁸ F.W. Sturz, Lexicon Xenophonteum, vol. 2 (Leipzig, 1801), 366 s.v. ἔρχεσθαι equates them in, among other passages, Xen. An. 7.7.17 (which he cites as 7.7.10: ἐρχόμεθα … βοηθήσοντες) and Ages. 2.7 (see n. 10). Sturz is cited by Dietrich (n. 9), 218. Wakker, 179 n. 13 notes that both ἤια with the future participle and ἔμελλον with the infinitive can be used for interrupted or postponed actions. ἔμελλον occurs only at 3.155.4, in Zopyrus’ speech to Darius. Wakker also points out that ἔμελλον never appears with a verb of saying in Herodotus. This is in contrast to Homer, who does use it with the present infinitive of verbs of saying, e.g. Il. 10.454–5: ἔμελλε … λίσσεσθαι, ‘he was about to entreat’.

³⁹ As Wakker, 178 n. 11 points out, Herodotus’ avoidance of μέλλω to announce what he is going to say immediately afterward is a remarkable contrast to Plato's use of both μέλλω and ἔρχομαι with a future participle to do so.

⁴⁰ I have noted these singularity markers in the Appendix. Though the sample is admittedly small, this may be another difference between the ἔρχομαι construction and the future indicative, since only eight of the narrator's 27 uses of first-person singular verbs of saying are accompanied by a singularity marker: ἐρέω (4.129.1 and 6.43.3), φράσω (2.155.3, 2.156.6 and 5.65.5) and δηλώσω (1.192.1–2, 2.101.1–2 and 4.81.4–5). In addition, two of the narrator's three uses of εἰρήσεται (2.35.1 and 4.16.2) are accompanied by a singularity marker, and two of them occur in close proximity to ἔρχομαι (2.35.1 = #4: see p. 7) or ἤια with the future participle (4.82). As J. Priestley, Herodotus and Hellenistic Culture: Literary Studies in the Reception of the Histories (Oxford and New York, 2014), 61 notes, the future indicative also appears ‘frequently … in his descriptions of marked thōmata’. But a neat illustration of the difference is provided by the contrast between ἐρέω with a singularity marker without a definite article in 4.129.1 (θῶμα μέγιστον ἐρέω, ‘I will tell of a very great wonder’; cf. also 6.43.3) and ἔρχομαι φράσων with the same superlative, but with a definite article, in 1.194.1 = #2 (τὸ … θῶμα μέγιστον … ἔρχομαι φράσων, ‘I am going to tell of … the greatest wonder’). Wakker, 180 notes that this difference suggests that only the future indicative can indicate ‘a kind of characterization of what the narrator is going to tell, rather than an announcement of what he is going to tell’.

⁴¹ Although Lightfoot (n. 10), 289 claims that in Lucian the expression has ‘a portentous significance in general absent from Herodotus’ (with the exception of 2.35.1 and 2.99.1), the singularity markers and the appearance of the construction in the proem and at the beginnings of other sections suggest that every occurrence is important.

⁴² Brock, 4 observes: ‘When he wants actively to draw the audience on, Herodotus frequently uses the carrot of an evaluative adjective or phrase’, each of which in his examples is a superlative (1.60.3, 8.105.1 and 9.37.2). Herodotus launches into the whole Egyptian logos with a superlative, leading the reader in medias res with a singularity marker without the ἔρχομαι expression (2.2.1): οἱ δὲ Αἰγύπτιοι … ἐνόμιζον ἑωυτοὺς πρώτους γενέσθαι πάντων ἀνθρώπων. ‘And the Egyptians … thought that they were the most ancient of all peoples.’

⁴³ I adopt the term ‘the historian's voice’ from Fowler, 69, who at 71 defines one way in which it appears as ‘giving unusual information that implies special knowledge or research’, e.g. with superlatives followed by τῶν ἡμεῖς ἴδμεν. 1.5.3 is the only instance of ἔρχομαι and the future participle connected with τῶν ἡμεῖς ἴδμεν: Herodotus delivers on his promise to tell of the first man he knows (τὸν δὲ οἶδα αὐτὸς πρῶτον) to have committed hostilities against the Greeks with the words οὗτος ὁ Κροῖσος βαρβάρων πρῶτος τῶν ἡμεῖς ἴδμεν …, ‘this Croesus was the first of the barbarians we know of …’ (1.6.2). Although the other uses do not include the phrase, they all underscore Herodotus’ personal knowledge. Dewald, 281–2 argues that ‘these superlatives [with τῶν ἡμεῖς ἴδμεν] express in an abbreviated form the diffidence that Herodotus insists on as a critic’, but on the basis of the polemical tone of 1.5.3 I agree with Fowler that they advertise Herodotus’ efforts.

⁴⁴ In fact, as Hartog (n. 34), 3 stresses, Herodotus notes that Scythia has hardly any wonders (4.82: θωμάσια δὲ ἡ χώρη αὕτη οὐκ ἔχει, χωρὶς ἢ …), in contrast to Egypt (2.35.1 = #4), which may partially explain the absence of a singularity marker in 4.99.2 = #9.

⁴⁵ Herodotus also uses future indicative verbs of saying in travel contexts. Although he does not explicitly mention his travels in connection with all of the instances of the ἔρχομαι construction in his voice, his statements about his autopsy or its limits suggest that the expression does introduce descriptions of places and phenomena that he claims to have seen or learned about abroad. Marincola, J., ‘Herodotean narrative and the narrator's presence’, Arethusa 20 (1987), 121–36Google Scholar, at 122–3 n. 5 lists the passages in which Herodotus claims or denies autopsy, mentions personal contact with a specific source or makes general remarks about his historiē (2.99.1 and 2.147.1).

I will go briefly through the passages in the Appendix to show the travel context, besides #1, which I discuss above. Herodotus does not tell us if he went to Babylon (#2), but Asheri et al., 197 think there is no reason to doubt that he did. By stressing the fact that he did not see the statue in the temple there (1.183.3), Herodotus may imply that he did see the other features he describes: compare how his other denials of autopsy appear to enhance the credibility of his account by defining the limits of his efforts (e.g. 2.34.1, 4.16.1–2 and 4.25.1–2). In the Egyptian logos, while it is not clear if he is claiming to have personally visited the gulf of Arabia (#3), he describes it to support his own observation and the priests’ report about the deposition of sediment by the Nile (2.10.1 and 2.12.1). Similarly, #4 and #5 are in the part of the logos that he attributes to his own autopsy, judgement and inquiry (#6). While in #7 he does not say outright that he noticed the absence of wine-jars in Egypt on his travels, his use of ἰδέσθαι may indicate that it is based on his eyewitness experience. Finally, his knowledge of the coast of Scythia in #9 may be connected with his claim of autopsy of that region in 4.81.1–2 (and possibly also in 4.86.4, though there it is uncertain if he is saying that he personally measured the Pontus, as Marincola, 123 n. 5 explains). As Purves, 142 points out, almost half of the θωμ- words in Herodotus occur with sights he claims to have seen himself, as may be the case in one of their two appearances with the ἔρχομαι expression (#2) and surely is in the other (#4).

⁴⁶ Brock, 14 cites this passage as an example of Herodotus’ attention to his audience.

⁴⁷ Munson, R.V., ‘Herodotus’ use of prospective sentences and the story of Rhampsinitus and the thief in the Histories ’, AJPh 114 (1993), 27–44 Google Scholar, at 36 notes that this is the only example of an open-ended direct question in Herodotus and that it is a common technique used by oral folk-narrators to connect parts of a story.

⁴⁸ Darius refutes Otanes’ proposal because it is contrary to their ancestors’ practice (3.82.5), Miltiades opposes the prevailing opinion among the Athenian generals (6.109.2), Artabanus had spoken freely to Xerxes previously (7.46.1) and had been rebuked (7.11.1), and Demaratus asks permission to speak the truth to the king (7.101.3).

⁴⁹ Here I consider the singularity marker to be this emphatic address to the one person in particular who has the power to determine the decision (τὸ κῦρος). V. Zali, The Shape of Herodotean Rhetoric: A Study of the Speeches in Herodotus’ Histories with Special Attention to Books 5–9 (International Studies in the History of Rhetoric 6) (Leiden and Boston, 2015), 265, with n. 111, well notes the ‘recurrent use’ of the second-person singular pronoun in Miltiades’ address to Callimachus (6.109.3, 4 and 6).

⁵⁰ As in the narrator's voice, so too in other speakers’ mouths the ἔρχομαι expression is more closely associated with singularity markers than are future indicative verbs of saying. Of the ten uses of first-person future verbs of saying and one of εἰρήσεται by characters within the narrative, four are accompanied by a singularity marker: ἐρέομεν at 4.118.4 (Scythian messengers), φράσω at 5.111.3 (Onesilus of Salamis’ squire), εἰρήσεται at 6.86δ (Leotychidas) and ἐρέω at 9.42.2–3 (Mardonius).

⁵¹ All except Otanes appear on the list in R. Lattimore, ‘The wise adviser in Herodotus’, CPh 34 (1939), 24–35. J. Gould, Herodotus (London, 1989), 78 and Thomas, 285 note the connection between wise advisers and the narrator.

⁵² C. Scardino, Gestaltung und Funktion der Reden bei Herodot und Thukydides (Beiträge zur Altertumskunde 250) (Berlin and New York, 2007), 350 and 752 stresses the interpretative function of speeches for the external audience.

⁵³ As the proem declares, ‘great and wondrous deeds’ (ἔργα μεγάλα τε καὶ θωμαστά) are one of the main subjects of the Histories. The frequency of θωμ- words (94 instances) shows their importance, on which see Munson (n. 22), 232–65 and Priestley (n. 40), 51–108.

⁵⁴ When Herodotus introduces the constitutional debate, he declares (3.80.1): ἐλέχθησαν λόγοι ἄπιστοι μὲν ἐνίοισι Ἑλλήνων, ἐλέχθησαν δ’ ὦν, ‘speeches were given that are unbelievable to some Greeks, but in fact they were given’. He defends the historicity of Otanes’ speech in particular by analogy with Mardonius’ actions in 492 (6.43.3): μέγιστον θῶμα ἐρέω τοῖσι μὴ ἀποδεκομένοισι Ἑλλήνων Περσέων τοῖσι ἑπτὰ Ὀτάνεα γνώμην ἀποδέξασθαι ὡς χρεὸν εἴη δημοκρατέεσθαι Πέρσας· τοὺς γὰρ τυράννους τῶν Ἰώνων καταπαύσας πάντας ὁ Μαρδόνιος δημοκρατίας κατίστα ἐς τὰς πόλιας. ‘I will tell of a very great wonder to those Greeks who do not believe that Otanes declared his opinion to the seven Persians that the Persians should have a democratic government: Mardonius in fact deposed all the tyrants of the Ionians and set up democracies in their cities.’

⁵⁵ This takes a step further the common observation that in this debate ‘non-Greeks are speaking and thinking in Greek-like ways’, as C. Pelling, ‘Speech and action: Herodotus’ debate on the constitutions’, PCPhS 48 (2002), 123–58, at 124 puts it.

⁵⁶ 3.80.2. Asheri et al., 473 note that the expression is common in Herodotus, and a TLG search shows that ἐμοὶ/μοι δοκέει or (ὡς) ἐμοὶ/μοι δοκέειν appear 22 times. Of the four occurrences of ἐμοὶ/μοι ἐδοκέε, one is in the voice of another character (the fisherman to Polycrates, 3.42.2), as is one of the six of ὡς ἐγὼ δοκέω (Xerxes to Demaratus, 7.101.2).

⁵⁷ See Lateiner (n. 14), 163–86 for many examples. As he notes at 171, ‘Herodotus does not categorically condemn one-man rule’, and one should evaluate the tyrants case by case.

⁵⁸ Roy, C.S., ‘The constitutional debate: Herodotus’ exploration of good government’, Histos 6 (2012), 298–320 Google Scholar, at 301 notes this correspondence.

⁵⁹ R.W. Macan, Herodotus: The Fourth, Fifth, and Sixth Books, with introduction, notes, appendices, indices, maps, vol. 1 (New York, 1973; orig. London, 1895), 366.

⁶⁰ Macan (n. 59), 366.

⁶¹ For Artabanus as an Herodotus-like figure, see Grethlein, J., ‘How not to do history: Xerxes in Herodotus’ Histories ’, AJPh 130 (2009), 195–218 Google Scholar, at 214 with additional bibliography at 199 n. 15.

⁶² Thomas, 109–11 analyses the debate from this perspective, and D. Branscome, Textual Rivals: Self-Presentation in Herodotus’ Histories (Ann Arbor, 2013), 54–104 discusses the parallels between Demaratus and Herodotus. On the figure of the ‘tragic warner’, see H. Bischoff, Der Warner bei Herodot (Diss., Marburg, 1932) and Lattimore (n. 51), 24–6.

⁶³ Herodotus affirms Demaratus’ prediction by describing the boldness of the Spartan ambassadors Sperthias and Boulis as worthy of wonder (7.135.1: θώματος ἀξίη) and by praising the Lacedaemonians’ courage in a notorious counterfactual scenario (7.139.3). Although he then envisions an alternative in which they might have medized (7.139.4), the narrative vindicates Demaratus’ view, since they did end up fighting. A Persian horseman marvels at the Lacedaemonians exercising and combing their hair before the battle of Thermopylae (7.208.3: ἐθώμαζε), which Xerxes finds so incredible that he summons Demaratus again (7.209.1).

⁶⁴ This may be emphasized by the fact that Solon, the archetypal wise adviser, does not use the ἔρχομαι expression even though both the narrator (1.29.1 and 1.30.1) and Croesus (1.30.2) stress his travels.

⁶⁵ Although Hecataeus treats Egypt in his Periegesis, in the extant fragments there is no evidence of first-hand observation, as West, S., ‘Herodotus’ portrait of Hecataeus’, JHS 111 (1991), 144–60CrossRefGoogle Scholar, at 152 n. 46 notes. She also points out that Hdt. 2.143 is the only evidence that Hecataeus ever left Ionia, except for Agathemerus’ description of him as ‘a much-wandering man’ (BNJ 1 T 12a: ἀνὴρ πολυπλανής), which may be derived from the Herodotean passage.

⁶⁶ Cf. R.L. Fowler, ‘Early historiē and literacy’, in The Historian's Craft (n. 5), 95–115, at 108–9. This is a trait that Herodotus shares with earlier philosophers and contemporary medical writers, as Thomas, 235–48 and 257–69 shows.

⁶⁷ As W. Rösler, ‘The Histories and writing’, in Brill's Companion (n. 1), 79–94, at 89 points out, Herodotus uses the metaphor of λέγειν for the message of a written text (e.g. 1.124.1 and 3.40.1) and uses λέγειν and γράφειν more or less interchangeably. γράφω occurs in the first person at 1.95.1, 2.70.1, 2.123.1 and 2.123.3, 4.195.2, 6.53.1 and 7.214.3, συγγράφω at 3.103 and οὐκ ἔχω … συγγράψαι at 6.14.1. Cf. also 1.93.1 (ἐς συγγραφήν) and 4.36.2 (ἐς γραφήν). The last passage, in which Herodotus ridicules others’ attempts to draw maps, alludes to the opening of Hecataeus’ Genealogies, which also juxtaposes a verb of saying with γράφω (BNJ 1 F 1a: ῾Εκαταῖος Μιλήσιος ὧδε μυθεῖται· τάδε γράφω, ‘Hecataeus of Miletus speaks as follows: I write these things’), on which see L. Bertelli, ‘Hecataeus: from genealogy to historiography’, in Historian's Craft (n. 5), 67–94, at 80–4 and Fowler (n. 66), 110–11. The allusion to Hecataeus in 4.36.2 is noted by D. Boedeker, ‘Herodotus’ genres’, in M. Depew and D. Obbink (edd.), Matrices of Genre: Authors, Canons, and Society (Cambridge, MA, 2000), 97–114, at 107, cited by Purves, 128 n. 24.

⁶⁸ I am grateful to one of the referees for suggesting I sharpen this point. The deictic pronoun οὗτος and the temporal adverbs νῦν and ἤδη, which occur in close conjunction with the ἔρχομαι expression in eleven of the twelve examples (all but 3.80.5 = #8) and in all eight of the narrator's uses, provide further direction. For this understanding of deixis in Herodotus, see E.J. Bakker, ‘The making of history: Herodotus’ historiēs apodexis’, in Brill's Companion (n. 1), 3–32, at 30 and id., ‘The syntax of historiē: how Herodotus writes’, in Cambridge Companion (n. 23), 92–102, at 95–101. I have marked deictics and temporal adverbs in the Appendix. It should be stressed that deictics are ubiquitous in Herodotus and not unique to the ἔρχομαι construction.

⁶⁹ Dionysius’ use of compounds of βαίνω and ἔρχομαι in his praise of Herodotus (Pomp. 3.8 and 3.14) may reflect Herodotus’ own practice, even if unconsciously, since the vocabulary was standard (e.g. Plut. de Malig. 3 and 22 = Mor. 855c and 859e uses the metaphor of the path of discourse in his attack on Herodotus). For the importance of audience engagement to Dionysius’ conception of history, see C.C. de Jonge, Between Grammar and Rhetoric: Dionysius of Halicarnassus on Language, Linguistics and Literature (Mnemosyne Supplements 301) (Leiden and Boston, 2008), 356–61 and N. Wiater, The Ideology of Classicism: Language, History, and Identity in Dionysius of Halicarnassus (Untersuchungen zur antiken Literatur und Geschichte 105) (Berlin and New York, 2011), 132–44.

⁷⁰ Marincola (n. 45), 122–3 and 133.

⁷¹ Luraghi (n. 11), 156.

⁷² The absence of the ἔρχομαι construction from the narrator's voice after Book 4 may be seen as evidence of the more focussed narrative of the Persian Wars, since Herodotus uses the expression to justify detours and transitions, which occur more frequently in the earlier books. Thus it is no surprise that 20 of the 27 first-person future indicative verbs of saying appear in Books 1–4. Brock, 14 observes: ‘The density of many of the signposting features thins out in the later books as the narrative line becomes more straightforward and robust’.

⁷³ Fowler, 69 with n. 61 discusses the forceful beginnings of other fifth-century prose works.