As Herodotus explores how much and in which ways human societies are mutually different, he takes for granted the generally recognized subdivision of the world into Greeks and barbarians. [1] But this common-view discourse provides only a partial framework of explanation for the text of the Histories. [2] The language of the Greek-barbarian antithesis stands in a dialogic relationship with a more complex and pluralistic opinion of the world. [3] This other voice never entirely silences the first, but it emerges as an insistent challenge in the flow of evidence throughout Herodotus’ work.
Whether supported or not in any given context, the very notion of “barbarian” testifies to a criterion of difference that is to a great extent linguistic. Beyond any internal linguistic differentiation, Greek perception defines barbarian speech as anomalous and hard to understand. [4] The word barbaros itself may or may not have initially encoded the idea of indistinct or “funny” sound, though it eventually comes to do so. [5] Our earliest literary sources, at any rate, contain references to language (or languages) different from that of the text even before they clearly formulate the notion of non-Greek in ethnic or cultural terms. [6] The Iliad identifies “foreign” with “foreign-speaking” when it calls the Carians barbarophōnoi (Iliad 2.867). [7] The leaders of these Carians, moreover, are said to be the sons of an individual, now dead, who “used to go to battle wearing gold, like a girl” (Iliad 2.872–873). This man is qualified as a fool, with a term that has a linguistic dimension: nēpios literally means ‘infantile’, like a baby who has not reached speaking age. [8]
Variations on this stereotype, which combines language difference with other negative features, become more pervasive and deliberate in later texts, especially with the politicization of the Greek-barbarian antithesis during and after the Persian Wars. [9] Foreign speech is almost invariably an index of primitivism, uncouthness, intellectual or cultural inferiority, irrationality, or madness. [10] Though all three tragedians make numerous references to the strangeness of foreign speech, Aeschylus has the greatest number in plays that have survived whole. [11] The Persians features exotic lamentations, the mention of Persian “noise,” and catalogues of proper names that seek to capture the bewildering sounds of a strange language. [12] In the Suppliants the Danaids refer to their own foreign way of speaking, and the Egyptian herald provides a demonstration of un-Greek phrasing. [13] Fifth-century comedy imitates barbarian gibberish or barbarized Greek. [14] Several passages in various texts compare barbarian speech to the voice of animals, especially birds. [15] Pindar associates the adjective barbaros with pali nglōssos ‘with the tongue put on backwards’ (Isthmian 2.24), and Sophocles’ Trachinian Women replaces the antithesis “Greek and barbarian” with “Greek and aglōssos,” ‘tongueless’ (1060).
If language comes to represent a mark of Greekness and Greek superiority, [16] and if the issue of cultural differentiation is central to the Histories, what views does this text communicate about Greek as a language and the various barbarian languages? [17] Does the role Herodotus attributes to language reinforce or undermine the authoritative Greek-barbarian antithesis of contemporary thought? The Histories resound with many different discourses in the voices of the narrator-historian, the narrator-ethnographer, characters, ethnographic subjects, and sources. Everyone speaks Greek, but the text provides irregular indications that in the world of the story they did not. Narratives of past events mention linguistic discrepancies among the speakers or give us glimpses of a community’s speaking style. Ethnographic descriptions include the language of this or that people along with other cultural traits. The narrator discusses the reality of language difference for a traveler in distant lands and shows what non-standard speech sounds like when he explains foreign names or introduces native terms.
These different passages do not create a monolithic representation of the importance of language differentiation; they are sometimes hard to interpret, mutually contradictory, answering or even neutralizing one another. The linguistic discrepancies recorded in the text are part of a broader cultural and ideological heteroglossia at different levels of discourse. [18] We can separate the voice of the histōr from the speeches and traditions he reports, but only up to a point. [19] The primary narrator sometimes appears to slide along the ideological spectrum in response to his sources, material and audiences. On the whole, however, the Histories display a view of language that is internally coherent and consistent with Herodotus’ message on other aspects of culture.
Within the ideological landscape of the Greek-barbarian antithesis, Herodotus paradoxically distinguishes himself for being both uniquely philobarbaros and uniquely Greek. [20] He appears uniquely Greek in the sense that even among fifth-century Greek texts his work professes an extraordinary degree of commitment to freedom and democratic equality. The Histories thereby celebrate the superiority of a culture that at a certain point in its development chose to implement various forms of broad-based constitutional government. [21] But Herodotus is also philobarbaros. His political egalitarianism is grafted onto the Ionian tradition of objective description of foreign peoples and shares in the broad-minded outlook of the sophists and it therefore extends horizontally beyond the boundaries of the Greek world. [22] In his mythical reconstruction of the origin of civil society, Plato’s Protagoras attributes the “civic art” not to Athenians or Greeks but rather to “all men.” [23] Similarly in Herodotus, just as the citizens of the Greek polis share in political decisions, and the various poleis are, or deserve to be, autonomous, so also “all men” are competent to pursue knowledge and regulate themselves. [24] Or, to reverse this principle in the direction of the ethnographer’s work, a study of the nomoi (‘customs’) of different societies will display in most cases the equal competence of all men. This view appears almost closer to our modern American pluralistic ideology than to the most commonplace attitude of fifth-century Greeks. One of the central tasks of Herodotus’ work is in fact to promote the paradox that the uniqueness of Greek values also entails respect for the equal worth of non-Greeks. [25]
The overarching thesis of this book is that the issue of language provides Herodotus with special opportunities to instruct his audiences. In Chapter One we begin by examining Herodotus’ attitude to the Greek linguistic community, his notions about the boundaries that separate it from non-Greeks, and his view of its internal differentiation both diachronically (different origins of the different groups of Greeks) and synchronically (different dialects). Chapter Two will turn to the ways in which the ethnographer Herodotus deals with languages not his own as a special trait of the foreign cultures he describes and as a factor to be reckoned with in the course of “field work.” Perhaps our most precious passages from the point of view of Herodotus’ ideology of language are the passages where Herodotus asserts his authority as interpreter by translating foreign words: these “metanarrative glosses,” scattered through both ethnographic descriptions and historical sections in the Histories, are the special focus of Chapter Three. [26] Finally, in Chapter Four, we shall examine the role of language difference in several narratives that explicitly mention the issue as an element of the plot. The preponderance of all this evidence in the Histories confirms Herodotus’ commitment to the notion of the equal competence of foreign societies in various spheres. It also shows that language difference is not a serious obstacle to understanding others, enhances rather than hampers learning about foreign cultures, and helps people to understand their own culture. Different aspects or mediums of communication provide abundant opportunities for uncertainty and error. [27] But language in the strictest sense is to Herodotus an area of objective and interesting difference (and unexpected similarities) that also turns out to be relatively unproblematic. It therefore represents a good model for coming to terms with other more emotionally charged features of the barbarian world.


[ back ] 1. The word barbaros occurs about 300 times in the Histories; see Powell 1938, s.v. Herodotus’ uses are surveyed by Laurot 1981; see also Payen 1997.176.
[ back ] 2. Cartledge 1993.356–362. Pelling 1997.
[ back ] 3. In Bakhtinian terms, this prevalent, homogenizing, ideology would be “centripetal,” in opposition to the “centrifugal” voice of dissent. Bakhtin 1981.270–308. The Bakhtinian notions of “dialogism” and “heteroglossia” in the novel have been applied to Herodotus by Pelling 2000.83, and to Homer by Peradotto 1990, esp. 53–58. For Bakhtin and the classics, see also the essays in Peradotto, et al. 1993 and Bracht Branham 2002.
[ back ] 4. E.g. Sophocles Ajax 1262–1263. For other examples of “barbarian” as a generalized language (none in Herodotus), see Harrison 1998, ch. 3. “The Greek Conceptualization of Foreign Languages.” For unintelligibility, see e.g. Plato Protagoras 341c, where Lesbian is said to be a “barbarian language” because hard to understand, and Thucydides 3.94.5 on the speech of the Eurytanes of Aetolia.
[ back ] 5. Strabo 14.2.28. The term’s possible Sumero-Babylonian derivation (Weidner 1913) does not preclude the onomatopeia ba-ba. Chantraine 1968–1980.164–165; Levy 1984; E. Hall 1989.4; Moggi 1991.36; Coleman 1997.178. But see the objections of J. Hall 1995 and 2002:111–112.
[ back ] 6. The Iliad refers to language difference only within the Trojan force (2.804; 4.436). Lejeune 1940–1948.52. Rotolo 1972.398. In the Odyssey, which displays a new preoccupation with cultural distinction between the “real world” of heroic epic and faraway lands, we find the epithets allothrooi (ἁλλόθροοι, 1.183, used by Athena/Mente of peoples to whose lands she has sailed; 3.302, of the Egyptians, where Menelaus traveled), and agriophōnoi (ἀγριόφωνοι, 8.294, of the Sinti of Lemnos). See also Odyssey 19.172–177 on the many languages (and dialects) spoken in Crete. Homeric Hymn to Apollo 3.162–169 mentions the existence of several phōnai, which could be either Greek dialects or foreign languages. Homeric Hymn to Aphrodite 5.113–114 mentions Trojan and Phrygian as languages different from one another and (presumably) from Greek.
[ back ] 7. Werner 1989; Lejeune 1940–48.52; Levy 1984; Colvin 1999.41–50; J. Hall 2002.112. The Homeric epithet barbarophōnoi may mean “speaking bad Greek” (Strabo 14.2.28), but see Herodotus 8.135, where the Carian language is not apparently understood or recognized by Greek speakers. In other respects Herodotus (1.146, 171; 2.152, 154, 163; 3.11) and other sources agree in representing the Carians as culturally very similar to the Greeks.
[ back ] 8. Cf. Iliad 3.237–238, where Nestor rebukes the Achaeans as being babies (νηπιάχοις) and speaking like children (παισὶν ἐοικότες ἀγοράασθε). On this passage, see Martin 1989.104.
[ back ] 9. On the evolution of the Greek-barbarian polarity, see Diller 1961, esp. 39–68; Campos Daroca 1992.27–30; Cartledge 1993, esp. 36–62; Nippel 2002.279–293; Coleman 1997.186–194; J. Hall 1997.45–47; 2002.174–181; Rosivach 1999; Tuplin 1999.55–57. For early references to foreign languages, see e.g. Anacreon fr. 423 Page; Hipponax’ frr. 3a, 92 West. Farina 1963.25–29. The word barbaros connotes intellectual inferiority already in Heraclitus DK I6, 22, B107.
[ back ] 10. Long 1984.130–131; E. Hall 1989.3–5. For barbarophōnos applied to the Persians in a derogatory context, see the oracles in Herodotus 8.20.2 and 9.43.3 (‘barbarian-voiced scream’: βαρβαρόφωνον ἰυγήν).
[ back ] 11. See especially Bacon 1961.15–24, 64–73, 115–120, who usefully surveys all references to foreign speaking in the three tragedians.
[ back ] 12. Broadhead; Bacon 1961.23; E. Hall 1989.76–79; Colvin 1999.77–78, 85. Aeschylus Persians 20–58, 302–330, 402, 958–972, 981–984, 994–1001; at 634, βάρβαρα σαφηνῆ ἡιέντος ‘speaking in clear barbarian language’ (uttered by a Persian) comes across as ironic.
[ back ] 13. Bacon 1961.15; Suppl. 117–119, 127–129, 825–902, 914, 972–974. Cf. Aeschylus Agamemnon 1059–1061; 1062–1063, esp. 1050–1053 (ἀγνέωτα φωνήν of Cassandra); Sophocles Trachinian Women 1060; Antigone 1001–1002; Ajax 1262–1263; Euripides Orestes 1369 ff.; fr. 139 N2; Phoenician Women 1302–1303, Bacchae 158–159; Rhesus 294–297. Aristophanes Birds 199–200; Frogs 679–682; Women at the Thesmophoria 1082–1135, 1176–1226.
[ back ] 14. Long 1984; E. Hall 1989.17–21; Colvin 1999.287–295. See especially the speech of the Persian “Eye of the King” in Acharnians 100 and of the Triballian god in Birds 1615, 1628–1629 discussed with references by Long 1984.134–135 and Colvin 1999.288–290. Elsewhere in comedy (e.g. the Scythian archer in Women at the Thesmophoria 1678–1679) the barbarian speaks bad Greek, as does the Persian prisoner in Timotheus’ nome Persians (158–173). Colvin 1999.55–56. See below, p. 70 and note 10.
[ back ] 15. E.g. Aeschylus Agamemnon 1050 (barbarian speech of Cassandra like the voice of swallows); Seven Against Thebes 463–464 (horses blowing a barbarian noise); Sophocles Antigone 1001 (birds with barbarian cry); Aristophanes Birds 199–200; Schol. Birds 1680 (Triballian god sounds like swallow); Frogs 679–683 (swallow on lips of Cleophon). For a later example, see the language of the satyr in Plutarch Sulla 27.2. Other passages in Harrison 1998, note 80; Moggi 1991.38–39.
[ back ] 16. E. Hall 1989.4. Contra J. Hall 1995.93–95.
[ back ] 17. For recent discussions of foreign languages in Herodotus, see Harrison 1999; Chamberlain 1999; Campos Daroca 1992; Colvin 1999.57–61.
[ back ] 18. For Bakhtin’s term heteroglossia see note 3, above.
[ back ] 19. Dewald 1987, esp. 153, was the first to call histōr the narrator and implied author who emerges from the Histories as opposed to the real author (on these distinctions, see Booth 1983.67–77; Chatman 1978.147–158; Genette 1980.213–214). Though never used by Herodotus, the term is especially apt and now accepted by many modern critics. Besides referring to the practice of historiē, it carries a juridical meaning that is also appropriate to Herodotus’ authorial persona and activity. See especially Nagy 1990.262, 315–320; also Connor 1993.
[ back ] 20. For Herodotus philobarbaros, see Plutarch Malice of Herodotus 12–14 = Moralia 857A–F.
[ back ] 21. See especially 5.78. Loraux 1986.205. The notion of Greek equality is sometimes expressed in terms of the Greeks’ adoption of a constitutional order and rejection of autocratic rule; e.g. Hippocrates Airs, Waters, Places 23; Aristotle Politics 1237b.23–34. For the origins and character of Greek ideas of freedom and equality and their limitations from a modern viewpoint, see Morris 1996, Ostwald 1996, Hansen 1996, Raaflaub 1996, Roberts 1966.
[ back ] 22. See, e.g 2.3.2; 3.38. For the objective aspect of early Greek ethnography, see Fornara 1983.12; Nippel 2002.278, 282–287. For Herodotus’ link with the sophistic and scientific thought of his time, see Thomas 2000.
[ back ] 23. Plato Protagoras 320c–323c. Roberts 1996.188. Herodotus (in a passage possibly derived from Protagoras) insists on Persian awareness of the nature of different forms of government, including democracy (3.80–83; 6), while at the same time noticing the imperfect fulfillment of the principles of freedom and democratic equality among the Greeks. On the connection between Herodotus and Protagoras, see Thomas 2000.125–127, 147–149, Munson 2001.165–166, 206–207.
[ back ] 24. Herodotus is also explicit in assuming the equality (ἴσον) of all men in the sphere of religious beliefs at 2.3.1.
[ back ] 25. 3.38. Munson 2001.168–172.
[ back ] 26. According to my definition, a gloss is a metanarrative statement in the voice of the narrator explaining some element of the text. See Munson 2001.32–44 for different types.
[ back ] 27. Though Herodotus is an expert interpreter of all types of codes of information. See especially Hollmann 1998 and Hollmann, forthcoming. I am grateful to the author for providing me with a typescript of the latter work.