Black Doves Speak: Herodotus and the Languages of Barbarians

  Munson, Rosaria Vignolo. 2005. Black Doves Speak: Herodotus and the Languages of Barbarians. Hellenic Studies Series 9. Washington, DC: Center for Hellenic Studies.

1. Greek Speakers

Greeks and Pelasgians

In his narrative of Miltiades’ conquest of Lemnos (6.137–140), Herodotus begins by reporting how at that time Pelasgians were occupying the island. In the heroic age these Pelasgians used to live in Attica but the Athenians expelled them, “either justly or unjustly,” depending on the source one believes. According to Hecataeus, the Athenians had allowed the Pelasgians to inhabit an infertile piece of Attica below Hymettus in exchange for their building the wall around the Acropolis; but when the Pelasgians transformed the land by cultivating it, the Athenians became envious and drove them out. The Athenians themselves, by contrast, claim that they expelled the Pelasgians because they had violated the Athenian women and planned an attack against Athens (6.137). These are the two versions of the story, concludes the narrator, and he declines to choose between them. [1]

From Attica the Pelasgians went to Lemnos, where they sought revenge from the Athenians for depriving them of their previous home. They raided Brauron during the festival and abducted a number of Athenian women whom they kept as concubines. The children born from these unions learned from their mothers “the Attic language and Athenian customs” (γλῶσσάν τε τὴν Ἀττικὴν καὶ τρόπους τοὺς Ἀθηναίους), formed a tight group among themselves, and soon “felt entitled to rule the other children and were much stronger” (ἄρχειν τε τῶν παίδων ἐδικαίευν καὶ πολλῷ ἐπεκράτευν. Cf. ἄρχειν πειρῴατο, 6.138.3). This worried the Pelasgian men, who then slaughtered the children together with their Attic mothers (6.138). Here the narrator explains that this crime and an earlier symmetrical female one—when the Lemnian women murdered their husbands—have become proverbial, so that “people throughout Greece term all atrocious deeds ‘Lemnian’” (6.138.4). The murderous Pelasgians were subsequently struck by a curse and received a Delphic prescription that they pay the requisite penalty to the Athenians. Many years later this came to pass when they were forced to turn their island over to Miltiades (6.139–140).

Pelasgians and barbarians

As they become Greek by learning Greek, barbarians enrich Greek culture, including the linguistic field. [19] Because the Pelasgians in Herodotus represent the collective embodiment of what links the Greeks to the non-Greek world, they are especially responsible for fundamental contributions at the intersection of language and religion. In another innovative passage, Herodotus declares that the Pelasgians learned from the Egyptians the “names” (ounomata) of most of the gods, devised a few of them themselves, and transmitted them to the Greeks. [20] This means, at the very least, that the Pelasgians learned to distinguish different divine figures by their functions and attributes and then gave each of these figures a name. [21] Herodotus, however, does not make clear in what linguistic form the Greeks received the Egyptian divine names via the Pelasgian intermediaries. [22] Were these the names the Egyptians still currently use, although their Greek counterparts now differ from them? Herodotus himself elsewhere says that the Egyptians call Apollo Horus, Osiris is Dionysus in Greek, and so on. [23] If the Pelasgians received these Egyptian names, did the Greeks eventually change or “translate” them into Greek? This could have happened as part of the codification effected by Homer and Hesiod, which included, Herodotus says, adopting new divine “names” (epōnumias). [24] Or are the current Egyptian names Herodotus records perhaps alternative or later forms, while the originals that spread to Greece resembled the names by which the Greeks called their gods in Herodotus’ time? [25]

Herodotus’ lack of specificity on the form of the names at 2.50 reveals that for him the most concrete linguistic dimension of the barbarian religious contribution the Pelasgians transmitted to Greek culture is here of secondary importance. [26] By the same token, this section also confuses the issue of what language the Pelasgians themselves spoke at the time they adopted divine names. Herodotus says that before learning (mostly from the Egyptians) about individual figures of gods, the Pelasgians conceived of them as a collectivity and called them theoi because divine beings “put things in order” (κόσμῳ θέντες, 2.52.1). Now, this is a competent word, one that testifies to a degree of philosophical speculation about the nature of things. [27] Moreover it looks like Greek, and its form and etymology point to resemblances between a barbarian language and Greek. [28] Alternatively, if this is supposed to represent a genuine Greek word, it indicates that the Pelasgians’ native tongue was at this point Greek, even though Herodotus still calls them “Pelasgians.” [29] In the discussion on ethnic origins, as we have already seen, Herodotus is deliberate in identifying the Pelasgians as original speakers of a barbaros glōssa ‘barbarian language’ (1.57.3) who ceased to be Pelasgian and became Greek once they learned Greek. Here, by exploiting the instability of the tradition about the ethnic identity of the Pelasgians, he makes them into a benevolent tertium quid that has bridged the gap between the Greek and the barbarian world. [30] In ancient Dodona, the center where cross-cultural trends join, the Pelasgians are equally at home in the antitheses “Pelasgians and Barbarians” and “Pelasgians and Greeks”:

When the Pelasgians asked at Dodona if they could adopt the names of the gods that came from barbarians, the oracle bid them to use them … The Greeks later inherited these names from the Pelasgians.


Greek language and languages

Barbarians and strangers

The Athenians refuse to negotiate with the Persians, but they need help for the resistance and repeatedly urge the reluctant Spartans to send it. In their first speech they proclaim they will never make peace with the invader for two reasons: they need to avenge the harm he has done to their polis, and they remain loyal to the common heritage of the Hellenes—defined in the passage quoted above (8.144). In a second speech, they complain that by their delay the Spartans are betraying the Greeks and recall how advantageous the terms of the Persian peace would be for themselves (9.6; 7α.1–2). Finally, in the face of further Spartan stonewalling, the Athenians intimate in a third speech that they are ready to become the allies of the King and follow him wherever he might lead (9.11.1–2).

This crescendo of Athenian appeals, protests, and threats never meets with an articulate response. But the Spartans, in the meantime, realize what danger they are in and take action in secret. At the last minute they inform the exasperated Athenians that their army is already on its way to central Greece, marching, as they say, “against the foreigners” (ἐπὶ τοὺς ξείνους, 9.11.2). Here Herodotus translates: “Foreigners (ξείνους) is what they called the barbarians” (9.11.3).


[ back ] 1. For their origin, see Bertelli 2001.87–88. I follow the edition of the Histories by Hude.

[ back ] 2. Sourvinou-Inwood 2003.136.

[ back ] 3. Cf. 4.145.2. The Athenian claim to moral superiority at 6.137.4 in turn recalls claims to superior justice in the exercise of power that we find in fifth-century Athenian diplomatic discourse; see Thucydides 1.75.1–4.

[ back ] 4. The reference to the story of the ethnic dominance of the children of the Attic women in Lemnos seems to allude to Pericles’ citizen law of 451 BC, which prescribed Athenian descent also from the mother’s side. Cartledge 1993.21, 26; Dewald 1998.692. On Pericles’ citizenship law, see esp. Cartledge 1993.73.

[ back ] 5. See J. Hall 1987.62, 1997.41–51; 2002: 25–29, 82. Malkin 1994.41–42.

[ back ] 6. On the terms ethnos and genos, see Walbank 2002.24; Jones 1996; J. Hall 1997.35–36. Barbarians/Pelasgians are the original inhabitants of Hellas also at 2.52–56 (Hellas as a whole), 1.146.1, 2.171, 7.94, (Arcadia/Peloponnese), 2.51.2, 7.94 (Attica); 2.52 (Dodona), 7.42, 95 (Aeolis).

[ back ] 7. On the wide spectrum of views concerning the barbarian or Greek ethnicity of the Pelasgians in ancient authors after Homer, see Sourvinou-Inwood 2003.117–121.

[ back ] 8. This is Herodotus’ original contribution to the “Pelasgian theory” of the ancients, (Myres 1907.197–203) and one of the few investigations in Herodotus that are linguistic in the proper sense of the word (Campos Daroca 1992.49).

[ back ] 9. I agree with the interpretation of 1.58 by Sourvinou-Inwood 2003.122–124, but the passage is ambiguous. Cf. McNeal 1985.17–20; Lévy 1991.220.

[ back ] 10. See especially Sourvinou-Inwood 2003. At the factual level, see Myres 1907.192, 201. Myres (who is attempting to extract from ancient sources historical information on the Pelasgians) regards the Pelasgians “who settled below Hymettus” of Herodotus 6.137.2 as a wave of (Hellespontine) Pelasgians that came later, when the Pelasgians of Attica had already been Hellenized. This is to some extent supported by 2.51, which talks about “Pelasgians who became neighbors of the Athenians in Attica when these were already counted as Greeks,” except that the same passage says that also these Pelasgians eventually began to be regarded as Greeks. See Laird 1933.

[ back ] 11. McNeal 1985, Georges 1994.131–133, 137; Thomas 2000.119–122, 216; 2001.222–223.

[ back ] 12. Cf. Herodotus 7.161.3, 9; Thucydides 1.2.1–3.4, 4.109.4. On Athenian historical and mythical discourse on autochthony, see Loraux 1993, esp. 37–71; Rosivach 1987; Connor 1993a.204–206; Georges 1994.135–136; Isaac 2004.114–124.

[ back ] 13. Cf. Plato, Menexenus 245c–d; Lysias 2.17–18; Loraux 1993.50–51; Rosivach 1987.302–303; Georges 1994.135; Dougherty 1996; Hall 1997.54; 2002.142; Thomas 2001.218. In Aeschylus Suppliants the Pelasgians of Argos clearly count as Greeks and symbolically represent the Athenians. Georges 1994.164, E. Hall 2002.142–143.

[ back ] 14. Cf. 6.52.1, 9.26.2–4, 27 and the genealogies of the Spartan kings at 7.204, 8.131.2; cf. 5.72. Malkin 1994.33–43; J. Hall 1997.56–65.

[ back ] 15. The usual antonyms are autokthōn (αὐτόχθων) and epēlus (ἔπηλυς), though Herodotus uses neither term in this passage (but see 1.171–172, 176.3, 4.197.2). Rosivach 1987.297–299. That barbaroi inhabited the Peloponnese before the arrival of the Greeks is already found in Hecataeus FGrHist 1 F 119 (= Strabo 7.7.1, C321).

[ back ] 16. 4.78–79, esp. 78.1; Campos Daroca 1992.45; Munson 2001.118–123. Cf. Harrison 1998, ch. 2, “Herodotus’ knowledge of foreign languages.”

[ back ] 17. Compare the slaughter of the Median boys by Cyaxares’ Scythian guests (1.73) with Psammetichus’ successful “Greek language program” (2.154; see below, p. 28–29).

[ back ] 18. The quoted expression is from Campos Daroca 1992.46; see also McNeal 1985; Asheri 1988.299; J. Hall 2002.194. That other factors were involved in the transition is implied by the wording ἅμα τῇ μεταβολῇ τῇ ἐς Ἕλληνας καὶ τὴν γλῶσσαν μετέμαθε (1.57.3).

[ back ] 19. Notably by the importation of the alphabet, which Herodotus attributes to the Phoenician companions of Cadmus (5.58).

[ back ] 20. 2.50.1–3 (cf. 2.4.2): from Egypt came all the gods except for the Dioscouri, Hera, Hestia, Themis, the Graces, and the Nereids, who are of Pelasgian origin, and Poseidon, who comes from Libya. For the polemical style of this discussion, see Thomas 2000.282.

[ back ] 21. 2.53. Linforth 1926; 1940; and see especially Burkert 1985 (cf. 1970.445–446), followed by Campos Daroca 1992.108. Thomas 2000.275–278. Mikalson 2003.167–173.

[ back ] 22. Hence the scholarly controversy centered on the meaning of the term ounoma as Herodotus uses it, especially in this passage—whether it must necessarily denote the very word composed of determinate sounds or letters, or whether it can more generally refer to the conceptual representation that name expresses, in whatever language. See the summary of the controversy in Lloyd 1976.203–205 and, most recently, the discussion by Harrison 2000.251–264. Irad Malkin explains in the following linguistic terms of the open–ended polytheistic mentality of the Greeks and other ancient peoples: “The gods of ‘others’ were either unfamiliar (‘new gods’) or, simply the ‘same’, but known by different names and attributes. Religion was langue; the names of their gods and their particular cults were parole. For example, when Herodotus says that ‘Ammon is the name of Zeus among the Egyptians’ (2.42), he sees a ‘Zeus’ in the Egyptian deity (langue), although his name, cult and even status may be peculiarly Egyptian (parole). In other words, he was saying, ‘Ammon is how you say ‘Zeus’ in Egyptian’” (Malkin 2004.350). In the terms of this passage, it appears that the word οὐνόματα at 2.50.1–3 has shifted semantically from the realm of parole to that of langue.

[ back ] 23. 2.156.5; 144.3. For Herodotus’ translations of the names of several foreign divinities, including Egyptian, see p. 31 and note 5.

[ back ] 24. 2.53.2. Epōnumia (ἐπωνυμία) can be used as a synonym of ounoma (οὔνομα; compare, for example 2.4.2 with 2.52.2), though it is generally interpreted as “epithet” in this passage (see e.g. Lloyd 1976.250). The difference is more properly that epōnumia tends to refer to a motivated name (given for some reason or after something or somebody, as at 4.45.2, where both terms occur). Campos Daroca 1992.85. For the term onoma (ὄνομα), see also below, p. 31–32.

[ back ] 25. Lattimore 1939; Lloyd 1976.204; Harrison 1998.4 “The Imagined Relationship between Greek and Foreign Languages”; 2000.256.

[ back ] 26. 1.131.2 suggests that Herodotus’ translations of the names of the gods might be more conceptual than linguistic: “The Persians call the whole vault of the sky ‘Zeus’.” On the other hand at 4.59.1–2 (see below, p. 44), the translation of the concept is accompanied by an etymological translation.

[ back ] 27. See also the later Pelasgian name of the goddess Themis (2.50). Cf. Aristotle’s idea of god as holding things together (On the Cosmos 397b10–401b24; Politics 1326a32). The etymology also recalls those of the Derveni commentator and of Plato’s Cratylus. See below, p. 22, note 18; p. 36.

[ back ] 28. For foreign words transparent to Greek in Herodotus, see Chamberlain 1999, who greatly emphasizes Herodotus’ sensitivity to “the occasional transparency of one language from the point of view of another (especially from the Greek point of view), and the role played in that transparency by minimal phonic changes” (Chamberlain 1999.278). Obviously transparent foreign names in Herodotus include the Scythian Papaios for Zeus at 4.59.2; Porata/Pyretos at 4.48.2; the Persian kings’ names at 6.98.3; Masistius/Makistius at 9.25.1; and the possibly “Egyptian” barbaroi at 2.158.5. See below, pp. 35, 44, 48–50, 65. Plutarch’s tendency in On Isis and Osiris to translate Egyptian names through the Greek is noticed by Donadoni 1947.

[ back ] 29. See also the ancient Greek-speakers of Dodona at 2.54–57, at a time when “Greece was called Pelasgia” (2.56.2). On the apparent lack of clarity in Herodotus’ treatment of the Pelasgians (which Myres 1907 and Laird 1933 attempt to dispel), see e.g. Campos Daroca 1992.75; Georges 1994.133–134; Thomas 2001.226–227; Sourvinou-Inwood 2003.121–131. For ethnea that are linguistically on the margin between the Greek and the barbarian world in historical times, see the Geloni, an originally Greek population of Scythia who speak a mixed Greek-Scythian language (4.108.2); and the Eretrians deported to Ardericca of Cissia, who “kept their old language” (6.119.4). The blurring of the boundaries is partially due to the unpredictable movement of peoples in the course of history (5.9.3).

[ back ] 30. See Sourvinou-Inwood 2003, esp. 121–131.

[ back ] 31. On the confusion of linguistic boundaries in Euripides, see S. Saïd 2002.69, 89.

[ back ] 32. Mosley 1971.1–2; Morpurgo Davis 2002.157, 169. J. Hall 1995.93; 1997.168–177; Harrison 1998; Colvin 1999.33–38. On Greek dialect difference, see esp. Thucydides 3.94.5, 4.3.3, 41.2, 112.4, 7.44.6, 57; Plato Cratylus 385d–e, 401c etc. (Rochette 1996.95–96); Xenophon Anabasis 3.1.26; and Aeschylus Seven Against Thebes, where an all-Greek heterophōnos stratos (ἑτερόφωνος στρατός) comes complete with barbarously whistling horses (170, 463). See S. Saïd 2002.89; Colvin 1999.75.

[ back ] 33. 6.138.2. Cf. Solon fr. 36.8–12, on Athenians sold abroad who no longer speak Attic. Athenian belief in the superiority of the Attic dialect seems to have increased in the fourth and third centuries (Isocrates Antidosis 295–296; Posidippus fr. 28 Kock). Colvin 1999.28–29. On the other hand, in the fifth century, the “Old Oligarch” complains that the Attic dialect is uniquely contaminated by barbarian speech (“Xenophon” 2.7–8).

[ back ] 34. On 1.139 (Dorian san/Ionian sigma), see below, p. 26. J. Hall 1997.152; 1985.86–89. For quotations of poetry and oracles in the epic-Ionic dialect, see Colvin 1999.60. For the Greek consciousness of internal language differentiation at the local level in the fifth century, see Hainsworth 1967.66–67.

[ back ] 35. Cf. 1.146. Asheri 1988.348. Campos Daroca 1992.42 observes that the linguistic differentiation noted by Herodotus (unverifiable for us on the basis of the epigraphic evidence) is part of his polemic against the Ionians; thus the Samians, who deserted at Lade, form a linguistic group by themselves.

[ back ] 36. 6.138.4 (see above, p. 7). See also 1.193.5 (“the palm trees which the Greeks call ‘male’”); 3.111.2 (cinnamon); 4.14.3 (poem Arimaspea), 4.189.2 (name of aigis of Athena). Some passages, where the name-users are implicitly Greek, rather refer to people in general: 4.45.2–5 (“Europe,” “Asia,” “Libya”), 3.33 (“sacred disease”), 3.122.2 (“the so-called ‘human generation’,” as opposed to the heroic age of Greek tradition). For Greek expressions for barbarian realities, see below, p. 33 and note 17.

[ back ] 37. The following are Spartan terms: 1.67.5, ἀγαθοεργοί; 6.57.2, Πύθιοι; 6.71.1, Zeuxidamos; 7.134, Talthybiads; 8.124.3, ἱππεῖς; for 9.11.2, see below, p. 16. For words from other parts of Greece, see 1.14.3, Γυγάδας; 4.18.1, Βορυσθενεῖται/Ὀλβιοπολῖται; 5.58, Φοινικήια and διφθέραι; 5.77.2, ἱπποβόται; 7.176.3, Χύτροι; 7.188.2, Ἑλλησποντίης; 7.197.2, λήιτον; 8.52.1, Αρήιος πάγος. A difference in terminology between Athenians and Boeotians on a borderline pass is noted at 9.39.1, Τρεῖς κεφαλαί/Δρυὸς κεφαλαί.

[ back ] 38. Cf. Plato Protagoras 341c, which labels Lesbian Greek as a barbaros phōnē.

[ back ] 39. Asheri 1988.297.

[ back ] 40. 1.56.2. Contrast the portrayal of fifth-century Athenians and Spartans in Thucydides 1.70–71.

[ back ] 41. On the double aspect of Greek identity expressed in this speech, see Moggi 1991.32.

[ back ] 42. As J. Hall (1995.92–93) observes, “the sometimes heterogeneous behavior of Greeks could begin to appear more uniform by contrast with the more alien practices of others.”

[ back ] 43. Thomas 2001. On similarities between different cultures in Herodotus, see Pelling 1997, Munson 2001.91–132. For “common blood” as an ethnic construct, cf. the rhetoric of Aristagoras at 5.49 with the narrator’s statements at 1.146–147 (mixed blood of the Ionians) and 1.151.2 (Methymnans of Lesbos enslave inhabitants of Arisba in spite of their common blood ties). Beside the Pelasgian origin of the Ionians, Herodotus also records the foreign origin of Greek aristocratic families (5.57, 5.66, two passages which Plutarch Malice of Herodotus 26 = Moralia 860E, regards as insulting; 6.53–54, on the Heraclids of Sparta; cf. 2.91), and the Greek ancestry of foreign peoples or dynasties (1.7). Lévy 1991.220–221.

[ back ] 44. Gould (1989.5) assumes that the Athenian definition represents a direct expression of Herodotus’ thought; J. Hall (1995.92; 1997.44–45; 2002.189–194) even argues that it constitutes a Herodotean innovation rather than a commonplace among fifth-century Greeks. But see Fornara 1972.84–86; also Raaflaub 1987.240; Georges 1994.130–131; Thomas 2001.214–215.

[ back ] 45. Campos Daroca 1992.48. Cf. Finley 1975.120–123.

[ back ] 46. Herodotus 8.136–139; cf. 5.21–22; 7.173.3; 8.34. In the case of the Macedonian royal family, by “Greek” Herodotus means Achaean (5.22). Badian 1994.

[ back ] 47. For markers, see Colvin 1999.21–26. In this case Colvin rightly notices, however (1999.59), that if the tradition is correct the Spartan lexical choice appears in its Ionic form (xeinous [ξείνους] as opposed to xenous [ξένους]). This is consistent with Herodotus’ practice when he reports speeches of non-Ionians.

[ back ] 48. Pace Jones 1996.315n4. From the point of view of the historical context of 479 BC, of course, the Spartan choice of words, if genuine, may be taken as reflecting the less polarized attitude of the Greeks in general before the Persian Wars (above, p. 2 and note 9); so Malkin 2004.349. Herodotus, at any rate, implies that it was almost inconceivable in his time.

[ back ] 49. Cartledge 1993.47. Cf. Moggi 1992.34: from the Greek point of view, the foreignness of the xenos is political, that of barbarians is also ethnic and cultural. Of course, barbarians of different ethnea (except for the Egyptians: see below, p. 65) refer to the Greeks, and each other, as xenoi.

[ back ] 50. The Spartan Amompharetus applies the word again to the Persians at 9.53.2. Contrast, in addition to the Spartan ambassadors’ references to “barbarians” at 8.142.1 and 5 (cited above, p. 16), Pausanias’ distinction at 9.79.1. When all the characters involved are Greek, the vocative xeine (ξεῖνε), is either uttered by a Spartan or addressed to a Spartan for a total of nine times out of ten (uttered by a Spartan: 5.49.9, 50.3, 9.79.1, 91.1 and 2, plus the inscription at 7.228.6; addressed to a Spartan: 1.68.2, 5.72.3, 7.160.1; single exception: 7.162.2). In other grammatical cases xeinos and xeinoi appear in the mouth of a Spartan (3.148.2, 5.51.2, 6.86b1, 7.226.2) or are used to denote strangers from the Spartan point of view (1.65.1, 3.55.2, 9.9.1, 9.91.1) or in reference to a Spartan (6.81) for a total of nine times out of eleven. (I have not counted instances where xeinos is used with the dative in the sense of ‘guest-friend of’: see Powell 1938, ad loc.) According to Herodotus’ account, lack of contacts with foreigners had been one of the problems in pre-Lycurgan Sparta (1.65.2), but for other sources this was still the case in the fifth century. Among the various stereotypes applied to the Spartans in Attic comedy (on which see Harvey 1994 esp. 39), we find the epithet dieironoxenoi, ‘treacherous toward outsiders’ in Aristophanes Peace 623.

[ back ] 51. On Spartan foreignness, see e.g. 6.58–60. Hartog 1988.152–156; Cartledge 1993.81; Munson 1993; 2001.96.

[ back ] 52. Cf. Thucydides 4.40.2. Colvin (1999.70–73, esp. 70) observes that when Xenophon “add[s] dialect colouring to a reported speech, it is always the Laconian dialect that is in question.” On Spartan local terms in Herodotus, see above, note 37; for Laconian brevity, see 3.46, 9.90–91, 7.226.1–2 and cf. Sthenelaidas in Thucydides 1.86.1. For silence in Spartan life, see David 1999.