A Phrygian Word Signifying Bread

At the city of Memphis in Egypt, the priests of Hephaestus’ temple told Herodotus a story about a king who wanted to know something no one could tell him. Before Psammetichus, who ruled Egypt in the seventh century BCE, the Egyptians had held themselves to be the oldest people on earth. Psammetichus wanted proof. Inquiry produced none, so he built an experiment. He took two newborn children from ordinary parents and gave them to a shepherd with instructions: no one was to speak in their hearing; they were to lie alone in an isolated hut; at the proper hour the shepherd was to bring goats to nurse them and see to the rest. Herodotus records the purpose in a single clause. The king did this "because he desired to hear what speech would first break from the children, when they were past the age of indistinct babbling" (Herodotus 2.2; trans. Godley 275–77).

The Greek is more exact than the English can be. Psammetichus waits for the children to be freed of τῶν ἀσήμων κνυζημάτων, whimperings without sign, and listens for ἥντινα φωνὴν ῥήξουσι πρώτην, whatever phōnē they will first break out with. After two years the shepherd opens the door and both children run to him, stretching out their hands, and say bekos. He keeps quiet. The sound recurs, visit after visit, until it has become, in Herodotus's phrase, τοῦτο τὸ ἔπος, this word, and he reports it. The king hears it for himself. Then he does the thing that makes this a story about language rather than about children: he inquires οἵτινες ἀνθρώπων βεκός τι καλέουσι, which people on earth call anything bekos. The answer comes back that the Phrygians call bread so. On the strength of it the Egyptians conceded that the Phrygians were older than they. Herodotus adds that he has all this from the priests at Memphis, and that the Greeks tell a sillier version in which the king first cut out the nurses' tongues (2.2; Godley 277).

Read slowly, the king's procedure is the strange part. He treats the syllable as a coin of unknown mint and sends out to find the treasury that struck it. The verb Herodotus gives the Egyptians for their conclusion is σταθμησάμενοι: they measured it off, as a builder measures with a line. The measuring works only if the coin has a mint, which is to say only if the sound already belongs to some language spoken somewhere. Rosaria Vignolo Munson states the presupposition directly. Psammetichus would never have proceeded as he did unless he assumed that a specific primordial language is innate in every human being, its words for common things already formed, and that untaught children will begin to speak it on their own (Munson, ch. 2). A modern reader is inclined to see a test of whether language is learned. The king had no such doubt. He had decided that language is found, like the source of a river, and he was looking for the spring.

Herodotus's own vocabulary leaves the matter more open than the king's. Phōnē in the Histories is the voice, the audible thing: he uses it for the human speech the people of Dodona finally recognized in an Egyptian priestess, and for the bird-cry they thought they heard from her before she learned Greek (2.55–57). For a language as such he reaches for glōssa. When he returns to Psammetichus's trial thirteen chapters later he does say glōssa, and the change of word marks a change of mind, to which I will come. At the door of the hut, however, the children make a phōnē. Whether it is a word is the very thing the king has to decide.

Antiquity supplied the alternative reading early, and the tenth-century Byzantine lexicon called the Suda preserved it. Retelling the experiment under the headword βεκεσέληνε, the Suda observes that if goats nursed the children and no woman did, then "it is no wonder that hearing the goat they imitated her voice," and the Phrygian word is a coincidence (Suda β 229; trans. Roth). Margaret Thomas notes that this objection has trailed the story for centuries (Thomas 37–62), and some modern readers think Herodotus planted the goats on purpose, meaning the attentive listener to catch the joke (Munson, ch. 2, notes 13–14). What matters about the objection is what it does. The bleat and the word are the same sound. Nothing in the air of that hut distinguishes them. The Egyptians heard a lexeme because a king had built the hut in order to hear one; the skeptic hears a goat because he knows what goats say. The evidence is the same in both accounts. The listener differs.

Athens got there before the Suda did. In the Clouds, staged in 423 BCE near the end of Herodotus's life, Socrates rounds on the old farmer Strepsiades: ὦ μῶρε σὺ καὶ Κρονίων ὄζων καὶ βεκκεσέληνε, you fool, you who smell of the age of Cronus, you bekkeselēnos (Aristophanes, Clouds 398). The scholiasts and the Suda derive the coinage from bekos and selēnē, the moon, two very old things fused into one insult for a man as old as the first word and the moon together (Suda β 228). While Herodotus was finishing his history, then, the syllable had already entered Attic slang as shorthand for hoary antiquity. The first word two children ever spoke had become a name for the man who is behind the times. It is a fair index of how Athens received the story: as a charming proof of nothing in particular.

Herodotus came to a similar view by a different road. He sets the experiment as the frontispiece to his Egyptian book and then spends the following chapters on geology. The Delta, he argues from shells on the hills and from mud brought up eleven fathoms down a day's sail from shore, is new land, the gift of the river (2.5, 2.12). If the Ionians are right that Egypt is the Delta alone, then there was once no Egypt for the Egyptians to be old in, and so, he concludes, "they needed not to make that trial to see what language the children would first utter" (2.15; Godley 293). He dismantles the experiment without touching its logic. The children's word holds no interest for him as evidence; the land does. And he pairs the trial with another. At Sais a temple scribe told him, joking, Herodotus suspected, that Psammetichus had once tested the springs of the Nile between two peaked hills by letting down a rope woven of many thousand fathoms, and that it never reached bottom (2.28; Godley 305–07). The same noun, διάπειρα, a trial, names both undertakings. The king sounded the river and found no floor. He sounded the human throat and found Phrygia. Herodotus reports both soundings and believes neither.

The story shows, in the end, where the meaning in that hut was. Herodotus records two things the children did when the door opened. They said bekos, and they stretched out their hands. Munson observes that Herodotus treats the gesture of entreaty as understood across every people he describes; a Phrygian exile at the Lydian court, Persian soldiers, and Theban captives all hold out their hands the same way (Munson, ch. 2, note 16). No one at Memphis needed to send to Phrygia to learn what the hands meant. The children were asking to be fed, and the shepherd, the king, and every reader since have understood them. The syllable was the part of the scene that required a philologist. The hands required no one. Psammetichus went looking for a first language in the sound and found one, because a search of that kind always finds one: with enough peoples in the world, some nation somewhere calls something by nearly any noise a child can make. What he could never have found by that method was a word, since a word is a sound to which a community has agreed to attach a meaning, and the only community in that hut was two infants and a herd of goats.

So what is usually called the first language experiment measured what its designer had already assumed. A great many experiments have that shape. It is why Herodotus, who trusted his own eyes on the shells in the hills over any priest's account of a hut, is the better guide to the question. He kept the story. He also kept his distance from it, and he let one word, phōnē, hold open the question the king had closed: whether, at the threshold of speech, a sound is yet a word, or only a voice.

Works Cited

Aristophanes. Aristophanes Comoediae. Edited by F. W. Hall and W. M. Geldart, vol. 2, Clarendon Press, 1907. Perseus Digital Library, www.perseus.tufts.edu/hopper/text?doc=Perseus%3Aabo%3Atlg%2C0019%2C003%3A435.

"Bekeselene." Suda On Line: Byzantine Lexicography, Adler numbers beta 228 and beta 229, translated by Catharine Roth, edited by David Whitehead, 21 July 2000, cs.uky.edu/~raphael/sol/sol-entries/beta/228 and cs.uky.edu/~raphael/sol/sol-entries/beta/229.

Herodotus. Herodotus, with an English Translation by A. D. Godley. Vol. 1, Loeb Classical Library 117, William Heinemann / G. P. Putnam's Sons, 1920. Internet Archive, archive.org/details/herodotus01hero. Proofread transcription of the same edition, English and Greek: LacusCurtius, penelope.uchicago.edu/Thayer/E/Roman/Texts/Herodotus/2A*.html and penelope.uchicago.edu/Thayer/H/Roman/Texts/Herodotus/2A*.html.

Munson, Rosaria Vignolo. Black Doves Speak: Herodotus and the Languages of Barbarians. Hellenic Studies Series 9, Center for Hellenic Studies, 2005. Chapter 2, "The Ethnographer and Foreign Languages," chs.harvard.edu/chapter/2-the-ethnographer-and-foreign-languages/.

Thomas, Margaret. "The Evergreen Story of Psammetichus' Inquiry into the Origin of Language." Historiographia Linguistica, vol. 34, no. 1, 2007, pp. 37–62. Post-print, eScholarship@BC, hdl.handle.net/2345/3222.