The Ideas

The myth of Theuth

Theuth is Plato’s form of the name of Thoth, the Egyptian ibis god of writing. Five voices meet in the passage, and it matters which one you quote: Plato wrote it, Socrates tells it as something he heard, Theuth makes the claim for writing, King Thamus gives the verdict, and Phaedrus doubts the story.

The numbers in brackets, such as 274c or 275a, are the standard references printed in the margins of the editions listed below and shown on Perseus; Jowett’s 1892 volume gives only the number (274, 275), not the letter. With these numbers every quotation here can be found in any printed edition below or on Perseus. Quotations from the Phaedrus in English come from two translations in the public domain: Benjamin Jowett’s (1892) and Harold North Fowler’s (Loeb Classical Library, 1913).

What the text says

Socrates has finished with the art of speaking and turns to “propriety and impropriety in writing” (274b, Fowler). He does not open with an argument of his own. He says he can tell “something I have heard of the ancients; but whether it is true, they only know” (274c, Fowler). The Greek word is akoē (ἀκοή), hearsay: the story arrives already marked as reported.

The story is short. In the region of Naucratis in Egypt (περὶ Ναύκρατιν) lived one of the old gods of the country, “the one whose sacred bird is called the ibis”, named Theuth. He “invented numbers and arithmetic and geometry and astronomy, also draughts and dice, and, most important of all, letters” (274c-d, Fowler). The king of all Egypt was Thamus, who lived in the great city of the upper region “which the Greeks call the Egyptian Thebes” (274d). Theuth showed him his arts one by one; Thamus asked what use each had and, as Theuth explained, praised or blamed what he said (274d-e). When they reached letters, Theuth made his claim: writing “will make the Egyptians wiser and will improve their memories; for it is an elixir of memory and wisdom that I have discovered” (274e, Fowler). In Greek the last clause reads μνήμης τε γὰρ καὶ σοφίας φάρμακον ηὑρέθη. Where Fowler writes “elixir”, Plato wrote pharmakon, a drug.

Thamus answers in three steps (274e-275b):

• The maker is not the judge. “One man has the ability to beget arts, but the ability to judge of their usefulness or harmfulness to their users belongs to another” (274e, Fowler). Theuth, “the father of letters”, has credited them with the opposite of their real power (275a).
• Reminding is not memory. Writing “will produce forgetfulness in the minds of those who learn to use it, because they will not practise their memory”. Its users will trust marks from outside (ἔξωθεν), “external characters which are no part of themselves”, instead of remembering from within (ἔνδοθεν). So Thamus hands back Theuth’s own word: οὔκουν μνήμης ἀλλὰ ὑπομνήσεως φάρμακον ηὗρες, “You have invented an elixir not of memory, but of reminding” (275a, Fowler).
• Appearance is not wisdom. “You offer your pupils the appearance of wisdom, not true wisdom, for they will read many things without instruction and will therefore seem to know many things, when they are for the most part ignorant and hard to get along with, since they are not wise, but only appear wise” (275a-b, Fowler). The last words render doxosophoi (δοξόσοφοι), wise in appearance.

The conversation does not stop at the verdict. Written words are like paintings: question them and “they always say only one and the same thing” (275d, Fowler). A written word “knows not to whom to speak or not to speak” and cannot defend itself (275e). Better is “the word which is written with intelligence in the mind of the learner” (276a). Even so, the man who knows will still write, for amusement, “to treasure up reminders for himself, when he comes to the forgetfulness of old age” (276d). Socrates does not throw writing out; he gives it the job Thamus named: reminding. Plato, who wrote the dialogue, did not throw it out either.

Four words in two translations

Both translations are in the public domain, and they render the key words differently. Jowett makes Theuth’s gift a medical “specific”, Fowler an “elixir”. Neither says “poison”.

What is disputed

Whether the story is Egyptian. The first doubt comes from inside the dialogue. Phaedrus answers at once: “Socrates, you easily make up stories of Egypt or any country you please” (275b, Fowler). Socrates does not defend the pedigree. He says that the people who listened to the oak at Dodona were content to hear an oak or a rock “provided only it spoke the truth”, whereas to Phaedrus “it makes a difference who the speaker is and where he comes from, for you do not consider only whether his words are true or not” (275b-c, Fowler). Phaedrus accepts the rebuke. No source checked for this page tells the story before Plato, and whether any Egyptian tale lies behind it is unsettled.

Our reading of that exchange. The exchange cuts against this site’s own method, so we say how we read it. Socrates asks only whether the words are true; we ask where and when they first appear. The two questions do different work. A dated source does not make a claim true, and a true claim can arrive with an invented pedigree. That is why we think the date is worth asking for, and why it is never the whole answer.

Who Ammon is. At 274d the manuscripts say that the Greeks call the city Egyptian Thebes “and the god Ammon” (καὶ τὸν θεὸν Ἄμμωνα). J. P. Postgate proposed in 1901 to read Θαμοῦν, “Thamus”, for θεὸν, “the god”, which would make the king himself the one the Greeks call Ammon. Burnet records the conjecture in the notes under his text (the critical apparatus) and keeps the manuscript reading in the text itself. At 275c Socrates speaks of “the prophecy of Ammon” (τὴν Ἄμμωνος μαντείαν), and Phaedrus calls the king “the Theban”. Both translations say “the god Thamus” at 274d, although the Greek there says only that Thamus was king of all Egypt. Whether Plato meant king and god to be one figure is a question about the text, not a settled fact.

Remedy or poison. Theuth and Thamus use the same word, pharmakon. The Liddell-Scott-Jones lexicon gives its first sense as “a drug, whether healing or noxious”, with “poison” among the sub-senses, but it files these two lines of the Phaedrus under “a means of producing something”. In the text Thamus does not call writing a poison. He calls it a means of reminding rather than of memory. Reading pharmakon as cure and poison at once belongs to the passage’s later life, notably Jacques Derrida’s essay of 1968 (see the table below).

When it was written. Harvey Yunis, in his Cambridge commentary of 2011, argues for a date of composition between 370 and 355 BCE, as Christopher Moore reports in his review. We give the range rather than a single year.

Whether it is the first such argument. Is this the oldest recorded argument about a new information technology? We have found no source that proves it, so this page does not claim it. The idea Theuth defends is older than Plato: a fragment of Euripides’ lost Palamedes, a play of the 5th century BCE, calls letters λήθης φάρμακα, remedies against forgetting (fr. 578, line 1, in August Nauck’s edition of 1889; the Liddell-Scott-Jones lexicon cites the line under pharmakon). Whether Thamus’ answer has an earlier parallel, we have not established.

From Theuth to Thoth to Hermes

Plato does not name Hermes in this passage. The bridge from his Theuth to the Hermes of the Hermetic tradition is built from other sources:

• The ibis. Plato’s Theuth is the god “whose sacred bird is called the ibis” (Phaedrus 274c).
• Letters. In the Philebus (18b) Socrates credits the first sorting of the sounds of speech into vowels and other elements to a god or a godlike man: “there is an Egyptian story that his name was Theuth” (tr. Fowler).
• Hermes. Herodotus, in the 5th century BCE, reports that dead ibises are taken “to the city of Hermes”, ἐς Ἑρμέω πόλιν (Histories 2.67, tr. Godley). A Greek writer already tied the ibis to a city named after Hermes.
• Laws and letters. In Cicero’s De natura deorum the Academic speaker Cotta lists five gods called Mercury, the Latin Hermes; the fifth is said to have fled to Egypt and given the Egyptians laws and letters (leges et litteras). The manuscripts give his Egyptian name as Theyn; H. Rackham’s Loeb text prints Baiter’s correction Theuth, Plato’s spelling (3.56).
• Thrice great. The epithet that turned this god into Hermes Trismegistus grew out of Thoth’s Egyptian titles; Copenhaver (1992) traces an early Greek form to the minutes of an ibis-cult council near Memphis in 172 BCE.

The links are documented identifications, not proof of one figure under several names. Hermes Trismegistus, a Greek-Egyptian creation, grew from the pairing of Thoth with the Greek Hermes.

How it was read later

The passage holds two questions: what a tool does to the people who use it, and whether a story is true when it comes with an ancient pedigree. Both return later, separately. The table is a reading list, not a chain of influence: Valla and Casaubon are listed for the kind of question they asked, not for any use of the Phaedrus.

One more echo, of a name rather than an argument: Corpus Hermeticum XVI is addressed to a king Ammon and warns that translating it into Greek will destroy its force (treatise XVI). We note the shared name, not a dependence between the two texts.

Ibis checks: the myth of Theuth

What the passage attests, and what retellings added to it.

Editions and translations

The Greek text and two public-domain English translations, with the pages that hold 274c-275b. Every link was opened on 29 October 2026.

The dates on these copies do not agree. One Greek text carries a preface of 1901, a digital header of 1910 and a reading-page label of 1903; one translation carries a title page of 1913, reprints that say 1914, and a reading page that says volume 9 and 1925 for a book whose title page says volume I. None of this changes a word of the passage. It shows why a citation names the edition and the page, and why a date is read from the title page rather than from a catalogue or a web header. The two Greek texts do differ in one small reading: at 274e Burnet prints τὰ τέχνης with the manuscript T and Stobaeus, while Fowler’s Greek has τὰ τῆς τέχνης with the manuscript B. The sense is the same.

How to check it

1. Find the passage by its number, not by a quotation: Phaedrus 274c-275b. The numbers are the same in every printed edition above and on Perseus, though Jowett’s 1892 volume gives only the number (274, 275), not the letter; the Project Gutenberg file has none.
2. Read the Greek and at least two translations side by side. Where they differ, as with “specific” and “elixir”, a dictionary entry for the Greek word shows what the text can and cannot carry.
3. Name the speaker. In this passage Plato writes, Socrates narrates, Theuth claims, Thamus judges and Phaedrus doubts. A quotation that begins “Socrates said” has already chosen one of five voices.
4. Take the date from the title page or the preface, not from a catalogue or a web header.
5. Look the claim up in Ask for the date, which keeps a dated card for “Socrates (or Plato) said writing destroys memory”, or take any other “ancient wisdom” quotation through the four layers of Provenance.

What this can and cannot tell us about AI

Source. Thamus judges one tool, writing, by what it does to the people who use it: reminders in place of memory, the appearance of wisdom in place of wisdom (275a-b). It is an argument, not a measurement. Our reading. Two of his worries become questions worth asking of an AI assistant: what do I still know without the tool, and does this answer only sound like knowledge? One of Socrates’ own points after the myth does not carry over: written words, he says, “always say only one and the same thing” when questioned (275d), while a language model answers back. A closer fit is his next point (275e): a written word “knows not to whom to speak or not to speak”. Practice. Write your own answer before you ask, as in the five questions before you ask AI. Hypothesis. Doing so may keep more of the work in your own memory instead of in reminders. No study cited here tests that. Not settled here. Whether handing memory to search engines or AI weakens it is a question for studies of today’s tools, read with their dates, not for Plato. Nor does the passage settle whether Thamus was right about writing; his verdict itself reaches us only because Plato wrote it down. The wider question, what to hand to a model and what to keep, is on the page human agency in the AI age.

Key facts

Where	Plato, Phaedrus 274c-275b
Written	c. 370-355 BCE (Yunis 2011)
Who speaks	Plato writes, Socrates tells, Theuth claims, Thamus judges, Phaedrus doubts
Theuth	Plato’s form of the name of Thoth, the Egyptian ibis god
Key words	pharmakon (drug); mnēmē (memory) and hypomnēsis (reminding); doxosophoi (wise in appearance)
First doubt about the source	Phaedrus, 275b
Public-domain translations	Jowett 1892, vol. I, pp. 483-485; Fowler, Loeb Classical Library 36 (1913), pp. 561-565

Fair questions

Did Socrates say writing destroys memory?

Not in those words, and not in his own voice. In Plato’s Phaedrus (274c-275b) Socrates tells a story he says he heard, in which King Thamus tells the god Theuth that writing “will produce forgetfulness” in those who learn it, because they will not practise their memory (275a, tr. Fowler). Plato wrote the dialogue, around 370-355 BCE.

Who were Theuth and Thamus?

In Plato’s story Theuth is an Egyptian god from the region of Naucratis, “the one whose sacred bird is called the ibis”, who invented numbers, geometry, astronomy, games and letters; the name is Plato’s form of Thoth. Thamus is king of all Egypt and lives at Thebes. At 275c Socrates speaks of “the prophecy of Ammon”, and a conjecture by J. P. Postgate (1901) at 274d would make Thamus the one the Greeks call Ammon.

Is the myth of Theuth a real Egyptian myth?

Its setting and its god are Egyptian, but the story is first attested in Plato, and the dialogue raises the doubt at once: Phaedrus says Socrates easily makes up “stories of Egypt or any country you please” (275b, tr. Fowler). No source checked for this page tells it before Plato; whether an Egyptian tale lies behind it is unsettled.

What does pharmakon mean in the Phaedrus?

A drug. The Liddell-Scott-Jones lexicon gives “a drug, whether healing or noxious” and files these lines under “a means of producing something”. Jowett (1892) translates “specific”, Fowler (1913) “elixir”. Thamus turns the word back on Theuth: writing is a pharmakon not of memory but of reminding (275a). Reading it as remedy and poison at once comes from later interpretation, notably Derrida’s “Plato’s Pharmacy” (1968).

Did Plato reject writing?

No. After the myth Socrates criticises written words because they cannot answer questions or defend themselves (275d-e), but he still lets the man who knows write “to treasure up reminders for himself, when he comes to the forgetfulness of old age” (276d, tr. Fowler). Writing keeps the role Thamus gave it, a reminder, and the argument itself reaches us because Plato wrote it down.

Sources & further reading

• Plato, Phaedrus 274b-276e, Greek text: John Burnet (ed.), Platonis Opera, vol. II, Scriptorum Classicorum Bibliotheca Oxoniensis (Oxford: Clarendon Press; preface dated March 1901), with critical apparatus (274d: Postgate’s conjecture; readings of Stobaeus and Hermias); Internet Archive scan (consulted 29 October 2026).
• Benjamin Jowett (tr.), The Dialogues of Plato, 3rd ed., vol. I (Oxford: Clarendon Press, 1892), pp. 483-485; Internet Archive scan (consulted 29 October 2026). Public domain.
• Harold North Fowler (tr.), Plato I: Euthyphro, Apology, Crito, Phaedo, Phaedrus, Loeb Classical Library 36 (London: Heinemann; New York: Macmillan, title page MCMXIII = 1913), pp. 560-565; Internet Archive scan (consulted 29 October 2026). Translation in the public domain in the US and, from 2026, in the EU.
• Perseus Digital Library, Tufts University: Plato, Phaedrus 274-276, Greek (text 1999.01.0173) and English, tr. Fowler (text 1999.01.0174); Plato, Philebus 18b, tr. Fowler; Herodotus, Histories 2.67, Greek text and tr. A. D. Godley (Loeb, 1920). Digital texts marked CC BY-SA 3.0 US (consulted 29 October 2026).
• Cicero, De natura deorum 3.56, Latin text at The Latin Library, which does not name its edition and prints Theyt. Critical text: H. Rackham (tr.), Cicero, De natura deorum; Academica, Loeb Classical Library 268 (first printed 1933), p. 340, apparatus: Theuth Baiter, Theyn manuscripts; Internet Archive scan of the 1967 reprint (both consulted 29 October 2026).
• H. G. Liddell, R. Scott and H. S. Jones, A Greek-English Lexicon, s.v. φάρμακον: sense I, “a drug, whether healing or noxious”; sense II, “remedy, cure”, citing λήθης φάρμακα, of letters, from Euripides fr. 578.1 (Nauck numbering; the lost Palamedes); sense II.2, “a means of producing something”, citing Phaedrus 274e and 275a; consulted at lsj.gr, 29 October 2026.
• August Nauck (ed.), Tragicorum Graecorum Fragmenta, 2nd ed. (Leipzig: Teubner, 1889), Euripides, Palamedes, fr. 578, p. 542, preserved by Stobaeus; the headnote on p. 541 dates the play to Olympiad 91.1, citing Aelian, Varia Historia 2.8; Internet Archive scan (consulted 29 October 2026).
• J. P. Postgate, “On Plato, Phaedrus 274 D”, The Classical Review 15.1 (1901), p. 27, doi:10.1017/S0009840X00029371.
• Harvey Yunis (ed.), Plato: Phaedrus, Cambridge Greek and Latin Classics (Cambridge: Cambridge University Press, 2011), x + 270 pp.; its date of composition, 370-355 BCE, as reported by Christopher Moore, review, Bryn Mawr Classical Review 2012.07.55.
• Lorenzo Valla, De falso credita et ementita Constantini donatione, ed. Wolfram Setz (Weimar: Hermann Böhlaus Nachfolger, 1976); G. W. Bowersock (ed. and tr.), On the Donation of Constantine, I Tatti Renaissance Library 24 (Cambridge, MA: Harvard University Press, 2007); the date 1440 from Marc van der Poel, review, Bryn Mawr Classical Review 2008.01.31.
• Isaac Casaubon, De rebus sacris et ecclesiasticis exercitationes XVI (London, 1614), I.10; Anthony Grafton, “Protestant versus Prophet: Isaac Casaubon on Hermes Trismegistus”, Journal of the Warburg and Courtauld Institutes 46 (1983), pp. 78-93.
• Brian P. Copenhaver, Hermetica (Cambridge: Cambridge University Press, 1992), introduction: the epithet “thrice great” and its early Greek form.
• Jacques Derrida, “La pharmacie de Platon”, Tel Quel (1968); in English as “Plato’s Pharmacy”, in Dissemination, tr. Barbara Johnson (Chicago: University of Chicago Press, 1981). Cited as reception only.
• Neil Postman, Technopoly (New York: Knopf, 1992), ch. 1, “The Judgment of Thamus”. Cited as reception only.