Early Journal Content on JSTOR, Free to Anyone in the World

This article is one of nearly 500,000 scholarly works digitized and made freely available to everyone in the world by JSTOR.

Known as the Early Journal Content, this set of works include research articles, news, letters, and other writings published in more than 200 of the oldest leading academic journals. The works date from the mid-seventeenth to the early twentieth centuries.

We encourage people to read and share the Early Journal Content openly and to tell others that this resource exists. People may post this content online or redistribute in any way for non-commercial purposes.

Read more about Early Journal Content at http://about.jstor.org/participate-jstor/individuals/early- journal-content.

JSTOR is a digital library of academic journals, books, and primary source objects. JSTOR helps people discover, use, and build upon a wide range of content through a powerful research and teaching platform, and preserves this content for future generations. JSTOR is part of ITHAKA, a not-for-profit organization that also includes Ithaka S+R and Portico. For more information about JSTOR, please contact support@jstor.org.


Vou. XLI, 1. WHOLE No. 161


Some years ago I contributed to this Journal (XXIII 261- 282; 362-387)? an article in which I undertook to reconstruct the plot of a popular tale of Gyges and the King of Lydia which appears to have been still current in the period of Herodotus and Plato. In the process of my investigation, which I carried as far as the fall of the Eastern Empire, it became more and more evident that the most notable feature of the later tradition of Gyges and Candaules was the increasing preponderance in it of the two versions by Herodotus and Plato respectively. The matter had no bearing upon the subject which I was then discussing, and I therefore mentioned it only in passing. But the fact itself is so characteristic of ancient literary tradition as such, and in some ways is so striking a commentary upon it, that it seems worthy of special consideration.

Let us begin with the later tradition of Plato’s story of Gyges and his Ring (Republic 359 D). It will be remembered that in this passage the spokesman, discussing the well-known doctrine that the only thing which prevents even the best of us from doing wrong in the end is the fear of detection, asserts that his point would be proved if both a good and a bad man could be given some power which would render detection impossible. “I

1 This article was transmitted to the JOURNAL a few hours before the author’s death, and so did not have the benefit of his final revision.—Eb.

2My investigation did not concern itself with the ultimate origin, meaning or credibility of the various accounts. For these points the best and most recent authority is Lehmann-Haupt, PRE s. v. Gyges.



mean,” he says, “such a power ... as they say was once pos- sessed by the ancestor of the Lydian.” Then by way of at once enforcing his point and explaining his reference, he tells the story in question. When he returns to the story at 612 B Plato couples the ring of Gyges with another more ancient and more famous method of going invisible, the Homeric “Aides κυνέη or Hades’ Cap of Darkness.”

A brief and interesting story told by a master and in his best style, a story with a moral, above all a story with a literary reference (Γύγου δακτύλιος) which could be used to great advantage by writers and speakers—so far as rhetoricians were concerned, here, as the old translator of Bayle says of books of extracts, was “meat already chawed.” Nevertheless, we hear nothing of the story until Cicero (Off. III 9, 38) translates it in connection with his discussion of the same question of con- duct. And strange to say I have been unable to find a single reference in any other Roman author.®

Even on the Greek side I find no mention of this story until Ptolemaeus Chennus (Myth. Graeci, p. 192 W) at the end of the first century of our era. Chennus was a sort of purveyor in ordinary to the literary chit-chat so characteristic of that period. As such he can tell us, for example, why the Queen was able to see Gyges in spite of his ring. She had a double pupil, also a dragon-stone. This shows of course that the literary world was on the whole quite well aware of the relation between the story of Plato and the story of Herodotus. Such a book as the Suasoriae and Controversiae of the Elder Seneca, not to mention a number of others, is enough in itself to show that in practi- cally every instance the source and associations of these semi- popular literary discussions were scholastic. It is fair, therefore, to assume that our passage in Plato had already been familiar to the Rhetorical Schools for a long time. However that may be, we know that it had entered them at least as early as the First Sophistic Renaissance. This we learn from the Progymnasmata of Theon, one of the most notable figures in the educational life of that period.

In the second chapter of this text-book (Rhet. Graeci, 1 159

3In N. H. XXXIII 8, Midae quidem anulum, quo circumacto habentem nemo cerneret, quis non etiam fabulosiorem fateatur? Pliny was hardly thinking of Plato’s story; see A. J.P. XXIII 273.


Walz) for the use of students and teachers, Theon recommends and in some cases discusses those passages from the great classical authors which every schoolboy was expected to learn by heart. These passages were selected and graded according to the age and training of the student, and for the most part fall into three classes: 1. anecdotes (χρεῖαι), 2. fables (μῦθοι), 3. stories (Smynoas)—these last being again subdivided into mythical stories and stories of actual fact. Under the first subdivision (διηγήσεις μυθικαί) four examples are recommended:

Διηγήσεως δὲ παραδείγματα ἂν εἴη κάλλιστα, τῶν μὲν μυθικῶν Πλάτωνος ἐν τῷ δευτέρῳ τῆς πολιτείας περὶ τοῦ δακτυλίου τοῦ Γύγου " καὶ ἐν τῷ συμποσίῳ, περὶ τῆς γενέσεως τοῦ ἔρωτος - περὶ δὲ τῶν ἐν ἽΑιδου, ἐν τῷ Φαίδωνι, καὶ τῷ δεκάτῳ τῆς πολιτείας - καὶ παρὰ Θεοπόμπῳ ἐν τῇ ὀγδόῃ τῶν Φιλιππικῶν τοῦ σελίνου.

It will be observed that the very first of these passages is Plato’s story. Is it at all surprising to discover that the Ring of Gyges suddenly becomes prominent in the writers of this particular period? We shall also find a practically unbroken tradition of its use as a literary reference until the fall of the Kastern Empire.

In his Bis Accusatus 21, Lucian makes Epicurus say, in his plea for pleasure as against the claims of the Stoa, that these apostles of toil and efficiency

Cannot bear to be detected in any relaxation, or any depar- ture from their principles: but, poor men, they lead a Tantalus’ life of it in consequence, and when they do get a chance of sinning without being found out, they drink down pleasure by the bucketful. Depend on it, if some one would make them a present of Gyges’s ring of invisibility, or Hades’s cap, they

would cut the acquaintance of toil without further ceremony, and elbow their way into the presence of Pleasure.”

Again in the Navigium 42, Timolaus is made to say

My wish is that Hermes should appear and present me with certain rings, possessed of certain powers. One should ensure its wearer continual health and strength, invulnerability, insen- sibility to pain. Another, like that of Gyges, should make me invisible.”

Epist. Graec. p. 619, 43 Didot (Aischines to Xenophon) we have:

κἂν πολλάκις περικρύπτηται περιθέμενος τὴν “Aidov κυνῆν τὸν Τύγου δακτύλιον καὶ δίκας γράφηται τοῖς ἐν τῇ πόλει - ζῇ γὰρ ἀπὸ βυρσοδεψικῆς.


Libanius, Orat. LVI (Contra Lucianum), 10 says:

ἀλλ᾽ ds ἤδειμεν χάριν μᾶλλον ὄψιν ἡδίστην θεώμενοι, Λουκιανὸν ἀσθενῆ καὶ ζητοῦντα τὴν ἀρχήν, οὐκ ἔχοντα, χρῆν οὐκ αὐτὸν ἐνθυμούμενον, κλέψαι τῇ νυκτὶ τὴν εἴσοδον, ἐπεὶ μὴ πρίασθαί γε ἐξῆν ποθεν τὸν Γύγου δακτύλιον μισθωσάμενον γοητείαν ὑπὸ τοῖς ἐκείνης μαγγανεύμασι δραμεῖν.

Again, Orat. LXIV (Pro Saltat.) 35, he says:

ἀλλ᾽ οὐκ ἂν ἔχοις, εἰ μή, νὴ Δία ye, τὴν ἴΑιδος κυνῆν, τὸν Γύγου δακτύλιον ἔχοντες ἀδικοῦσιν, ὑφ᾽ ὧν λανθάνουσιν.

And finally in his Epistles 1031, we have (as quoted by L.-S. Paroem. Graec. II, Ὁ. 20):

σὺ δ᾽ οἴου μετὰ τοῦ δακτυλίου τοῦ Γύγου πάντα δρῶν λανθάνειν.

The use of the phrase by Gregory of Nazianzus is glib but evidently quite mechanical; cp. Orat. Contra Julianum (35, p. 628 Migne) :

ἀλλ᾽ οὐκ ἔστιν ὅπως ἑαυτὸν ἀποκρύψει, οὐδ᾽ ἂν πολλὰ στραφῇ καὶ παντοῖος γένηται ταῖς ἐπινοίαις, οὐδ᾽ εἰ τὴν Αἴδος κυνέην, δὴ λέγεται, περιθέμενος τῷ δακτυλίῳ Γύγου, καὶ τῇ στροφῇ τῆς σφενδόνης χρησάμενος, ἑαυτὸν ἀποκλέψειε, ete. : Orat. 48, 21 (L. andS., Paroem. Graeci, 1, Ρ. 21), πλέον ἐφρονοῦμεν τῇ στροφῇ τῆς σφενδόνης

Τύγης, εἴπερ μὴ μῦθος ἦν, ἐξ fs Λυδῶν ἐτυράννησεν : Carmina, Lib. I, 2, 30 (87, p. 685 Migne),

κέρδος τοσοῦτον κἂν τρέχειν ὅρους δοκῇς,

κἄν σοι τὰ Γύγου τοῦ πολυχρύσου παρῇ

στρέφῃς τε πάντα τῇ στροφῇ τῆς σφενδόνης

σιγῶν δυνάστης, οἷο. The emphasis which Gregory lays on σφενδόνη indicates in itself that this old word used by Plato for the bezel of a ring had long been obsolete or obsolescent.

Doubtless the hereditary reference to Gyges’ Ring occurs here and there in the huge Corpus of Greek Fathers edited by Migne—one would expect it for instance to be used by such a firebrand of rhetoric as Joannes Chrysostomus—but I have made no effort to examine this field systematically.

That the phrase continued to live, however, and to be used more or less frequently, is shown among other things by the frequency of its occurrence in the Paroemiographi Graeci, cp. Apostolius 5, 71 (P. G. 2, 353) ; Macarius, 3, 9 (P. G. 2, 154) ; Diogenianus, 3, 99 (P. G. 1, 232 and 2, 20); Greg. Cyp. 2, 5 (P. G. 1, 358).


So much for the later literary reference to Gyges’ Ring. Among the authors whose interest in the story evidently went beyond the mere phrase which we have been discussing, the most notable is Philostratus. In the Heroicus, 2, 137, 29 sqq., he gives a brief rhetorical version of the old story, as follows:

΄ a Kat μήν, εἰ μυθολογικὸς ἦν, τόν τε τοῦ ᾿Ορέστου νεκρὸν διήειν ἄν, ὃν ε < 2 , 4 a 3 cad αλι a” cal ἑπτάπηχυν ἐν Νεμέᾳ Λακεδαιμόνιοι εὗρον, καὶ τὸν ἐν τῷ χαλκῷ ἵππῳ τῷ A vot a rf Ν >. A δέ ν Tv: Ε} nw δὲ fol aA iy ὃς κατωρώρυκτο μὲν ἐν Λυδίᾳ πρὸ Τύγου ἔτι, σεισμῷ δὲ τῆς γῆς a “- e gy διασχούσης θαῦμα τοῖς περὶ Λυδίαν ὥφθη ποιμέσιν, οἷς ἅμα 6 Τύγης 2, , 3 x a“ νΝ 4 4 ΕΣ ε ΄ a ἐθήτευσεν. ἐς γὰρ κοῖλον τὸν ἵππον θυρίδας ἐν ἑκατέρᾳ πλευρᾷ ἔχοντα νεκρὸς ἀπέκειτο μείζων ἀνθρώπου δόξαι.

In his life of Apollonius of Tyana, ITI 8, describing how the wonderful Indian dragons are hunted, he says:

a ΄ a 2 i“ 4 4 Ν tod Led κοκκοβαφεῖ πέπλῳ χρυσᾶ ἐνείραντες γράμματα τίθενται πρὸ τῆς χειᾶς ὕπνον ἐγγοητεύσαντες τοῖς γράμμασιν, ὑφ᾽ οὗ νικᾶται τοὺς ὀφθαλμοὺς 6 a 3 δράκων ἀτρέπτους ὄντας, καὶ πολλὰ τῆς ἀπορρήτου σοφίας ἐπ᾽ αὐτὸν ἄδουσιν, οἷς ἄγεταί τε καὶ τὸν αὐχένα ὑπερβαλὼν τῆς χειᾶς ἐπικαθεύδει τοῖς γράμμασι : προσπεσόντες οὖν οἱ “Ivdol κειμένῳ πελέκεις ἐναράττουσι, καὶ A \ 3 ΄ , BS 3 2A i 2 ~ 6. δέ τὴν κεφαλὴν ἀποτεμόντες λήζονται τὰς ἐν αὐτῇ λίθους. ἀποκεῖσθαι δέ φασιν ἐν ταῖς τῶν ὀρείων δρακόντων κεφαλαῖς λίθους τὸ μὲν εἶδος ἀνθηρὰς καὶ πάντα ἀπαυγαζούσας χρώματα, τὴν δὲ ἰσχὺν ἀρρήτους κατὰ τὸν

δακτύλιον, ὃν γενέσθαι φασὶ τῷ Γύγῃ."

Of the passages remaining to be considered some are merely notes designed to explain the reference to Gyges’ Ring, others are rhetorical abstracts, all are directly due to the scholastic tradition.

The story, for instance, is told by Nonnus in his note on Gregory of Nazianzus, Invect. 1, 55 (text in Westermann’s Mythographi, p. 366, XVI) as follows:

Πλάτων φιλόσοφος ἐν πολιτείαις (ἔστι δὲ οὕτως αὐτοῦ λεγομένη πραγματεία) εἰσφέρει τὸν μῦθον τοῦτον, οὕτω λέγων, ὅτι Γύγης ἦν τις ποιμὴν περὶ τὴν Λυδίαν - οὗτος ποιμαίνων ἔν τινι ὄρει τὰ πρόβατα

΄ ν΄ vr x 3 Ν ¥ > a = φ Ν

περιέτυχε σπηλαίῳ τινί, καὶ εἰσελθὼν ἐν αὐτῷ εὗρεν ἵππον χαλκοῦν καὶ ᾿ς οἵ

ἔνδον τοῦ χαλκοῦ ἵππου ἄνθρωπον νεκρὸν καὶ δακτύλιον - οὗ δακτυλίου

κεφαλὴ στρεπτὴ ἦν καὶ ἐστρέφετο. ἔλαβεν οὖν Τύγης τὸν δακτύλιον

καὶ ἐξῆλθε - καὶ ἡνίκα μὲν ἦν ἐν τῇ τάξει δακτύλιος, ἑωρᾶτο ὑπὸ

΄ Δ. ὧδ Ν Ἂν: ΄ a ed my 3 Ν a. πάντων, ἡνίκα δὲ τὴν σφενδόνην τοῦ δακτυλίου ἔστρεφεν, ἀφανὴς ἐγίνετο

a 3 5 f 3 sf Ν a a ε ΄ 3 , πᾶσιν. οὖν Πλάτων εἰσφέρει τὸν μῦθον τοῦτον, ὅτι δίκαιος ἀνήρ, κἂν τοῦ Γύγου λάβῃ δακτύλιον, ἵνα μὴ δρᾶται ὑπό τινος, οὐδ᾽ οὕτως

‘In my article on Gyges, A. J.P. XXIII 370, I somehow managed to translate κατὰ τὸν δακτύλιον, “even against the ring,” as though it were a genitive instead of an accusative, according to the ring.”


ὥφειλεν ἀδικεῖν - δεῖ γὰρ τὸ καλὸν δι’ αὐτὸ τὸ ἀγαθὸν ἐπιτηδεύειν Kal μὴ

δι’ ἄλλους τινάς.

Nonnus practically repeats the same note in his commentary on Gregory, Orat. in Basil. 5.

Poor old Ioannes Tzetzes, at the dawn of the Renaissance—a man who would have been a distinguished scholar if he had had half a chance—was especially interested in our two passages and saw more or less clearly the original relation between them. His Chiliades is now so rare a book that I quote here in full the passages in point:

Chiliades I 137-166:

Ῥύγης τὸ πρότερον ποιμὴν κατά τινας ὑπάρχων, Ποιμαίνων εἷρέ που χαλκοῦν ἵππον ἐγκεχωσμένον, Εἰς ὅνπερ ἵππον ἔνδοθι νεκρός τις κεκλιμένος Στρεπτὸν περὶ τὸν δάκτυλον δακτύλιον ἐφόρει. Τοῦτον γοῦν τὸν δακτύλιον οὗτος λαβὼν Γύγης Καὶ γνοὺς ὡς ἔχει δύναμιν, σφενδόνης στρεφομένης, Συγκρύπτειν τὸν κατέχοντα καὶ πάλιν ἐμφανίζειν, Κτείνας Κανδαύλην ἔλαβε Λυδῶν τὴν βασιλείαν. Ἡρόδοτος τὸν Γύγην δὲ ποιμένα μὲν οὐ λέγει, Υἱὸν Δασκύλου δέ φησιν, ὑπασπιστὴν Κανδαύλου - 7 , ΄ Ὅστις Κανδαύλης γυναικὸς ἔρον οἰκείας τρέφων Ῥυμνὴν αὐτὴν ὑπέδειξε τῷ Γύγῃ λεληθότως. δὲ καὶ γνοῦσα σιωπᾷ, εἶτα καλεῖ τὸν Γύγην, Αἵρεσιν λέγουσα λαβὲ Γύγη τῶν δύο piav, σὺ Κανδαύλην ἀνελεῖν φονευθῆναι τούτῳ. Γυμνὴ δυσὶν ἀνδράσι γὰρ οὐ στέγω θεαθῆναι. Οὕτω Κανδαύλην ἀνελὼν εἷλε τὴν βασιλείαν. Ἔκ τῆς Κανδαύλου γυναικὸς ΓΑρδυς υἱὸς τῷ Γύγῃ, ἼΛρδυος Σαδυάττης δέ, καὶ τούτου ᾿Αλυάττης, Ἐξ ᾿Αλνυάττου Κροῖσος δέ, ὅστις ἡττᾶται Κύρῳ. "ANN ἤδη σε σφαδάζοντα καὶ κεχηνότα βλέπω, Τὴν Γύγου χρήζοντα μαθεῖν πᾶσαν ἀλληγορίαν. Ποιμὴν 6 Τύγης λέγεται τῷ στρατηγὸς τυγχάνειν " Ἵππος χαλκοῦς ἀγέρωχός ἐστιν βασιλεία, Ναὶ μὴν καὶ τὰ ἀνάκτορα " νεκρός, γυνὴ Κανδαύλου, Τῶν ἀνακτόρων ἄπρακτος ἔνδοθεν καθημένη. *Hs τὸν δακτύλιον λαβὼν ὑπασπισταῖς δεικνύει, Καὶ σὺν αὐτοῖς ἀπέκτεινε λαθραίως τὸν Κανδαύλην. Στρέψας δὲ τὸν δακτύλιον πάλιν πρὸς τὴν γυναῖκα Γίνεται πᾶσιν ἐμφανής, λαβὼν τὴν βασιλείαν.

Id. VI 481-484:

Νυσσία οὖσα σύζυγος Μυρτίλου τοῦ Κανδαύλου - To δὲ Κανδαύλης Λυδικῶς τὸν σκυλοπνίκτην λέγει "


Ἐπεὶ Κανδαύλης ἔδειξε γυμνὴν αὐτὴν τῷ Deyn, Κτανεῖν τὸν Γύγην ἔπεισεν αὑτῆς τὸν συνευνέτην.

Id. VII 195-202:

Τυμνὴν Κανδαύλης ἔδειξε τῷ Γύγῃ σφὴν γυναῖκα - Ἥτις καὶ ᾿συγκαλέσασα τὸν Ῥύγην κατιδίαν

Δίδωσι τὸν δακτύλιον αὐτῆς, ὡς ἀποκτείνῃ Κανδαύλην ταύτης σύζυγον, δείξας κρυφῆ συμμάχοις. οὗ γεγονότος κτείνας τε λαθραίως τὸν Κανδαύλην Καὶ στρέψας τὸν δακτύλιον πάλιν εἰς τὴν γυναῖκα, Τίψεται πᾶσιν ἐμφανὴς λαβὼν τὴν βασιλείαν.

“Eyes ἐν πρώτῳ πίνακι τρίτην τὴν ἱστορίαν.

Last of all, we have the following account in the so-called Violarium of Eudocia (now considered the work of some scholar of the Renaissance), 247:

Τύγην οἱ Ἕλληνες ἐμυθεύοντο τῇ στροφῇ τῆς σφενδόνης, ἣν ἐφόρει, ἀφανίζεσθαι καὶ μὴ ὁρᾶσθαι παρόντα καὶ εἰς ὄψιν ἔρχεσθαι - ὃν καὶ Πλάτων 6 φιλόσοφος ἐν ἸΠολιτείαις εἰσφέρει μυθικῶς οὕτω λέγων, ὅτι Γύγης τις ἦν ποιμὴν περὶ τὴν Λυδίαν. οὗτος ποιμαίνων ἔν τινι ὄρει τὰ πρόβατα περιέτυχε σπηλαίῳ τινί. καὶ εἰσελθὼν ἐν αὐτῷ εὗρεν ἵππον χαλκοῦν, καὶ ἔνδον τοῦ χαλκοῦ ἵππου νεκρὸν ἄνθρωπον φοροῦντα δακτύ- λιον, οὗ δακτυλίου κεφαλὴ στρεπτὴ ἦν καὶ ἐστρέφετο. yris σφενδόνη ἐκαλεῖτο. ἔλαβεν οὖν 6 Τύγης τὸν δακτύλιον, καὶ ἐξῆλθεν. καὶ ἡνίκα μὲν ἦν ἐν τῇ τάξει δακτύλιος, ἑωρᾶτο ὑπὸ πάντων, ἡνίκα δὲ τὴν σφενδόνην τοῦ δακτυλίου ἔστρεφεν, ἀφανὴς ἐγίνετο ἐν πᾶσιν. οὖν Πλάτων εἰσφέρει τὸν μῦθον τοῦτον, ὅτι, φησίν, δίκαιος ἀνήρ, κἂν τοῦ Γύγου λάβῃ τὸν δακτύλιον, ἵνα μὴ ὁρᾶται tad τινος, οὐδὲ οὕτως ὀφείλει ἀδικεῖν. δεῖ γὰρ τὰ καλὸν δι’ αὐτὸ τὸ ἀγαθὸν ἐπιτηδεύεσθαι, καὶ μὴ δι᾽ ἄλλο τι. ἔχων οὖν Τύγης τοῦτον τὸν δακτύλιον, ἐλθὼν ἐπὶ τὰ βασί- λεια τῶν Λυδῶν καὶ ἀντιστρέψας τὴν σφενδόνην ἐγένετο ἀφανής. καὶ εἰσελθὼν ἀπέκτεινε τὸν βασιλέα καὶ ἔλαβε τὴν βασιλείαν. διὸ καὶ Τύγου δακτύλιος ἐπὶ τῶν πολυμηχάνων καὶ πανούργων λέγεται. δὲ Ἡρόδοτος ἄλλως ἱστορεῖ τὰ κατὰ τὸν Γύγην, ὅτι ἐπιτροπῇ τῆς δεσποίνης ἀπέκτεινε τὸν Κανδαύλην 6 Τύγης καὶ ἐβασίλευσεν.

It will be seen that this note was written entirely for practical purposes. The author explains the Platonic application and points out the origin and meaning of the familiar proverb. He is not affected by the allegorizing of Tzetzes, but on the other hand he also seems to have been sufficiently modern to have quite lost track of the good old tradition, as we saw it for instance in Chennus and Philostratus, according to which Plato and Herodotus really go back ultimately to a common source.

As we look back over this sometimes thin but always persistent literary tradition of more than a millennium, the most notable


feature of it is the fact that with the possible exception of Cicero’s translation, I have been unable to find a single reference which does not go back either directly or indirectly to the school- house. There is something portentous in the length, the strength and the persistence of such a pedagogical tradition. Fancy our eminent educators allowing anything, no matter what it was, to remain in the schools for more than thirty generations! It would be hasty, however, to assume that this extraordinary conservatism was entirely due to the fact that no one had the brains or the energy to think of anything new or better. It was a long, long time before the Imperial system of education ceased. to be distinctly superior in its own particular way to that of any other nation or period.

Finally, it may be worth observing that apart from the trans- lation of Cicero already mentioned, I find no reference to Gyges’ Ring, no sign of familiarity with the story of it, in the entire range of Latin literature. One would have guessed that the paramount authority of a writer like Cicero would have given his version the entrée of the Roman schools. But this does not seem to have been the case.

Let us now investigate and test in the same manner the later tradition of the story told by Herodotus. This, too, begins with Rhetoric. The first, and one of the most important references now surviving, belongs to the Age of Augustus. It is found in Dionysius of Halicarnassus, De Compositione Verborum, § 16.

The author’s main object and the point which he especially desires to make is (8 9) that “it is upon arrangement, far more than upon selection, that persuasion, charm, and literary power depend.” “Every utterance,” he continues (88 11 sqq.), “by which we express our thoughts is either in metre or not in metre. Whichever it be, it can, when aided by beautiful arrangement, attain beauty whether of verse or prose. But speech, if flung out carelessly at random, at the same time spoils the value of the thought. Many poets and prose writers (philosophers and orators) have carefully chosen expressions that are distinctly beautiful and appropriate to the subject matter, but have reaped. no benefit from their trouble because they have given them a rude and haphazard sort of arrangement: whereas others have invested their discourse with great beauty by taking humble, unpretending words, and arranging them with charm and dis-

5 The translation in this and the following sections is that of Roberts.


tinction. It may well be thought that composition is to selection what words are to ideas. For just as a fine thought is of no avail unless it be clothed in beautiful language, so here, too, pure and elegant expression is useless unless it be attired in the right vesture of arrangement.

But to guard myself against the appearance of making an unsupported assertion, I will try to show by an appeal to facts the reasons which have convinced me that composition is a more important and effective art than mere selection of words. I will first examine a few specimen passages in verse and prose. Among poets let Homer be taken, among prose-writers Herodotus: from these may be formed an adequate notion of the rest... .

“There is in Herodotus a certain Lydian king whom he calls Candaules, adding that he was called Myrsilus by the Greeks. Candaules is represented as infatuated with admiration of his wife, and then as insisting on one of his friends seeing the poor woman naked. The friend struggled hard against the constraint put upon him; but failing to shake the king’s resolve, he submitted, and viewed her. The incident, as an incident, is not only lacking in dignity and, for the purpose of embellishment, intractable, but is also vulgar and hazardous and more akin to the repulsive than to the beautiful. But it has been related with great dexterity: it has been made something far better to hear told than it was to see done. And, that no one may imagine that it is to the dialect that the charm of the story is due, I will change its distinctive forms into Attic, and without any further meddling with the language will give the conversation as it stands.”

Dionysius then rewrites Herodotus I 8-10 (Tvyn, οὐ ydpce . . . μὲν δὴ ὡς οὐκ ἐδύνατο διαφυγεῖν, ἕτοιμος ἦν) in Attic and continues:

Here again no one can say that the grace of the style is due to the impressiveness and the dignity of the words. These have not been picked and chosen with studious care; they are simply the labels affixed to things by Nature. Indeed, it would perhaps have been out of place to use other and grander words. I take it, in fact, to be always necessary, whenever ideas are expressed in proper and appropriate language, that no word should be more dignified than the nature of the ideas. That there is no stately or grandiose word in the present passage, any one who likes may prove by simply changing the arrangement. There are many similar passages in this author, from which it can be seen that the fascination of his style does not after all lie in the beauty of the words but in their combination.”

*“The truth seems to be,” says Roberts in an interesting passage


For the purposes of our present inquiry this discussion of Dionysius is very instructive. We may almost begin with the assumption that this passage of Herodotus had already been associated with scholastic rhetoric for an indefinite period. Otherwise a man of the type and time of Dionysius would hardly have used it as an illustration in a technical treatise on rhetoric. By the time of Augustus, the examples and illustrations used by the rhetoricians were for the most part veterans in the service. That this was actually the case with this particular passage is suggested for one thing by the fact that it is such an extraordinarily good example of the λέξις εἰρομέν. And the well-known passage in Aristotle’s Rhetoric (3, 9) in which the author discusses and characterizes the two great types of compo- sition, indicates that even then Herodotus had become the classic of that type. If now we add that as Volkmann observes (Rhetorik?, p. 28), Dionysius as a technical rhetorician harks back to Isocrates, it is at least quite possible that the Herodotean tale of Gyges entered the scholastic tradition of rhetoric at some time between Isocrates and Dionysius.

At any rate—and, after all, that is enough for our present purpose—it actually does appear in a rhetorical treatise of the Augustan Age. There it is used in connection with the claim that composition is more important than selection. This, too, must have been a traditional claim. At all events, it is one which this passage of Herodotus was peculiarly fitted to support, inasmuch as the biblical simplicity of the language used is such a marked contrast to the more or less rare and recherché vocabulary which was cultivated, for instance, by an author like Tacitus, and which was characteristic of rhetoric in general during and after the time of Dionysius himself. Indeed, although it is quite certain that Dionysius thoroughly believes in Herodotus, he, nevertheless, takes up the cudgels for him in a way that almost seems apologetic.”

{Dionysius of Halicarnassus, The Three Literary Letters, p. 11, n. 1), “that, in this instance, the charm lies not so much in the dialect, or jndeed in the vaunted σύνθεσις itself, as in the attitude of the writer’s niind as revealed in the entire narrative, style being interesting (here if anywhere) as the revelation of personality.” Roberts has a similar note in his Ὁ. H. on Literary Composition, pp. 84 sq., where he also bids the reader compare and contrast the narrative of Livy 39, 9.

τοῦ course the Ionic dialect of Herodotus, as Dionysius himself must


This passage of Dionysius besides being of unusual impor- tance in itself is also the only one, so far as I know, in which the Herodotean tale of Candaules is used to illustrate a question of literary style. It will be observed that the portion of the story selected by Dionysius for discussion is the dialogue, not the narrative. This is entirely characteristic of rhetorical training in the schools. It is, therefore, no surprise to find that later references, in so far as they are scholastic in origin, are so largely confined to this particular portion of our story. But before considering these references, let us take up another important discussion of the story as a whole.

This belongs to the fifth century and is found in the Progym- nasmata of the sophist Nikolaus. Long before the time of

have felt, undoubtedly does have a charm of its own, especially in a story like this. If we distrusted our own judgment, we might appeal to such ancient critics as Quintilian, 9, 4, 18, and Hermogenes, De Ideis, 362, 14 Spengel (cp. Norden, Antike Kunstprosa, p. 36). We may grant perhaps that σύνθεσις as Dionysius defines it is superior to selec- tion. We may even grant that his experiment on our passage of Hero- dotus has proved it. Nevertheless the fact still remains that he has failed to prove that the charm of the story is not due to the dialect. The reason—though he himself was apparently quite unaware of the fact—was because neither he nor anyone else could get rid of the Ionic dialect merely “by changing its distinctive forms into Attic, and with- out any further meddling with the language giving the conversation as it stands.”

In its form, as well as in its associations, the Ionic dialect has the dignity, the harmony, the flexibility of the old Epic. Tonic prose is not primitive in the sense of being inartistic. But it is old. Artistically as well as chronologically it is anterior to Attic prose. The same is true of the λέξις εἰρομένη, the type of literary composition—or, as Dionysius would call it, σύνθεσις---οἵ which Herodotus has always been the great classical exemplar. ‘H μὲν οὖν εἰρομένη λέξις, says Aristotle, Rhet. 3, 9, ἀρχαία ἐστίν: ‘Hpoddérov Θουρίου ἥδ᾽ ἱστορίης ἀπόδεξις - ταύτῃ yap πρότερον μὲν ἅπαντες, νῦν δὲ οὐ πολλοὶ χρῶνται. “The λέξις εἰρομένη is the ancient type . . . formerly it was used by everyone, now by comparatively few.” In short, to state it in a slightly different fashion, Aristotle means that the λέξις εἰρομένη---ΟΥἩἉ, as Dionysius might have said, the type of σύνθεσις which suggested that term—is eminently characteristic of Ionic prose as opposed to later Attic prose. Anyone who is really acquainted with modern dialect at first hand, knows that it is characterized by its arrangement of thought quite as much as by its vocabulary. For the whole question of the λέξις εἰρομένη as developed by Herodotus for his special purpose, see Jacoby 5. v. “Herodotus” in PRE, Suppl. II.


Nikolaus, προγυμνάσματα had assumed a very important place in the scheme of education (see above, pp. 2 sq.). Among the most interesting were the practice declamations, more particu- larly the so-called ἀνασκευαί or confutationes (Quintilian 2, 4, 18 etc.). These were given the young students and were sup- posed to be learned by them. The third in the collection of Nikolaus (Rhet. Graeci, I 28% Walz) is entitled:

Ὅτι οὐκ εἰκότα τὰ κατὰ Κανδαύλην

“That the story of Candaules is not credible.”

“There was a time when I had a wonderful opinion of historians as compared with poets; for the object of history is truth, the object of poetry is stories. But now it seems to me that Herodotus differs in no respect from the poets; for he obliterates the distinction between the two, and consequently preserves neither the charm of the metre nor the truth of history. One might criticize him for many things, but especially for the story he has told about Candaules. It runs as follows: Candaules, he says, who was a descendant of Hercules and in love with his own wife, exhibited her to Gyges. For he took Gyges with him into his own palace, stationed him behind the bedroom door, and gave him the opportunity of witnessing the Queen from there. She was aware that he saw her and was highly incensed; but she waited until morning, sent for Gyges and gave him the choice of two things—either to slay Candaules, or if he shrank from it, to be slain himself. Gyges chose to survive, Candaules fell, and marriage with his wife was the reward given for his murder.”

“This is the story as Herodotus tells it. All the statements in it can be picked to pieces in regular order. Candaules is the descendant of Hercules.’ What indications of that pedigree are brought forward? The energy and ambition of Hercules were all in the direction of virtue and his deeds saved Greece; but Candaules had an eye only for pleasures. If he were a descendant of Hercules, how could he so belie his ancestry? How again could Candaules be in love with his own wife? For either he did not live with her or else he did live with her and therefore did not desire her; for intercourse destroys love, and the impulse of desire is killed by marriage. How too could he take Gyges into his palace? The palace was full of guards and crowded with people in every direction. Gyges would, therefore, be dragged off to execution before the King got him to the place proposed, and the trick would come to nought before Gyges saw the woman. And where in those rooms was he stationed for the view? Why, behind the door! If so, he would have escaped notice and therefore would not have seen her. For that which


is hidden from people is itself the first to escape notice. How could he see the woman naked? It was not the custom among the Lydians to strip oneself. Not even the men went without some covering, least of all the women. And why should a woman who is merely going to bed take off all her clothes? Women who derive an income from their favours, even if they were to strip themselves before men, would do so for the purpose of inspiring them with passion. Women who are chaste in their intercourse do not bring themselves to strip for the benefit of their husbands. How then could Gyges be present and look at 8 woman who, even to begin with, had not intended to take off all her clothes? Why did the woman send for Gyges and give him the choice of marriage, if she could not bear his seeing her, and why did she honor as a husband him whom she shrank from having as a spectator at such a time? How could she deliver the kingdom of the Lydians into his hands? Kings are chosen by peoples and by states. I really fail to see then how in the opinion of Herodotus a woman chooses a king and aspires to a fortune which a whole army does not confer. Herodotus ought not to have said these things and such things as these. And when he does say them, all we can do is to disbelieve him.”

This confutatio is carefully worked out in accordance with the rules given for this type of composition by the sophist Theon (Rhet. Graeci I, p. 216 Walz). Some of the arguments touch on themes which had long been familiar to the schools. The reference to the virtue of Hercules, for instance, suggests a discussion which had seldom had an opportunity to rest since the time of Prodicus himself. It has no great value as an argument here, in fact none of the arguments presented here will impress the modern reader as of any great value. Nor indeed did they make any deep impression at that time. Herodotus had long since attained the position of a more or less impeccable classic and therefore no argument against him was taken very seriously. But this had not always been the case. Note, for example, that of all the themes used for confu- tationes by both Theon and Nikolaus, this is the only one taken from history. The rest are all taken from mythology. This in itself would suggest that there was a long tradition of adverse criticism of Herodotus with which the world was fairly familiar. We know that such was actually the case, although little is now left of it except Plutarch’s essay De Herodoti Malignitate. This essay was written by a great man and one who was evidently more nearly in touch than was Theon with a living tradition of


the subject ; but when it comes to the arguments presented, there is little to choose between the two.

Another version of our story as a whole is found in the section given to διηγήματα or rhetorical narrationes in the Progymnas- mata attributed to Libanius (vol. VIII, p. 48 F). The text is as follows:

“Hpa τῆς ἑαυτοῦ γυναικὸς 6 Κανδαύλης καὶ παρεκάλει tov Γύγην ἐπὶ τὴν θέαν τῆς ὧρας. δὲ τὸ πρῶτον ἀρνούμενος ἐγκειμένου τοῦ Κανδαύλον συνεχώρησεν. ὕφ᾽ οὗ δὴ καὶ καταστὰς ὄπισθεν τῆς θύρας τὴν γυναῖκα καταγυμνουμένην ἰδὼν ἀπηλλάγη. δὲ μεταστραφεῖσα τὸ πραχθὲν οὐκ ἠγνόησεν, ἤνεγκε δὲ σιγῇ. μεταπέμπεται δὲ τὸν Τύγην, ἐπειδὴ ἡμέρα ἦν, καὶ ἐκέλευσεν ἀποθνήσκειν ἀντὶ τῆς θέας τοῦτο δρᾶν τὸν Κανδαύλην ὑπισχνουμένη συνοικήσειν αὐτῷ μετὰ τὸν φόνον. τὸν Γύγην ἤρεσκε μὲν οὐδέτερον, εἰς δὲ τὸ κτείνειν ἀπέκλινε. καὶ διαχρησάμενος καθεύδοντα τὸν δεσπότην γαμεῖ τε ἐκείνην καὶ βασιλεύει Λυδῶν.

The version of the scholiast on Alius Aristides, XLV, 56 (III, p. 411 Dindorf) was, so to speak, a mere matter of busi- ness, but it is a good example of the type of rhetorical narratio just quoted :

Κανδαύλης Λυδῶν ἦν βασιλεύς, παγκάλην ἔχων γυναῖκα ' νόμου δὲ ὄντος, μή τινα τῶν ἔξωθεν ὁρᾶν τὰς βασιλίδας, 6 Κανδαύλης ἐνέκειτο βιάζων τὸν Τύγην εἰς θέαν τῆς γυναικός, ὑπηρέτην ὄντα αὐτῷ - δὲ τὴν μὲν πρώτην ἀπεπήδα, χρόνῳ δὲ ὑπείξας τῷ Κανδαύλῃ βιάζοντι, καὶ εἶδε τὴν αὐτοῦ δέσποιναν. αὕτη οὖν λάθρα τουτονὶ μεταστειλαμένη, θνήσκειν αὐτόν, κτείνειν τὸν δεσπότην ἔλεγε" καὶ ὃς αἱρεῖται τὸ δεύτερον, καὶ ταύτην γαμήσας βασιλεύει Λυδῶν.

We have next to consider the political verses of JIoannes Tzetzes, Chiliades, I 137-166 and VII 195-202, the text of which has already been given above.

Finally, and this is almost the last word in ancient literature, Georgios (born 1241), later known as Patriarch Gregorios, who, it seems, was deeply interested in elementary education, com- posed a school-book (preserved in Harleianus 5735 and other MSS). True to the pedagogical tradition which had prevailed for more than a millennium, it consists of a prose paraphrase of ZEsopic fables, and some mythological pieces, among the rest the story of Iphigenia, of Aineas, of Pandarus and Diomedes, and of Candaules and Gyges. (See Krumbacher, Gesch. der byzan- tinischen Litteratur, 2d. edit., Munich, 1897, p. 477.) The persistence of our story at this late date shows in itself that it had long been familiar to the schools. How familiar it was,


and how persistent the scholastic tradition of it was, is shown by the fact that so far as rhetorical narrationes are concerned, it is one of the rare exceptions to the rule of a mythological rather than a historical or quasi-historical subject. In the forty- odd narrationes of Libanius, for example, this story and two others are the only exceptions. Even in the confutationes and refutationes of Theon and his successors, the same rule holds good.

Such is the tradition of the entire story. It was characteristic, persistent and, so far as we can see, entirely scholastic. But this was only one aspect of the tradition. The passage, for example, in Ptolemeus Chennus (p. 192 W), already referred to above, shows that the Herodotean tale of Gyges was quite as much a subject of literary chit-chat in the First Century as was Plato’s story of the Ring. It follows, therefore, that it had long been familiar to the Rhetorical Schools.

But the longest and perhaps the most important chapter in the tradition of this passage is concerned with two phrases. Both are found in the dialogue between Gyges and Candaules. The fact also that they are both sententious explains why they, and incidentally the dialogue in which they occur, were referred to so much oftener in the later tradition than anything else in the story. One of the notable features in the growth of rhetoric and rhetorical study under the Empire was the increasing fond- ness for sententia—using that word in the sense of sayings of general application—sometimes proverbial but not necessarily so.2 Tacitus, as everyone knows, is famous for them and, as we shall see later, Herodotus was greatly admired for his skill in making them spring naturally from the context.

Turning now to the first of the two phrases which we have to consider, Herodotus makes Candaules say, Gyges, when I tell thee of my wife’s beauty, methinks thou dost not believe me (in fact men’s ears are naturally less trustworthy than their eyes). ὦτα yap τυγχάνει ἀνθρώποισι ἐόντα ἀπιστότερα ὀφθαλμῶν. “Seeing is believing,” to use the parallel phrase in English. The thought was of course not new. Indeed, the artistic value and fitness of it in this particular connection are due to the fact not

* Ernesti, Lex. Techn. 8. v.; Seneca, Controv. I, Praef.; Quint. IV 2, 121; Theon, I, p. 200 W.


only that it was not new but that it was a commonplace familiar to everyone.

So far as Greece is concerned, however, the only notable occur- rence of the thought, before Herodotus, seems to be in a frag- ment of Heraclitus quoted by Polybius 12, 27, 1.2 The passage teads:

δυεῖν yap ὄντων κατὰ φύσιν ὡσανεί τινων ὀργάνων ἡμῖν, ols πάντα πυθανόμεθα καὶ πολυπραγμονοῦμεν, ἀκοῆς καὶ ὁράσεως, ἀληθινωτέρας δ᾽ οὔσης οὐ μικρῷ τῆς ὁράσεως κατὰ τὸν Ἡράκλειτον - ὀφθαλμοὶ γὰρ τῶν ὥτων ἀκριβέστεροι μάρτυρες. Ἡράκλειτον here was changed to Ἡρόδοτον by Leutsch, etc., but Ἡράκλειτον is the reading of the MSS, and there is no good reason for doubting it.

Sophocles, (Βα. Tyr. 1237, αὐτὴ πρὸς αὑτῆς. τῶν δὲ πραχθέντων τὰ μὲν ἄλγιστ᾽ ἄπεστιν * γὰρ ὄψις οὗ πάρα, though sometimes quoted in this connection, is hardly parallel. Latin cognates are fairly numerous,’® but the only passage which one might suspect of being an echo of the Herodotean phrase is Horace, Ars Poetica, 180:

Segnius irritant animos demissa per aurem Quam quae sunt oculis subiecta fidelibus, et quae Ipse sibi tradit spectator.

On the Greek side, it is again Dionysius of Halicarnassus—if indeed Dionysius is the author of the following passage—who furnishes the first reference, Rhetoric, 11, p. 401: 11

Furthermore, figures of speech also indicate the distinctive quality of the barbarian mind, as was undoubtedly the case when Herodotus makes Candaules say to Gyges in the course of his

* Frag. XV Bywater; frag. 1018 Diels. See Diels’ note and especially R. von Scala, Studien des Polybios, Stuttgart, 1890, I, pp. 88 ff.

19 Plautus, Asin. 202: Semper oculatae manus sunt nostrae, credunt quod vident; Plautus, Truc. 490 (also quoted by Apuleius, Flor. 2 and Festus, 179M): Pluris est oculatus testis unus quam auriti decem; ference, Eun. 350: Vidi, novi; Seneca, N.Q., 4, 3, 1: Itaque ex his me testibus numero secundae notae, qui audivisse quidem se, vidisse negant, ete.; Seneca, Epist. 6, 5: Homines amplius oculis quam auribus credunt; Hieronymus, Epist. 64, 10: Multoque plus intellegitur quod oculis vide- tur quam quod aure percipitur. Cicero, De Orat. 3, 161, though quoted in this connection, is not in point.

11] doubt whether Strabo 2, 5, p. 117 is in any sense an echo of Hero-



conversation with him: “In fact, men’s ears are naturally less trustworthy than their eyes.” For he did not speak of hearing’ and ‘sight,’ but transferred the thought to the parts of the body concerned.

It is quite true that a large use of figurative speech, especially in ordinary conversation, is more or less characteristic of the barbarian mind. But the long tradition of this particular use in Greek itself, beginning as we have seen as early at least as Heraclitus, suggests that Dionysius might have done better to select some other example. This, however, is a point with which we are not directly concerned.

Chronologically the first to consider after Dionysius is Philo Judaeus. He displays an extraordinary fondness for this thought, but, after a careful examination of his entire works, I can give no example which seems to be suggested by our phrase.

We now come to Lucian—in the discussion of a question like