At the beginning of Plato’s Gorgias, Socrates wishes to ask the sophist what he professes and teaches, and especially what the power of his art is (τίς ἡ δύναμις τῆς τέχνης τοῦ ἀνδρός, 447c1–2). Socrates does not address Gorgias directly, however; instead, he invites his friend Chaerephon to pose the question. As Chaerephon proceeds to ask what Gorgias professes to be, Polus intervenes to answer in his place. Chaerephon and Polus agree, first, that a practitioner of the same art (τῆς τέχνης, 448b5) as Herodicus is to be called a physician, and second, that a practitioner of the same art (τέχνης, 448b12) as Aristophon is to be called a painter. Then Chaerephon asks: νῦν δ’ ἐπειδὴ τινὸς τέχνης ἐπιστήμων ἐστίν, τίνα ἂν καλοῦντες αὐτὸν ὀρθῶς καλοῖμεν; (448c2–3), ‘Since Gorgias has knowledge of a specific art, what is the right name we should call him?’ In other words, Chaerephon assumes that Gorgias is skilled in a particular art and wishes to know what designation Gorgias claims for himself.

This flawless manuscript text is not what editors print. Ever since Heindorf,Footnote ¹ all editions of the Gorgias read νῦν δ’ ἐπειδὴ τίνος τέχνης ἐπιστήμων ἐστίν, τίνα ἂν καλοῦντες αὐτὸν ὀρθῶς καλοῖμεν; (448c2–3),Footnote ² as if Chaerephon were asking not one question, but two. This text is reflected in the translations of the passage, for example: ‘In fact, though, what craft has he knowledge of, and what would we rightly call him?’ (Irwin).Footnote ³ But Heindorf’s suggestion is grammatically and stylistically questionable, and the reasons that he provides for his reading are ultimately unconvincing.

Heindorf’s proposal contains two interrogative clauses placed in asyndetic parataxis: the first with a predicate in the indicative mood, and the second with the predicate in the optative mood accompanied by ἄν. Such a construction seems rather odd, to say the least. Therefore, its proponents must demonstrate (a) that it is grammatically admissible, and (b) that it aligns with Plato’s style. Moreover, Plato never introduces a main interrogative clause with ἐπειδή.Footnote ⁴ According to the TLG database, that word appears 422 times across Plato’s dialogues, yet nowhere does it introduce a direct question. A possible exception may be found in Alcibiades 1, a dialogue of disputed authenticity, where the manuscript text reads: ὦ Ἀλκιβιάδη, ἐπειδὴ περί τινος Ἀθηναῖοι διανοοῦνται βουλεύεσθαι, ἀνίστασαι συμβουλεύσων; (106c6–8). On this passage, Henricus Stephanus commented: ‘For(tasse) ἐπειδὴ περὶ τίνος scribendum est: hoc sensu, τί ἐστι περὶ οὗ ἐπειδὴ Ἀθ(ηναῖοι) διαν(οοῦνται) βουλεύεσθαι, ἀνίστασαι συμβουλεύσων; (ut paulo post dicit ὅταν οὖν περὶ τίνος βουλεύωνται, pro τί οὖν ἐστι περὶ οὗ ὅταν βουλεύωνται). Quod si retinendo περί τινος, praefigatur καὶ verbo ἀνίστασαι, ne sic quidem cohaerebunt satis quae sequuntur.’Footnote ⁵ Subsequent editors have followed this suggestion, printing περὶ τίνος instead of περί τινος.

Let us examine Heindorf’s arguments for his proposal: ‘Vulgo τινὸς, quo mendo simillimum locum Stephanus liberavit Alcib. I § 7 ὦ Ἀλκιβιάδη, ἐπειδὴ περὶ τίνος Ἀθηναῖοι διανοοῦνται βουλεύεσθαι, ἀνίστασαι συμβουλεύσων; similiterque de rep. I. p. 232Footnote ⁶ c. et d. bis scribendum, ἡ – τίσι τί ἀποδιδοῦσα τέχνη etc. Conf. not. ad Hipp. Mai. § 18.’Footnote ⁷ Heindorf thus appeals to three parallel passages.

The first is Alc. 1 106c6–8. As explained above, the ἐπειδή-clause in this passage is not interrogative in the manuscripts and Heindorf’s point is that Grg. 448c2 should be reinterpreted because (a) the reading suggested by Stephanus at Alc. 1 106c6–8 makes sense and (b) the two loci are simillimi. Against this argument, the following aspects should be considered. (i) Alc. 1 106c6–8 indeed seems to invite editorial intervention, as Stephanus noted, because in the very next sentence (ἆρ’ ἐπειδὴ περὶ ὧν σὺ ἐπίστασαι βέλτιον ἢ οὗτοι; 106c8–9) Socrates answers—in interrogative form—precisely the question περὶ τίνος. By contrast, Grg. 448c2, as shown above, is perfectly intelligible as it stands; neither the context nor the argument require any intervention. (ii) The reading of Alc. 1 106c6–8 represents a stylistic anomaly: as mentioned above, Plato never introduces a main interrogative clause with ἐπειδή elsewhere in the corpus. Therefore, this passage, the authenticity of which is more disputed now than in Heindorf’s time, provides a weak basis for altering the text in Grg. 448c2. (iii) Apart from the (minor) fact that both instances of ἐπειδή are followed by an indefinite/interrogative pronoun, the two passages do not share a comparable syntactic structure. In Alc. 1 106c6–8, ἐπειδή functions as a genuine conjunction, introducing a clause subordinated to the main clause. By contrast, Heindorf’s reading of Grg. 448c2 produces a parataxis in which two main clauses—with predicates in different moods—are introduced by a false conjunction. It is therefore unclear why Heindorf considers the two passages simillimi. (iv) There are numerous passages in Plato where ἐπειδή introduces a causal clause preceding a main interrogative clause, for example Menex. 82d1–2 ἐπειδὴ δὲ δυοῖν ποδοῖν καὶ ταύτῃ, ἄλλο τι ἢ δὶς δυοῖν γίγνεται; Grg. 462b4–5 ἐπειδὴ Γοργίας ἀπορεῖν σοι δοκεῖ περὶ τῆς ῥητορικῆς, σὺ αὐτὴν τίνα φῂς εἶναι; Tht. 197a3–4 ἐπειδὴ οὖν ἐσμεν φαῦλοι, βούλει τολμήσω εἰπεῖν οἷόν ἐστι τὸ ἐπίστασθαι; There is a strong case to be made that these and other such passagesFootnote ⁸ are more similar to Grg. 448c2 than Alc. 1 106c6–8 and speak against Heindorf’s reading.

The second parallel passage mentioned by Heindorf is Resp. 332c–d ὦ Σιμωνίδη, ἡ τίσιν οὖν τί ἀποδιδοῦσα ὀφειλόμενον καὶ προσῆκον τέχνη ἰατρικὴ καλεῖται; Here the meaning is τί ὀφειλόμενον καὶ προσῆκον ἀποδίδωσι τίσιν ἡ τέχνη ἣ ἰατρικὴ καλεῖται; Again, it is unclear what this passage, where the main question is embedded between article and noun, has in common with Grg. 448c2, or why it should be taken to support Heindorf’s reading.

Heindorf’s third parallel passage is Hp. mai. 288a8–9, where the manuscript text reads ταῦτα πάντα ἃ φῂς καλὰ εἶναι, εἴ τι ἐστὶν αὐτὸ τὸ καλόν, ταῦτ’ ἂν εἴη καλά; Heindorf read here … εἰ τί ἐστιν αὐτὸ τὸ καλόν …, explaining in a footnote: ‘Quod rescripsi εἰ τί, est ex certissima Schleiermacheri emendatione. Gorg. p. 448. C. Νῦν δ’ ἐπειδὴ τίνος (sic scr[ibendum]) τέχνης ἐπιστήμων ἐστί, τίνα ἂν καλοῦντες αὐτὸν ὀρθῶς καλοῦμεν; in quibus ex vulgato illo Graecorum more duplex est interrogatio, in unam coartata propositionem.’Footnote ⁹ This suggestion has also been adopted by all subsequent editors. But as with Alc. 1 106c6–8, Heindorf’s reading is prompted by the context. The quoted question is followed by an answer (ἐγὼ δὲ δὴ ἐρῶ ὅτι εἰ παρθένος καλὴ καλόν, ἔστι δι’ ὃ ταῦτ’ ἂν εἴη καλά, 288a9–11) which presupposes a previous εἰ τί ἐστιν αὐτὸ τὸ καλόν.

Heindorf’s interpretation of Grg. 448c2 also found its way into Greek grammars. Kühner–Gerth cite both Alc. 1 106c6–8 and Grg. 448c2 as examples of direct questions embedded in syntactic structures introduced by subordinating conjunctions or relative pronouns.Footnote ¹⁰ Yet none of the other examples they quote is truly similar to Grg. 448c2. Certainly, even if an example of two paratactic main interrogative clauses introduced by a conjunction could be found in Greek literature, this would not justify Heindorf’s reinterpretation of an otherwise clear and flawless text.