In this short note I would like to reflect again upon paragraphs 39–40 of Epicurus’ Letter to Herodotus and, more specifically, on the text printed in the last (superb and now indispensable)Footnote ¹ critical edition by Tiziano Dorandi,Footnote ² which I quote below together with the critical apparatus (758–9, lines 495–502):

In particular for Book 10 of Diogenes Laertius’ Lives of Eminent Philosophers, which—as is well known—includes the Life of Epicurus, Dorandi establishes a precise editing criterion based on a firm distinction between Epicurus and Diogenes Laertius. To be more precise, Dorandi distinguishes between the ipsissima verba Epicuri and the text of Epicurus that Diogenes Laertius used for the composition of the Life of Epicurus. From this point of view, specifically with respect to Diogenes’ Book 10, Dorandi's aim is to edit Diogenes Laertius and not Epicurus. On the basis of this innovative editing criterion, in the case of a very difficult work such as the Letter to Herodotus, Dorandi is interested in restoring not the text of Epicurus but the ‘Epicurean edition’ or, better, the manuscript of Epicurus that passed through the hands of Diogenes Laertius. Obviously this ‘edition’ of Epicurus could be different from the text directly written by Epicurus.Footnote ³

Paragraphs 39–40 of the Letter to Herodotus have been widely discussed in the secondary literature; here I will focus exclusively on Dorandi's text. Dorandi decides to fill the lacuna of § 39 (lines 496–7) with σώματα καὶ κενόν (Von der Mühll) on the basis of Pierre Gassendi's conjecture πῇ μὲν σῶμα, πῇ δὲ κενόν, thereby rejecting Hermann Usener's σώματα καὶ τόπος. There are several reasons in support of either Gassendi's or Usener's conjecture, but, as I have already said, I will focus only on the text established by Dorandi. At the beginning of § 40 (line 499), Dorandi, unlike in the previous case, decides to follow Usener and writes τόπος δὲ, while part of the manuscript tradition has τὸ πρόσθε (B¹), τὸ πρόσθεν (B² P Φ) and πρόσθεν (F). Unlike Usener, however, Dorandi does not then write ὃν κενὸν and so on but ὃ κενὸν, by removing also the comma before the relative. Usener, who is editor of Epicurus,Footnote ⁴ is certainly consistent in the construction of his text, since in the lacuna of § 39 he writes τόπος, that is, the same term that he conjectures at the beginning of § 40, by considering τὸ πρόσθε (τὸ πρόσθεν or πρόσθεν) as a redundant addition compared to ὥσπερ προεῖπον.Footnote ⁵

It is useful to remember that Usener's text is also welcomed by David Sedley (to whom I am grateful), who translates it as follows:

Dorandi cannot accept such a translation, since in Usener there is ὃν, which refers to τόπος and is the object of ὀνομάζομεν, while in Dorandi one reads ὃ, which for obvious reasons linked to gender cannot directly refer to τόπος. Dorandi's translation of lines 499–500, therefore, could and should read as follows (my translations):

or:

While ὃ is readable in Φ (Vaticanus gr. 96, ff. 29v–88r, XII cent.), ὃν is probably present in B¹ (B = Neapolitanus III B 29, XII cent.).Footnote ⁷ In B² (the corrector of B) there is a correction in rasura (ὃ_*); it is likely that the ν of ὃν has been scratched out; it is very difficult for me to understand the precise reasons for this correction. Finally, in P (Parisinus gr. 1759, XI–XII cent.) and in F (Laurentianus 69.13, XIII cent.) one reads ὂν (with the smooth breathing).

I believe that the right reading is the ὃν of B¹, which in my opinion is to be preferred regardless of whether one wants to edit Epicurus or Diogenes Laertius.Footnote ⁸ In addition to Dorandi, Peter Von der Mühll and Miroslav Marcovich print the same text (τόπος δὲ εἰ μὴ ἦν ὃ κενὸν καὶ χώραν καὶ ἀναφῆ φύσιν ὀνομάζομεν): this would confirm that the Greek is correct. Nevertheless, what does not convince me is the strange ordo verborum of this text: the Greek might work but only in a rather strained manner. Usener has no doubt that ὃν, preceded by a comma, refers to τόπος, which, in his opinion, is together with σώματα one of the constitutive principles of τὸ πᾶν. Hence, having said that everything is body and place, Epicurus would be explaining that by τόπος he means κενόν, χώρα and ἀναφὴς φύσις.Footnote ⁹ By the way, I believe that it is necessary to fill the lacuna in § 39 with σώματα καὶ κενόν when editing Epicurus’ ipsa verba, since (among other reasons) the two terms σώματα and κενόν are explicitly quoted together shortly afterwards, at the end of § 41, where—by good luck—there are no textual problems (καὶ μὴν καὶ τῷ πλήθει τῶν σωμάτων ἄπειρόν ἐστι τὸ πᾶν καὶ τῷ μεγέθει τοῦ κενοῦ; my emphasis).Footnote ¹⁰ Given this integration, τόπος appears for the first time only at the beginning of § 40. If τόπος is printed here, one must conclude that Epicurus is interested in explaining mainly what place is. For this reason, if τόπος δὲ εἰ μὴ ἦν is considered correct, by assuming that Epicurus’ main interest is to clarify the meaning of place (not least because of the position of τόπος at the beginning of the sentence),Footnote ¹¹ immediately afterwards it is necessary to read ὃν and not ὃ also because Epicurus has never mentioned κενόν, χώρα and ἀναφὴς φύσις before. If, by complying with the conceptual and doctrinal consistency of the passage under examination, the editor of Epicurus maintains (with Usener) that Epicurus wrote τόπος δὲ εἰ μὴ ἦν, he/she must print ὃν immediately afterwards.

I believe that much the same applies to Diogenes Laertius. Per litteras David Sedley writes: ‘if one assumes the corruption process proposed by Usener, it seems clear that the change of ὃν tο ὃ happened after the change of τόπος δὲ to τὸ πρόσθε (and the latter's later normalization to τὸ πρόσθεν). Since we can see this progressive corruption occurring in the Borbonicus, there is every reason to assume that the transmitted text of DL [sc. Diogenes Laertius] had ὃν. […] There is no reason to consider the alternative possibility that the relative pronoun has been attracted to the gender of the following κενόν, for which I can find no parallels in Widmann [sc. H. Widmann, Beiträge zur Syntax Epikurs (Stuttgart/Berlin, 1935)].’ Although it is impossible to have any absolute certainty on the matter, it cannot be ruled out that the text of the model used by Diogenes Laertius did not correspond to the ipsissima verba Epicuri and that, therefore, in the copy of Epicurus’ letter used by Diogenes Laertius one should read τόπος δὲ εἰ μὴ ἦν ὃ and so on. Hence, the editor of Diogenes Laertius surely has the right to print this text. Anyway, I take the view that the ordo verborum of this text is rather strange, and its meaning basically unacceptable; furthermore, it does not seem to me that it is possible to find parallels of this kind in Epicurus.

Finally, two further points cannot be left out: in a perspective (rightly) inspired by textual conservatism, (1) the oldest continuous manuscripts of the Lives (B, P and F) have either ὄν (P F) or (probably) ὃν (B¹); (2) τόπος δὲ is a conjecture by Usener, while in the manuscripts one reads either τὸ πρόσθε (B¹) or τὸ πρόσθεν (B² P Φ) or πρόσθεν (F).Footnote ¹² In truth, however, the ὃν of B¹, as Dorandi scrupulously reports in the critical apparatus, is a probable but not certain reading of the manuscript. Be that as it may, it seems to me plausible to conclude that, regardless of whether one is editing Epicurus or Diogenes Laertius, by bearing in mind that ὃν is in all likelihood present in one of the most authoritative manuscripts, if one accepts τόπος δὲ εἰ μὴ ἦν, because of the rather controversial ordo verborum, it is more compelling to print ὃν rather than ὃ, as already consistently done by Usener.