A passage near the start of the Rhetorica ad Herennium may well cause the reader to stumble (1.2):Footnote ¹

It is the task of the orator to speak about public matters and the law, obtaining the agreement of the audience as far as possible. The last seven words of this passage were translated by Harry Caplan in his Loeb edition as ‘and to secure as far as possible the agreement of his hearers’, and by Gualtiero Calboli in his recent edition with commentary as ‘e parlare riscuotendo, per quanto sarà possibile, l'approvazione degli ascoltatori’.Footnote ² According to Sext. Emp. Math. 2.62 (pages 96,29–97,1 Mau), this definition goes back to the second-century b.c.e. teacher of rhetoric Hermagoras of Temnos: ῾Ερμαγόρας τελείου ῥήτορος ἔργον εἶναι ἔλεγε τὸ τεθὲν πολιτικὸν ζήτημα διατίθεσθαι κατὰ τὸ ἐνδεχόμενον πειστικῶς. None of these versions has any place for eius, which fits poorly into the syntax. How can this genitive be attached not to a noun or pronoun but to the prepositional phrase quoad?

However, the standard dictionaries treat quoad eius as a proper Latin phrase. The OLD (s.v. 3a) notes that quoad can mean ‘To the degree that, as far as, as much as’ and mentions its use with a partitive genitive, giving the present passage as the only example; it adds the comment that ‘cod[ices] s[ome]t[ime]s vary between this and quod’. Lewis and Short (s.v. quoad B2) translate ‘So far as, as much as’ and note the usage with eius in the phrase quoad eius facere possum to mean ‘as far’ or ‘as well as I can’. They refer to parallels at Cic. Att. 11.12.4 (noting the variant quod eius), Fam. 3.2.2, Inu. rhet. 2.20 and Livy 39.45.7. But if one consults recent critical editions of all passages save Rhet. Her. 1.2, one finds that all of them read quod eius!

Two grammars of classical Latin also enter the fray. Hofmann–Szantyr regard both quod eius and quoad eius as genuine phrases, present in early and colloquial Latin, which we cannot tell apart because of the vagaries of the transmission.Footnote ³ According to Kühner–Stegmann on the other hand, quoad eius was used in the phrases quoad eius facere possum and quoad eius fieri potest, but it has been corrupted almost everywhere to quod eius.Footnote ⁴ This implies that in other kinds of phrases quod eius is (or may be) authentic.

Both grammars refer to the scholarly debate of the late nineteenth and early twentieth centuries.Footnote ⁵ A key contribution was a detailed discussion by Heinrich Jordan, who reached rather different conclusions from what has been presented so far.Footnote ⁶ Jordan's principal interest was not textual criticism but the history of the Latin language. He studied how the pronoun quod came to be used gradually as a conjunction, and paid close attention to the construction quod eius. In the language of Roman law, this phrase often introduced limiting clauses. The origins of the construction are clarified by passages where the antecedent to eius is repeated after it in the genitive, as in the Lex agraria of 111 b.c.e. at CIL I² 585.25 (Crawford, Roman Statutes no. 2.24) [ager locus quei sup]ra screiptus est, quod eius agrei locei post <h(anc)> l(egem) r(ogatam) publicum populei Romanei erit and with the spelling variant quot in the Lex Coloniae Genetiuae Iuliae siue Vrsonensis of 47–44 b.c.e. at par. LXXII (Crawford, Roman Statutes no. 25 tablet b col. I.30–5) quotcumque pecuniae stipis nomine in aedis sacras datum inlatum erit, quot eius pecuniae eis sacris superfuerit, quae sacra … facta <fuer>i<nt>, ne quis facito … quo minus in ea aede consumatur. Jordan also quotes variants in which eius lacks a specific antecedent: thus in the same Lex Coloniae Genetiuae at par. LXXVII (Crawford, Roman Statutes no. 25 tablet b col. II.29–33) si qu<a>s uias fossas cloacas IIuir aedil(is)ue publice facere … munire intra eos fines, qui colon(iae) Iul(iae) erunt, uolet, quot eius sine iniuria priuatorum fiet, it is facere liceto. As for the literary attestations of quod eius and quoad eius, Jordan states that the latter is not supported by strong manuscript evidence; at Rhet. Her. 1.2 and Cic. Inu. rhet. 2.20 (which are the only attestations that he discusses) it must be a corruption of quod eius.

In sum, the scholarly debate about this complex problem has not come to a close. I will summarize the three hypotheses that have been put forward, adding a fourth position for the sake of logical symmetry:

1. 1. Both quoad eius and quod eius are authentic Latin phrases. (This position is taken explicitly by Hofmann–Szantyr, and it is implied by the standard dictionaries.)

2. 2. quoad eius facere possum / quoad eius fieri potest is authentic, but it has been corrupted almost everywhere to quod eius. (This is the position of Kühner–Stegmann, who do not call into doubt the authenticity of quod eius in other constructions.)

3. 3. quoad eius is authentic; quod eius is a corrupt form that derives from quoad eius. (This has not been proposed so far by anyone, as far as I am aware.)

4. 4. quod eius is authentic; quoad eius is a corrupt form that derives from quod eius. (This is the view of Jordan.)

What sources support either phrase? Here follows a brief survey, grouped according to the forms that are attested in each passage.Footnote ⁷ Since any hypothetical reconstruction must be based on the evidence of the sources, textual conjectures are not taken into consideration at this stage. This leads to the omission of two of the three parallels adduced for quoad eius by Lewis and Short—namely, Cic. Att. 11.12.4, Fam. 3.2.2 and Livy 39.45.7—as there appears to be no manuscript evidence for the reading quoad eius in these passages.

For the sake of brevity, I focus on those parallels in which eius is not accompanied by a noun; adding those passages in which it is accompanied by a noun would increase the number of attestations of quod eius. Of course, I do not include passages in which eius does not depend on quod, such as CIL XI 600.13 ob merita quod eius mortem dolui.

1. i. quod eius (occasionally written as quot eius in the Lex Coloniae Genetiuae Iuliae) transmitted unambiguously: 35x in all.

Attested 26x in epigraphic sources: in the Lex repetundarum, possibly of 123–122 b.c.e., CIL I² 583 (Crawford, Roman Statutes no. 1), at 67; in the Lex agraria of 111 b.c.e., CIL I² 585 (Crawford, Roman Statutes no. 2), at 5 (ter), 33, 38, 64, 65, 66, 67, 75 and 80; in the Lex Municipii Tarentini of c.80 b.c.e., CIL I² 590 (Crawford, Roman Statutes no. 15), at col. I.41; in the Lex Antonia de Termessibus, possibly of 68 b.c.e., CIL I² 589 (Crawford, Roman Statutes no. 19), at col. I.32 and col. II.25; in the Lex Coloniae Genetiuae Iuliae siue Vrsonensis of 47–44 b.c.e. at paragraphs 13 (L'Année Épigraphique 2006, no. 645), 70 (Crawford, Roman Statutes no. 25 tablet b col. I.9), 71 (tablet b col. I.22), 72 (tablet b col. I.31), 77 (tablet b col. II.32), 80 (tablet b col. III.12), 102 bis (tablet c col. IV.30 and 32) and 128 (tablet e col. II.13); in the edict of Emperor Augustus found at Venafro (CIL X 4842.26); in the Lex Flauia Malacitana of 81–96 c.e. at 27 (CIL II 1963 col. II.16–17).

Attested 9x in literary sources: Cato, Agr. 32.1 and 33.1; Cic. Fam. 3.2.2 and 5.8.5; Att. 11.12.4; Q. Cicero (?), Comment. pet. 43; Livy 39.45.7 and 42.8.7; Apul. De deo Soc. 1.1.

1. ii. quoad eius transmitted unambiguously (but see below): attested 1x in literary sources only, at Rhet. Her. 1.2.

2. iii. quod eius and quoad eius transmitted in parallel, with both forms present in the manuscripts (but see below): attested 2x in literary sources only, at Cic. Inu. rhet. 2.20 and Q. Cicero (?), Comment. pet. 36.

quod eius is attested over twenty times in seven different laws and edicts, which have reached us on inscriptions in Spain and Italy; and it is attested nine times in literary texts and private letters by five different authors. It is unlikely in the extreme that all these attestations, especially those in the inscriptions, should be the results of textual corruption. This support for quod eius in our sources is matched by the convincing linguistic explanation of the phrase that has been put forward by Jordan, as we have seen. In short, quod eius is a well-documented and understandable Latin phrase. We must rule out hypothesis 3 above.

What about the sources that transmit quoad eius? Group iii above comprises two passages where this form stands in some manuscripts, while other sources read quod eius. In both cases, the sources that read quoad eius carry little weight; the authoritative textual witnesses of both passages read quod eius. At Cic. Inu. rhet. 2.20 quo(a)d (eius) fieri possit, the authoritative manuscripts and the lemmata in the commentary of Marius Victorinus read quod eius, while quoad eius and quoad are found in some of the more recent manuscripts known as the integri; the reading quoad eius was added by the second hand to the codex Sangallensis 820 and quoad was added by the third hand to Parisinus lat. 7774a.Footnote ⁸ At Comment. pet. 36 quo(a)d eius fieri poterit, the authoritative manuscripts read quod eius, while quoad eius is attested in the codices recentiores.Footnote ⁹ In sum, the transmitted reading in both passages is quod eius. The variant quoad eius may well have entered the manuscript tradition of these two passages from Rhet. Her. 1.2, a passage containing a key definition at the start of a work that was read fairly often during the Middle Ages.

Hence, Rhet. Her. 1.2 is the only passage where the authoritative textual witnesses support the reading quoad eius. This phrase is not only unparalleled but also awkward: it has already been noted how unusual it is for the genitive eius to be attached to the prepositional phrase quoad, to which it does not add anything, since quoad already means ‘so far as’ in and of itself. Contrast quod eius, where the genitive depends on quod and the meaning is clear: ‘that [part] of it which’, ‘to the extent that’.

Since quoad eius lacks linguistic credibility, it is very likely to be corrupt even here. It is likely to have arisen as a conflation of the phrases quoad and quod eius, both of which are attested at the start of similar clauses in Latin texts of this period. Here quod eius has been restored by Jordan, while quoad finds a precedent in a manuscript: Karl Ludwig Kayser's ‘Emmeranus 2’, which is identified by Ruth Taylor as Munich, Bayerische Staatsbibliothek, Clm 14601 (formerly Regensburg, St Emmeram, F 104).Footnote ¹⁰

Which of these two forms is more likely to be original? It is not easy to decide, given that both are attested in this kind of context. At Rhet. Her. 1.2, the transmitted text is quoad eius fieri poterit. The same phrase is attested with quod eius at Comment. pet. 36, while quod eius fieri possit is read at Cic. Inu. rhet. 2.20 and Fam. 5.8.5. On the other hand, quoad fieri potest is used by Cic. Timaeus 50 and quoad fieri poterit at Att. 8.2.2. The Rhetorica ad Herennium yields some less close parallels for quoad (4.34 quoad possem, 4.48 quoad potestis), but none for quod eius. That constitutes one argument in favour of reading quoad. Another argument can be drawn from the context: at the start of a clause that makes a general limitation rather than dividing up a specific entity, quoad is more apt than quod eius, especially in an author with a pedantic eye for precision. On the other hand, it is perhaps easier to explain quoad eius as a result of corruption from quod eius than to derive it from quoad; but the introduction of eius under the influence of the phrase quod eius is not unthinkable. On balance, I prefer quoad fieri possit, as it follows more closely the usus scribendi of the author.