Dice and Facie: Quintilian, Institutio Oratoria 1.7.23 and 9.4.39

In his discussion of orthography in book 1 of his Institutio Oratoria Quintilian mentions several examples of archaic spelling conventions, among them a practice of the Elder Cato, which I present here with readings I shall propose and defend during the course of this essay:

Non Cato Censorius dicam et faciam dice et facie scripsit eundemque in ceteris quae similiter cadunt modum tenuit? (Inst. 1.7.23 5 Cato, Inc. Lib. fr. 55 Jordan, p. 90)

Dice et facie is the reading of one manuscript (N); two other manuscripts (BH) have dicae et faciae; the other major manuscripts (APMS) have the similar but obviously corrupt dice et face. The vast discrepancies between the readings accepted by various editors demonstrate the need for a clarification of the evidence: dice et facie; ¹ dicem et faciem; ² dicae et faciae; ³ . ⁴ The reading dicae et faciae found its way into the most recent edition, the Oxford Classical Text of Michael Winterbottom ([1970] 1989). A close examination of the evidence will confirm that this is not the correct reading. It is time to put to rest a good deal of needless controversy by showing that Cato and his contemporaries used a distinct first-person form in the future indicative of the third conjugation and that they usually or always spelled the active form without the final -m. The reading dice et facie accepted by Radermacher (1959), Fierville (1890), and Halm (1869, 367) should stand.

A century ago (1898) Moore argued, in a brief note influential enough to be cited in Winterbottom's apparatus, that behind Quintilian's citation as preserved lies an innovative orthographical practice not otherwise attested. Moore constructed his hypothesis on the assumption that the variants adduced were simply different ways of spelling the forms dicam and faciam. ⁵ Cato was not satisfied, Moore proposed, simply to follow the usual practice of the time and omit spelling the final -m, which was in many environments only partly articulated. Since the -m did, however, have some impact on the sound, Cato invented a new custom, turning M on its side at the end of the word to indicate the nasal quality of the final vowel. This anomalous symbol was preserved by Quintilian but later misconstrued as E. In this way Moore attempted to render the otherwise nonsensical forms dicae and faciae comprehensible as textual corruptions of Cato's innovation.

Klotz (1931) dealt in greater depth with the problem, and although his solution is almost certainly inadequate, he touched on most of the evidence needed to solve it. Because Quintilian quotes Cato's archaic spelling convention again in book 9 (Inst. 9.4.39) in a discussion of sounds which are not spelled, Klotz conjectured (142) that Verrius Flaccus had quoted Cato as a precedent for his own practice of indicating the attenuated pronunciation of an elided final -m by writing only the first half of the letter. So, Klotz's argument continues, it was from Verrius Flaccus (GL 7.80) that Messala drew the example, and from him Quintilian derived the extant version. The "half-M" was misconstrued by Quintilian as E, and so the textual tradition preserves what he wrote: dicae et faciae.

Both Moore and Klotz have read Quintilian too narrowly. Each of their arguments is predicated on the assumption (in Moore's case) or argument (in Klotz's) that Cato's forms were morphologically identical to Quintilian's dicam and faciam. They dismissed the possibility that Cato was spelling forms distinct from the classical counterparts listed by Quintilian. This is not a surprising dismissal: among the examples cited by Quintilian, these are the only ones which turn out (if the current argument is accepted) to have been morphologically distinct from the classical "equivalents." Nevertheless, a careful examination of all the evidence, and in particular a close reading of the cognate passage...