On Varro’s and Cicero’s spelling and pronunciation: A clarification

I am grateful to Dr. Federico Biddau for his summary of the problematic issue I pointed out in my review of his paper without the necessary accuracy. Moreover, I am also glad that I can publish Dr. Biddau’s defense or explanation in the same issue of JoLL, where also the excellent paper by Prof. Marco Mancini was published on a similar subject [some editorial problems forced us to publish Mancini’s paper in the issue immediately preceding this issue].

As for my criticism of Dr. Biddau’s paper, first of all I want to highlight two points of method, (i) reading exactly the ancient texts, (ii) quoting them completely. The first passage taken into account by Dr. Biddau is a well-known passage of Varro’s, Ling. 9.80. Varro attests that some words were pronounced with the diphthong ei, not written (as Dr. Biddau thinks), because Varro used the verb dico three times, a mistake Dr. Biddau reiterated two times in his paper (pp. 51b and 62) and also in his “Clarification”:

Varro, Ling. 9.80 (I checked Gӧtz-Schӧll and Kent but now I give the reading by Biddau, pp. 51 and 62)

Negant esse analogias quod aliei dicunt ‘cupressus’, aliei ‘cupressei, item de ficeis, plataneisque, plerisque arboribus, de quibus aliei extremum ‘us’, aliei ‘ei’ faciunt. Id est falsum: nam deberent dici ‘e’ et ‘i’: ‘ficei’ ut ‘nummei’, quod est ut ‘nummei<s>’ ‘ficei<s)’, ut ‘nummorum’ ‘ficorum’. Si essent plures ‘ficus’, essent ut ‘manus’: diceremus ut ‘manibus’ sic ‘ficibus’, et ut ‘manum’ sic ‘ficum’.

I give also the commentary by Biddau (p. 51): “Qui Varrone dice apertamente che il nominativo plurale della seconda declinazione ‒ quindi un morfema, anche se lui non si esprime così ‒ esce in – ei.” He would have had to say not “esce” but “è pronunciato,” as well as in the following example:

Lucil. 364–65 Marx

Iam ‘pueri uenere’: ‘e’ postremum facito atque ‘i’ ut ‘puerei’ plures fiant.

Here Biddau’s translation is right (albeit he is wrongly referring to both Varro’s and Lucilius’ passages) (p. 51, and see also p. 62): “Cosa ci dice questo primo genere di fonte? Come quella parola, quel morfema o quel fonema veniva scritto all’epoca dell’autore.” Quintilian (inst.1.7.14–15), quoting the passage from Lucilius just quoted, writes scripserunt and informs us that this use (ut e et i iungendis eadem ratione qua Graeci ει uterentur) lasted until Accius and a little further (usque ad Accium [dead ca. 85] et ultra). Biddau’s insistence on writing (instead of pronouncing) of ei is connected with his idea that even Cicero wrote deico, by pronouncing dīco, which is the reason for our discussion. This question is grounded on Cicero’s passage which Biddau quoted only partially, not enough to make the issue clear in every respect. Here I give Cicero’s whole two paragraphs:

Cic. De orat.3. 45–46 (I employed Kumaniecki’s edition)

equidem (L. Crassus is speaking) cum audio socrum meam Laeliam ‒ facilius enim mulieres incorruptam antiquitatem conservant, quod multorum sermones expertes ea tenent semper, quae prima didicerunt ‒ sed eam sic audio, ut Plautum aut Naevium videar audire; sono ipso vocis ita recto et simplici est, ut nihil ostentationis aut imitationis adferre videatur; ex quo sic locutum eius patrem iudico, sic maiores; non aspere, ut ille, quem dixi (sc. Q.Valerium Soranum), non vaste, non rustice, non hiulce, sed presse et aequabiliter et leviter. Quare Cotta noster, cuius tu illa lata, Sulpici, non numquam imitaris, ut Iota litteram tollas et E plenissimum dicas, non mihi oratores antiquos, sed messores videatur imitari.

Literature about this passage may be found in Leeman, Pinkster and Wisse (1996: 192). Of this passage I would like only to highlight the adverbs Cicero used to describe the pronunciation of what derived from -ei, either -ī, or -ē. Particularly instructive is the adverb hiulce and, for the meaning of this adverb, cf. Thes.L.L.VI, 3 2847, 14–2848,33: s.v. hiulcus, -a, -um, B. t.t. Rhet. et gramm. de hiatu […] a. respicitur ordo vocalium, […] Mart. Cap. 5.516 hiulcae sunt [sc. litterae], cum in ea parte, quam diximus, similes vocales ac similiter longae collisam hiantemque structuram faciant. Of the three forms, namely -eī; -ī and -ē, obviously only the first one could be hiulca, and the adverb hiulce can concern only the pronunciation, not the writing. Even more so if the pronunciation of the diphthong ei was as Adams (2007) supposed for eiram of Plaut. Truc. 262 comprime sis eiram: “Since the one certainty about the passage is that iram must have appeared in the text as eiram, the actor may have pronounced the word [I am stressing] with an exaggerated (and perhaps old-fashioned) diphthong such that both elements were clearly heard by the audience [hiulce, I would say] but not by Truculentus, who, accustomed to articulating ei as a lung close e, was made to pick up only the e element” (Adams 2007: 53) ^[1] and understood eram.

For his part, Dr. Biddau, after quoting only partially Cicero’s exposition, ^[2] namely from Cotta noster to videatur imitari (which Biddau presumed to write imitarei), developed an argumentation in order to reach the unacceptable conclusion that Cicero wrote deico instead of dīco, whereas he pronounced dīco and reproaches Sulpicius and in particular Cotta for pronouncing dēco. I give the words Dr. Biddau himself used (pp. 52–53):

Crasso quindi rimprovera a Sulpicio di pronunziare certe i togliendo la iota littera e dicendo un e plenissimum. Si possono fare alcune considerazioni interessanti: la prima è che Cicerone almeno qui usa per le lettere latine nomi greci (ricordo che l’epsilon in antico si chiamava semplicemente e, e quindi iota ed e sono i nomi greci delle lettere latine scritte allo stesso modo: i ed e); un’altra è che Cicerone, parlando di suoni, di pronunzia, allude alla forma scritta delle parole⁷. [⁷La scrittura resta spesso riconoscibile sotto traccia come rappresentazione mentale privilegiata della lingua in contesti eminentemente orali] Ma quello che più ci interessa ora è che Cicerone sta pensando a parole scritte col digramma ei, a cui, togliendo la i, resta la semplice e. Questo vuol dire che a Cicerone era familiare la scrittura ei per parole che nel latino che noi chiamiamo «classico» si scrivono con una semplice i. Purtroppo Cicerone non fa esempii concreti; ma se quelle parole scritte con ei che lui ha in mente potevano essere pronunziate con un e plenissimum, vuol dire che in esse ei era originario: com’è noto [e:] era l’esito, al tempo di Cicerone ormai limitato al latino non romano (ai messores direbbe [dice] lui), del più antico dittongo [ej], che nella Roma del I sec. a.C. era ormai chiuso in [i:]. Quindi abbiamo non solo una prova che Cicerone scriveva ei, ma anche un piccolo indizio per capire dove. (Biddau, pp. 52–53):

To this claim of Dr. Biddau (“Ma quello che più ci interessa […] resta la semplice e”) and that Cicero wrote deico, I reacted, writing (p. 123): “This solution is founded on the idea that Cicero actually [i. e. in the present moment, in point of fact] pronounced ei, instead of i.” Now Biddau prefers to think “in his mind” and this is the core of our discussion. Dr. Biddau writes now: “nor could I ever have imagined being attributed such a notion by a reader of my paper”. Unfortunately, I am not so accustomed to walk inside of Cicero’s mind as Dr. Biddau, and as I read that Marcus Tullius switched from written to pronounced deico, I did not realize that the switch was blocked in the middle, completely developed in this case, but excluded in every other one, I am sorry. Anyhow, outside of every fancy, my mistake was that I accepted Biddau’s data and entered into the tangle of his argumentation instead of taking into account Cicero’s whole passage. For two data show that Biddau, after the right remarks about the Greek name of the letters, fell into error over Cicero’s use of deico: (i) because he missed out Cicero’s description of the pronunciation of deico, in particular that it happened hiulce and (ii) the whole development from deico to deco and dico ^[3] described by Crassus with reference to Laelia, his mother-in-law (who was alive in Naevius’ and Plautus’ time, in a direct contact with the pronunciation of deico). On the other hand, that in Plautus’ time the pronunciation ei was usual is proved by the famous joke in Plaut. Truc. 262–264, throughout explained by Adams (2007: 52–54) (and confirmed by Plautus himself, see note 6), a joke which presupposes a literate, urban audience to be understood. Therefore Biddau, in absence of a pronounced deico, which existed however in Laelia’s time quoted by Cicero, had to invent or better to fancy the transition from written deico to pronounced deico in Cicero’s mind, while Cicero’s description was not concerning such a transition but only the use and judgement of deico as it was pronounced. The history from Laelia onward excluded the necessity of this invention. If Dr. Biddau had paid more attention to Laelia and the whole story, he would not have needed to invent such a transition in Cicero’s mind and discussion. Cicero in his mind did not need to switch from writing to pronunciation. On the other hand, Varro (Lat. 9.80) gave enough of a pronounced ei in this time, namely in nom.pl. of the names like cupressei.

That in epigraphic Latin the diphthong -ei was employed instead of -ī and even largely employed does not need any proof, because it occurs in several documents. However, Adams (2003: 42–49; Adams 2007: 52–64) showed that even in epigraphic use -ei, and Greek ει, was prevailing, but not exclusively employed: and he adds that it is possible that in some areas the e became dialectal (a question treated by my colleague and friend Alfredo Ghiselli 1961 in connection with Livy’s patavinitas). I only contested that such a written -ei did not influence the oral pronunciation, though I agree that in Latin epigraphic script the use of diphthong -ei to represent a long ī was largely prevailing. Dr. Biddau denies now that he had such a foolish idea, namely that written deico was replicated in an oral deico through a kind of influence or connection, whereas he uses even this idea in the only case Cicero treated this question. But is this idea so absurd as Dr. Biddau wrote? The pronunciation of -ei has been attested by Varro, the most influential scholar on these questions in Cicero’s time.

Another point concerns the συνπάχειν theory as this theory has been developed by Ferdinand Sommer (1909) starting from Lucilius’ passage. In Dr. Biddau’s opinion, I gave too much confidence to this theory and he wants to show that such a trust is wrongly placed. While I am waiting for the new discussion by Dr. Biddau about the συνπάχειν theory, I tried to answer the first point taken into account by Dr. Biddau and now I still give credit to Sommer’s συνπάχειν theory. Then this theory is a kind of analogy and actually as a clue of analogy is given. Anyhow I agree with Dr. Biddau that -ei was the prevailing (not exclusively) epigraphic form, while -ī was the prevailing pronunciation of the city Rome, -ē⌦ not excluded (Adams 2007: 58).

Cicero in this passage says that the pronunciation of dico from the previous deico was urbane dīco, rustice deco, and an important point in this question is the reduced number of inscriptions from the city Rome (about 37) in comparison with the inscriptions found outside Rome (more than 200 from Latium and Italy, cf. J. N. Adams 2007: 42 n. 32). This concerns the question of urbanitas ~ rusticitas as it has been pointed out by Marouzeau (1949: 12) and Ramage (1973) and it should certainly be taken into account in order to have a clear picture of Cicero’s situation (and his reaction against ē and possibly ei) but I cannot involve myself in so complicated a question which concerns both grammar and rhetoric (my discussion [Calboli 1986: 1108–1122], seems to me too old now and rather not specific for this question). On the other hand, J. N. Adams (2007: 54–64) gave already a consistent collection of matter to meditate upon for me and also for Dr. Biddau.

Finally, Dr. Biddau challenged me to find some other material to open this question, some written dico in Cicero’s time and outside epigraphic Latin, I mean. By reading Adams’ (2003) book on “Bilingualism,” I realized that also in this case the corresponding Greek use is enlightening: only in epigraphic Greek we can speak of a regular use of -ει to render Latin -ī. More difficult was the challenge to find an example of literary Latin dīco in Cicero’s time, I mean in a work we call of Latin literature. It was not easy to find such a material outside epigraphic documents, better said it would be impossible because of the almost absolute absence of such material in republican time. Nevertheless, I could find a couple of examples of dico in a Latin papyrus of the Pisones’ Villa, where the Obolostates siue Foenerator by Caecilius Statius has been transmitted, ^[4] namely in PHerc.78 cr.04 lin. 5 and 6, in the reading by Knut Kleve and Giulia Carosi (a reading I myself have checked), while the writing deico was totally absent (which suggests the hypothesis that deico should have been rather an epigraphic use): d⌦i⌦ces⌦ e⌦s⌦[se do]cile⌦m ‒ dice⌦ mi⌦ Kleve (cives dica⌦m Car. and myself). ^[5] Therefore, Cicero too was free to write dico.

In conclusion, I confirm my very positive judgement of Federico Biddau’s new studies on Latin orthography but nevertheless I would like to conclude my answer and reply with the same words used by Dr. Biddau, with only an addition: “the research field is open, but to people free from too strong persuasions.”