Classical World

The extent of silent reading in antiquity has been debated for a century, as scholars have continued to look for evidence and examples of the practice. William Bedell Stanford laid out most of the relevant passages in his classic discussion of the topic. Plutarch would suggest that Julius Caesar could read a document silently (άναγινώσκειν σιωπή, Brut. 5), Suetonius that a group of unfortunate Roman equites could do so as well (pugillarium, quostaciti . . . legerent, Aug. 39).¹ A few other passages Stanford sees as inconclusive (Hor. Sat. 1.3.64–65, 1.6.122–123, 2.5.68; Ov. Her. 21.1–4). The first attested silent reader of literature in Latin according to Stanford is St. Ambrose in the fourth century, as described by St. Augustine, himself subsequently becoming the second attested reader (Conf. 6.3 and 8.12). Discussions subsequent to Stanford have acknowledged the persistence of reading aloud throughout antiquity, if not adopting Stanford's assertion that silent reading essentially did not exist. But Stanford, along with his predecessors and subsequent commentators, neglected a series of passages in Roman elegy, written a century earlier than those in Plutarch and Suetonius, where silent reading is clearly implied.

The least ambiguous of these passages is in Amores 1.4, where Ovid is visualizing and pre-enacting a cena to be attended by Ovid, Corinna (apparently), and Corinna's uir. Since communication will be difficult, Ovid gives Corinna a list of prearranged signs and aids to communication, among them the following (19–20): uerba superciliis sine uoce loquentia dicam; / uerba leges digitis, uerba notata mero. The words Ovid will "speak" with his eyebrows are specifically said to be silent (sineuoce), but those Corinna will read on the table must be so as well. For this scenario to make minimal [End Page 43] sense (and minimal sense is probably the most we can expect of Ovid here), Corinna must be reading the words silently, or otherwise her uir would hear her saying the words aloud.

Many scholars besides Stanford have not deduced silent reading in Amores 1.4 and in some similar passages in the elegists.² One very similar and probably derivative passage appears in ArsAmatoria 1.571–572, compared to the above-quoted Amores passage by Smith, Barsby, Ford, and McKeown.³ Here Ovid offers the prospective lover advice to use the same tactics employed in Amores 1.4, including: blanditiasque leues tenui perscribere uino, / ut dominam in mensa se legat illa tuam. Again the words in wine must be read silently (sermonetecto, 569). Two earlier passages, Tibullus 1.2.21–22 and 1.6.19–20, also compared by Smith, Barsby, Ford, and McKeown, as well as Munari,⁴ when read in this context must also refer to the silent reading of a lover's message: illa uiro coram nutus conferre loquaces / blandaque compositis abdere uerba notis; and neu te decipiat nutu digitoque liquorem / ne trahat et mensae ducat in orbe notas. It is possible that the notae in Tibullus are communicated through some sort of sign language. But in the first Tibullan passage the notae represent uerba, and in the second they are written in wine on a table with a finger, so when read with the Ovidian passages quoted here, the most likely conclusion is that all of these texts indicate that in the first century B.C. it was possible for the female lover to read, and to read silently. It seems simply not to have occurred to modern scholars that the dramatic situations call for the reading to be silent. Dating of the elegists' poems is difficult, but the passages in Tibullus would seem to be the earliest clear references (other than Julius Caesar in Plutarch) to silent reading in Latin, with Ovid's next.

What sort of message would Ovid send to Corinna in this way? In his literary biography Ovid tells us that he was unable to follow the wishes of his father that he become a lawyer like his brother. He tried to write prose (uerbasolutamodis), but the words came out in meter (carmennumerosueniebatadaptos, / et . . . uersuserat, Tr. 4.10.24–26). Consequently, Ovid...