Ovid in exile shows one of the many problems with any theory that
the only real poetry for the Romans was performed poetry. If so, we
should then expect to find a clear difference between ‘‘normal’’ poetry,
meant for performance by the poet in front of an audience, and ‘‘abnor-
mal’’ poetry, which the author was forced to send to unknown readers.
In short, theAmores (in this theory written for recitation at Rome)
ought to have not just a different subject but ought to be an utterly
different kind of composition from theTristiaor theEx Ponto.^154 We
ought to be able to hear the difference in theFastibetween verses
written at Rome to be performed and those written from Tomis to be
read. We ought to be able to hear at once that Martial Book 12, sent from
the ends of the earth to Rome, is utterly different from his other books
of epigrams, sent from Rome to the ends of the earth (2.1, 11.3, 5.61,
12.2, 12.5).
That poets expected their poems to be read out of books is shown not
only by the descriptions of presentation copies of the verses (Cat. 1; [Tib.]
3.1, Ov.Trist. 1.1, Mart. 4.10), but also by the fact that draft versions
were sent to selected readers, even when some of the poems had been
recited to the very person now receiving the finished volume or prepu-
blication proofs.^155 Catullus’s friend Caecilius sent him a draft of his
Magna Mater(Cat. 35). Vergil sent drafts of portions of theAeneidto
Augustus when he was away on campaign (Vit. Don. 31; cf. Macr. Sat.
1.24.11). These poets could very well have recited these works.^156 They
did. But they also chose to treat their verses as written words intended
to be read by someone at a distance even in the earliest stages of dissem-
ination. Quinn maintains (1982, 156):
Performance is always implied. Even when contact with a writer takes place
through a written text, that text was thought of as recording an actual
performance by the writer. .. it is offered as, so to speak, a transcript of a
performance which the reader recreates for himself.But this is simply not the case. As often as poems offer themselves as
fictive representations of the poet’s speaking voice (e.g., Cat. 4, 5), they
come in the guise of fictive letters, drawing deliberate attention
- Cf. the opening ofPont. 4.1.
- So for the finished volume: Cat. 1 to Nepos, though he has heard or read some of
the poems; Horace’sOdesto Augustus (Ep. 1.13). For a draft to be criticized: Pliny 3.15.
Augustusreadthe first book of Horace’sSatires(Suet.Vit.Hor.). So, too, for prose, for
example, Cic.Att. 13.21a (SB 327), 15.14.4 (SB 402), 15.27.2 (SB 406), 16.11.1 (SB 420).
For Pliny, see above, esp. 1.8, 9.28. See Starr 1987, 213; Valette Cagnac 1997, 145. - Horace pictures the critic listening as the poet reads his verses (Ars438 44),
but also picking up the written book and reading, emending, annotating, and crossing out
(445 50). Even in the case ofrecitatio, as Dupont notes (1997, 45): ‘‘We are dealing with
a real ‘writer,’ that is someone who has entrusted his text to the page.’’ Cf. Valette Cagnac
1997, 116 25.
224 Books and Texts