Richard Tarrant
Editors of classical texts expend great effort in attempting to recover the original wording
of those texts, and are equally diligent in removing transmitted material that is clearly or
probably non-authorial (such as the titles prefixed to poems in manuscripts of Horace,
Propertius and Ovid). When it comes to the grouping and ordering of works, however,
those same editors often perpetuate arrangements that have no claim to authorial
standing, some of which are likely to be the product of late antique editorial decisions.
I first look briefly at three authors whose works as grouped in modern editions do not
correspond to any arrangement chosen by the authors themselves: Catullus, Virgil and
Ovid (the so-called Carmina Amatoria, comprising Amores, Medicamina, Ars Amatoria
and Remedia Amoris). The organization of Catullus' poems in three categories (short
poems in a variety of meters, longer poems, poems in elegiac couplets) may be the work
of an editor close in time to the poet; the gathering together of Eclogues, Georgics and
Aeneid, while not authorial, reflects a widespread ancient consensus about the canonical
works of Virgil; while the Ovidian collection may simply be a grouping of works on
related subjects, created for a late antique codex that is the ancestor of the earliest
Carolingian manuscripts.
The case of Horace then receives a somewhat fuller discussion. The order of works in all
modern editions (Odes I-IV, Carmen Saeculare, Epodes, Satires I-II, Epistles I-II, Ars
Poetica) bears almost no relation to the order in which the several collections were
written and published, but is instead the product of two organizing principles: metrical
(lyric vs. hexameter) and hierarchical (with the Odes placed first as Horace's magnum
opus and the other works following). There is no reason to think that Horace oversaw a
collected edition of his poetry along such lines; in the only place where he refers to more
than one of his own works (Epistles 1.19.23-34), he does so in chronological order,
mentioning Epodes before Odes. The conventional order must be ancient in origin, since
it broadly corresponds to the lists of Horace's works in the Vitae attached to Porphyrio's
commentary and the ps-Acronian scholia, but it does not possess absolute authority, and a
modern editor should be free to adopt a chronological arrangement that more accurately
reflects Horace's evolution as a poet. That arrangement would order the collections as
follows: Satires I and II; Epodes; Odes I-III; Epistles I; Epistula ad Florum (= Epistles
2.2), Carmen Saeculare, Odes IV, Epistula ad Augustum (= Epistles 2.1), Ars Poetica.
(Producing a consistently chronological arrangement entails dismantling the traditional
"Epistles Book 2"—comprising just two works, the Epistula ad Florum and the Epistula
ad Augustum—which I would argue is an editorial construct.) This is a situation in which
an editor could bring about a noticeable improvement in our appreciation of an author's
work without altering a single letter of the text.
If time permits I would like to raise a broader question, namely, to what extent these
issues of grouping and ordering are consequences of the replacement of the roll by the
codex as the preferred medium of transmission for literary texts. It seems likely that
decisions about the order of works only became pressing when it was possible to
encompass some or all the works of an author in a single volume.
It is obvious that the textual choices made by editors can have a significant effect on
readers' understanding of an author. The same is true of the way editors choose to group
and order texts; that aspect of the editorial process should therefore be similarly
transparent and open to scrutiny.