Surely there is something puzzling about 13–14? What Horace was the first to do was to write Latin poetry using the metrical schemes of the Greek lyricists, principally Alcaeus and Sappho, who wrote in the Aeolic dialect of Lesbos. There can be no reasonable doubt that Aeolium carmen refers in the first instance to Horace's adoption of Aeolic metre.Footnote 1 For deduxisse there are two possibilities, and conceivably Horace meant both to operate. The first involves motion, bringing something or someone home or into port (for example, captives in triumph to Rome or a colonizing expedition to its destination in a foreign land). So Horace took Greek metres and moved them in one way or another to Italy. Secondarily deducere might mean ‘to produce (by spinning)’, a usage found at Sat. 2.1.2-4 similisque meorum mille die uersus deduci posse.Footnote 2
It is the phrase ad Italos … modos that causes difficulty. Modi are ‘measures’, and in poetry the principal thing measured is the succession of long and short syllables, so that when the subject is verse, as here, ‘metres’ is the likeliest meaning, as at Am. 1.1.1–2 arma graui numero uiolentaque bella parabam | edere, materia conueniente modis, and Hor. Sat. 1.4.58 tempora certa modosque. Yet if Aeolium carmen means, in effect, Aeolic metre, it makes no sense for Horace to be bringing it home to Italic metre: if you join Aeolic metres with Italian X, that X should logically be something different from metres: themes, perhaps, or settings, or subject matter. West, who translates ‘the first to bring Aeolian song to Italian measures’, must have seen that ‘Italian measures’ as the destination for Aeolian song is difficult since he offers a long explanation: after noting a number of passages where Horace combines, say, the Latin Muse Camena with the adjective Graia or speaks in the same breath of the Greek barbitos and the Latin fides, he goes on to explicate 3.30.13–14 by saying: ‘He has domesticated Greek song so thoroughly to Latin measures that Greek has become as Latin and Latin as Greek.’ But this paraphrase, which West himself admits may be over-subtle, has Horace domesticating Greek song ‘to Latin measures’, the very difficulty with which we began.Footnote 3 Since it fails—after a long divagation unconnected with anything in the text—to address this point, it must be judged unsuccessful.Footnote 4
An alternative is to take modos as ‘melodies’, as do Nisbet and Rudd: ‘Italos modos means, not “Italian metres” (for the metres were taken from Greece), but “Italian melodies”, so described because they are played on the Roman lyre, i.e. they are in Latin’.Footnote 5 But a melody is a sequence of notes in a certain rhythm, and any melody that fits a poem written in, say, Alcaics must accommodate a stanza of exactly 41 syllables with characteristic alternation of single and double short: in other words, it would be an Aeolic melody. No pre-existing Italian melody would suit this syllable count or this metrical shape, and if Horace wrote a new one that did, it does not follow that it is an Italian melody merely because an Italian wrote it or because Latin words are sung to it. So this explanation too is unsatisfactory.Footnote 6
What about the possibility that modi are musical ‘modes’, i.e. scales? Might there have been scales, sequences of intervals, that were specifically Italian? Might it have been to these that Horace set his imported metrical patterns? Yet neither OLD nor OLD 2 cites any instance of modus in that meaning.Footnote 7 Furthermore, Epod. 9.5–6 sonante mixtum tibiis carmen lyra, | hac Dorium, illis barbarum seems to show that to express ‘Dorian mode’ and ‘Phrygian (or Lydian) mode’ the word we need is not modus but carmen.
But could modi not mean ‘poetry’, as attested at OLD or OLD 2 s.v., 8b? It is true that, since verse is composed in modi, ‘metres’, this word is sometimes used by metonymy for ‘poetry’. And it is also true that in some ten places in Horace the word could be plausibly translated as ‘poetry’.Footnote 8 But in none of these cases is it important that the sense ‘metre’ be excluded on the grounds that it would create confusion. By contrast, here the fact that Aeolic metre figures so prominently in what Horace deduxit means that ad Italos modos would be a maladroit expression for the destination to which he brought it. (There likewise seems to be no way on this reading to accommodate the possibility of ‘produce by spinning’ as the meaning of deduxisse.) This defence is therefore unsuccessful. Can we do better? Can we rescue the more intuitively appealing sense ‘bring Aeolic metre to Italian X’ or (secondarily) ‘spin out Aeolic verses to the accompaniment of Italian X’?
Surely the point Horace is making is that he has enriched Italian song by the importation of the Aeolic metrical schemes. How might he have made this point in a way that does not cause learned and otherwise careful commentators to become vague, implausible or apologetic?Footnote 9 Others before me, without venturing to alter the text, have indicated the point Horace should be making. Garrison writes, ‘[M]odos are meters. Horace might more accurately have said that he adapted Aeolic modos to Italian carmina, i.e. poems set in Italy and addressed to a Roman audience.’Footnote 10 I suggest that this is exactly what Horace did say and that he wrote
Horace then would be claiming that he brought Aeolic metres to Italian song, an unexceptionable statement on which not a single vague or apologetic note would ever have been written. A close parallel to the sense, where foreign modi are added to Italian song, is Epist. 1.3.12–13 fidibusne Latinis | Thebanos aptare modos, where the poet Titius is doing for Pindar what Horace did for Alcaeus. The following illustrate its conformity with Horatian practice.
(1) The order noun–preposition–adjective exemplified in carmen ad Italum, though not as common as adjective–preposition–noun, is amply attested in the Odes: see 1.22.9 silua … in Sabina, 1.36.4 Hesperia … ab ultima, 1.36.12 morem in Salium, 2.3.1 rebus in arduis, 2.7.1 tempus in ultimum, 2.9.6 mensis per omnis, 2.13.25 puellis de popularibus, 3.1.6 reges in ipsos, 3.4.9 Vulture in Apulo, 3.27.43 fluctus … per longos and 4.5.29 collibus in suis.
(2) Nisbet points out that adjectives derived from proper nouns, which in Latin generally follow the nouns they modify, usually precede them in the Odes because they are felt to be ‘ornamental and emotive rather than factually descriptive’.Footnote 11 They follow, he says, when they refer to a particular place. To this general practice he notes certain exceptions, for example the postposition of Cypria in 1.1.13 (‘another chiasm’, says Nisbet [p. 142], which applies also to our passage). To these add the collocations in 1.20.9 (prelo Caleno), 1.29.9 (sagittas Sericas), 1.36.12 (morem Salium), 2.4.2 (Xanthia Phoceu), 2.7.8 (malobathro Syrio), 2.13.8 (uenena Colcha), 2.13.15 (nauita Poenus) and 2.18.3 (trabes Hymettiae), to go no further than the first two books of Odes. So carmen … Italum has plenty of company.
(3) The use of an adjective–noun pair in one grammatical case to bookend another pair in a different case (pattern aBbA or abBA, where upper-case letters represent nouns and lower-case their adjectives) is common in Horace: see Carm. 1.1.14 Myrtoum pauidus nauta secet mare, 1.1.28 teretes Marsus aper plagas, 1.2.13–14 retortis | litore Etrusco uiolenter undis, 1.3.32–3 semotique prius tarda necessitas | leti, 1.4.9 uiridi nitidum caput impedire myrto, 1.12.11–12 auritas fidibus canoris | ducere quercus, 1.12.22–3 saeuis inimica uirgo | beluis, 1.35.7–8 Bithyna … | Carpathium pelagus carina, 2.17.3–4 mearum | grande decus columenque rerum, etc. A particularly close parallel, where the outer pair encloses an inner pair belonging to a prepositional phrase, is Carm. 1.26.2–3 proteruis in mare Creticum | portare uentis.Footnote 12
(4) For carmen as a collective singular, the whole body of a certain kind of verse, see Carm. 1.6.2 Maeonii carminis alite. Horace, therefore, brought Aeolic measures ‘to Italian song’.
(5) Furthermore, since deduco can mean not only ‘lead home in triumph’ or the like but also ‘produce (by spinning)’, it is pleasing to see that the lines make sense with this second meaning: Horace boasts that he was the first to produce or spin out Aeolic verses to be sung to Latin scales. For carmen as ‘scale’ see Epod. 9.5–6 cited above, where the lyre produces a carmen Doricum and the tibiae a carmen barbarum, for example Phrygian or Lydian.Footnote 13 For ad, ‘to the sound of’, see OLD or OLD 2 s.v. 39.
A possible objection is Aeolio carmine nobilem at 4.3.12, which looks like a reference to our passage and seems to support Aeolium carmen. Yet Horace, who often reverts to earlier themes in later poems, has no objection to joining adjectives to different nouns the second time round. Thus at Carm. 2.13.24–5 he writes Aeoliis fidibus querentem | Sappho puellis de popularibus, with ‘Aeolian’ modifying the lyre, whereas in the similar phrase at 4.9.11–12, commissi calores | Aeoliae fidibus puellae, it modifies Sappho herself. Horace also varies phrases by using synonyms, for example 2.17.20 Hesperiae … undae compared with 1.28.26 fluctibus Hesperiis. The objection must be regarded as without force.
At this point, readers may be willing to grant that the alteration I propose produces clarity where it was lacking but might still wonder (a) whether conjectural intervention is a prudent course of action in an author whose text is as generally sound as Horace's; and (b) how this particular corruption is to be explained.
As to (a), Horace's text in general is good, but the list of places where a significant number of editors agree that it is faulty, for example limina Pulliae at Carm. 3.4.10, is not small.Footnote 14 The united witness of the manuscripts and ancient quoters sometimes has to be set aside.
(b) It is well known that scribes often unconsciously adjust the ending of a word to make it agree with a neighbouring word. Here it has happened twice in one line: Aeolios was adjusted to agree with carmen and Italum to agree with modos. Analogous double corruption—where two adjectives in the same line have had their endings corrupted each into the gender, number and case of the other—is attested in actually occurring variants in the manuscripts of Horace: see Carm. 3.8.19–20, where the chief witnesses read Medus infestus sibi luctuosis dissidet armis while the deteriores read Medus infestis sibi luctuosus dissidet armis.Footnote 15 At Sat. 1.6.75 octonos referentes Idibus aeris is similarly corrupted in some manuscripts to octonis referentes Idibus aera.
After making the above case I discovered that my conjecture was anticipated in 1840 by one Johann Daniel Fuss (1800–60), Professor of Latin in Lüttich.Footnote 16 Fuss made his case poorly and obliquely.Footnote 17 The 1850 third edition of Orelli and Baiter contained a one-sentence rejection of it,Footnote 18 and in subsequent editions even that was omitted, with the result that, as far as I can determine, no one after 1850 took any trouble with Fuss. Yet though he did not do his conjecture justice, the arguments I have put forward here will, I hope, restore his name to the foot of the Horatian page.
ADDENDUM: WHAT ABOUT ‘IT SHALL BE SAID OF ME IN APULIA’ (10–12)?
I take this opportunity to revisit another problem in this passage to which I once drew attention and to propose a more satisfactory solution. In 2011Footnote 19 I challenged Fraenkel's viewFootnote 20 —which had become, and probably still is, the communis opinio—that in Carm. 3.30.10–12 Horace boasts that he will be spoken of in Apulia as the great innovator in Latin verse. Fraenkel tries to argue that Italians, then as now, felt pride in what was said of them in the place of their birth, but I demonstrated that the passages he cites do not show this. I argued, as others had before me, that after Horace's temporal claim that his poetry will endure as long as Rome itself, the spatial claim that he will be renowned in Apulia cannot be what Horace meant, and that ‘where Aufidus roars and Daunus once ruled’ must refer to the place of his birth, not of his fame. I also showed that, pace Fraenkel, the word order this reading requires—a clause of relative import preceding its antecedent or equivalent—is Horatian (cf. the ex quo clause at 3.3.21–2, which is anchored in damnatum Mineruae), and that all that is needed in our passage is a word meaning oriundus to anchor qua … populorum. I showed that such a word once stood where potens now stands, since potens gives impossible sense. The phrase ex humili potens could mean only ‘once humble, now powerful’, since in a phrase of the form ex albo ater, where the contrasted terms are adjectives, they must be direct antonyms. Yet Horace, though clearly describable as humilis, never became powerful: see, for example, Sat. 2.6.40–6, where he has no influence even on his friend Maecenas.Footnote 21 A second point is that in every other occurrence of the phrase ex humili gleaned from the Brepols collection, the adjective is neuter or modifies an unexpressed loco. Ex humili (with ‘person’ understood) is unique. The upshot is that in place of potens we need a word showing that ‘where Aufidus roars and Daunus once ruled’ is where Horace was born or started out. I still regard this reasoning as sound.
Hereafter I became less plausible. I rejected all present participles, on the grounds that to express the antecedent action ‘having been born’ or ‘having arisen’ a perfect participle is needed, and settled on editus, whose corruption to potens I explained by the supposition that the eye of a scribe early in the tradition wandered up to impotens in 3. But although the echo of 1.1.1 atauis edite regibus would be pretty, the explanation of the corruption is unconvincing. More decisive still, Horace never uses the prefix e or ex to reinforce that preposition in the same clause.Footnote 22 Another approach is called for.
Had I found my way to Kühner–Stegmann 1.757, I would have learned that Latin sometimes does use the present participle for an antecedent action, so that a word in -ens or -ans cannot be ruled out a priori. (Horace has such participles at Epod. 1.19, 2.45 and 2.47; Sat. 2.2.91; and Epist. 1.17.10.) It would be lovely to have him ‘appear’ in Apulia from humble station (ex humili parens). Disobligingly, however, the first vowel of pareo is long. But there is more than one present participle that could mean ‘appearing’ in the sense of ‘having appeared’: we should consider the claims of patens, interpreted as ‘having become visible’: for pateo as ‘be visible’ see OLD or OLD 2 s.v., 6 (Lewis and Short s.v., B.4 say ‘cf. appareo, detegor’). It was in Apulia that Horace first came to light, appeared, out of humble origins. The corruption of ex humili patens to ex humili potens is not difficult to imagine and indeed might be thought virtually inevitable. This emendation thus gives the sense we expect from the context (no merely local fame) and rids us of a Horace self-described as potens: ‘It will be said of me that where the Aufidus roars and where Daunus, poor in water, ruled over farmer folk, appearing out of humble circumstances I was the first’, etc. The improvements are cheaply bought by the alteration of a single letter. Taken together, these points, though falling short of a Euclidean ὅπερ ἔδει δεῖξαι, constitute a strong case for according the conjecture a place in the apparatus criticus.
