Short URL for this page:

[image ALT: Much of my site will be useless to you if you've got the images turned off!]
Bill Thayer

[image ALT: Cliccare qui per una pagina di aiuto in Italiano.]

[Link to a series of help pages]
[Link to the next level up]
[Link to my homepage]

This webpage reproduces an article in
Classical Philology
Vol. 15, No. 2 (Apr. 1920), pp176‑183.

The text is in the public domain.

This page has been carefully proofread
and I believe it to be free of errors.
If you find a mistake though,
please let me know!

Thayer's Note: For the full text of the inscription, taken for granted in this article, see my note to Platner & Ashby's article Columna Rostrata C. Duilii.

 p176  The Elogium Duilianum
By Edwin W. Fay

1. In a recent paper in Classical Philology (XIV, 74 ff.) Professor Tenney Frank has discussed the authenticity of the inscription on the Columna Rostrata Duili. He dissents from the conclusions of Mommsen and Ritschl — and Lommatzsch — that the inscription is a forgery, contending with Woelfflin that it is a restoration, indeed, a second restoration. Woelfflin set the date of the restoration, the time of its actual chiseling, in the reign of Augustus; Lommatzsch has left it in the time of Claudius. Professor Frank suggests the time of Tiberius and cites Dio Cassius to show that a Commission to Restore Public Monuments (δεμόσια γράμματα) was in fact appointed by Tiberius. I cast my vote for the Commission of Tiberius, but in favor of a forgery. The reader must not, however, feel all the moral indignation that the term forgery is likely now to provoke. Morally, the authors of the Elogium Duilianum no more committed a forgery than Cicero when he couched his Leges in archaic terms.

2. What impulse moved the Imperial Commission, restoring or erecting anew the Columna Rostrata Duili — Servius ad Georg. 3.29 mentions two such columns — to provide it with an appropriate inscription? The Commission would have known of the (two or more) columns erected by Augustus to commemorate the battle of Actium (cf. Vergil G. 3.29). They doubtless knew, for Livy had reported it (40.52), the triumphal inscription on the temple of the Lares Permarini, commemorating the naval victory of L. Aemilius Regillus over Antiochus (B.C. 179). Of this inscription the pontifex, M. Aemilius Lepidus, caused a copy to be exhibited in the temple of Jupiter on the Capitol (Livy 40.51.3). The Commission must have heard of the destruction by lightning (B.C. 172) of the Columna Rostrata Aemili, erected to commemorate a naval victory five years after the victory of Duilius. Knowledge of these things was to be expected of the Commission of Tiberius. If we allow them such knowledge, we provide the Commission with a motive and with  p177 models for restoring, or supplying de novo,​1 a proper elogium for the restored Col. Rostr. Duili. In supplying the Elogium, the Committee on Restoration, particularly if it had an admiral in its member­ship, would have felt it a praiseworthy act to publish and preserve thereon the triumphant entrance of the Roman navy into history. The motive was not to forge or falsify evidence, but to display before the eyes of the people the glories of their navy. This was their motive for erecting anew near the Rostra the Col. Rostr. Duili. By supplying it with an inscription they were but documenting the Rostra, so to speak.

3. The tendency to archaize ran strong in Roman public inscriptions — witness the continued use of moirus and coirare. Only greater authenticity, a greater solemnity, was sought for the Elogium Duilianum by the use of archaic language, and I think it can be shown that the language of our inscription is compatible with the fact of its composition in early imperial times. Nor do I mean to beg the question by saying that, in its survey of public monuments, the Commission of Tiberius must have gathered all the orthographic data necessary for their essay in archaism. I shall rather try to show that any diligent grammaticus acquainted with Varro and Verrius Flaccus was competent to have composed our inscription.

4. Archaistic aspect of Roman education. — It must not be forgotten that Roman education turned predominantly on language, and especially cultivated the ancient forms of Latin. As late as Quintilian (1.1.35; 1.8.15) it was recommended that the difficulties of the ancient tongue, glossae and glossemata, be taught to children as a valuable side issue. Cicero could write his Leges in archaic words because he had been made to commit to memory the XII Tables. Orbilius taught Horace the text of Livius Andronicus by dictation. Who can doubt that Horace might have written the epitaph or Elogium in the language of 250 B.C.? It is enough merely to glance at the titles of the lost works of Verrius Flaccus to realize that, under the employment of Augustus, he can scarcely have refrained from giving courses in Old Latin to the imperial heirs. A schoolmaster will teach what he knows. It was by virtue of their  p178 linguistic education that Romans in the great fields of action wrote of grammar. In his rhetorical works Cicero discusses prosody and forms; and in his correspondence touches on syntax in passing. Caesar expressly wrote on grammatical questions and interested himself, as Augustus did after him, in matters of orthography. Varro was a wealthy man of the world and man of letters to whom grammar and linguistics constituted an absorbing avocation. Much earlier the men of letters had written of things grammatical, as one might expect, seeing that the Odyssey of Livius was done into Latin to serve as a schoolbook. Ennius may still be cited for matters of etymology and for etymological constitutions of orthography. Lucilius, the satirist, hitched the orthography into verse, by way of protest against the spelling rules of the tragedian, Accius. To be sure, such questions had been introduced into Greek literature by the sophists and treated by Plato, the antisophist, but the sophists were professors, forsooth, and grammar was shop. But in Rome grammar awakened the keen interest of men of letters and of the great men of affairs. Does any other literature record so keen an interest and great activity in grammatical matters on the part of statesmen and littérateurs, not schoolmasters?

5. To come now more definitely to grips with our elogium, Woelfflin's defense of its substantial integrity is misleading. Let us examine — often in tacit dependence on Mommsen in CIL 1.40 — his arguments as synopsized by Professor Frank.

6. The use of -que for et. An archaizer would have found a model in the Elogium Aemilianum recorded by Livy in XL.52.

7. For the bombastic tone of the Elogium, so far as this was not incidental to a desire to glorify the Roman navy, the archaizer had a model to hand in the above-mentioned Elogium Aemilianum. The bombastic note in a new-made inscription to Duilius would have been suggested by Cicero's brief but illuminating characterization of that vainglorious worthy in the de Senectute (§ 44). For a bombastic model we need not go with Professor Frank to Sicily, via Athens; nor need we appeal to the innate boastfulness of human beings. The Elogium Aemilianum had a literary model to hand in the first act of Plautus' Amphitruo, where Sosia makes heraldic proclamation of his master's victory.

 p179  8. In the phrase copiasque clasesque nauales the adjective may be justified, without recourse to the primary sense of classis, by a reference to copiae navales in Livy.

9. As regards the words praesented dictatored Mommsen correctly compared inspectante eopse Antiocho in the Elogium Aemilianum (Livy XL.52.6).

10. The turn in altod marId need not be considered antique; see the fairly copious usage in which altum is combined with mare from the time of Laevius and Catullus on (Thes. LL. I.1781.66).

11. The "Method" of Varro and Verrius Flaccus. — Broadly speaking there are but two ways of expressing the relation­ship of fact between variant forms of words: a chronological and an inverted way, as Catullus (C 4) took his yacht backward from her last berth in Lake Benacus to her launching dock at Amastris. Cicero (Orator 153) described the change of duellum to bellum as a contraction. Though his metaphors are not acoustic, Varro habitually derives the later form of a word from the earlier, saying of Casmena/Camena (LL. VII.27) or of *osmen/omen (ibid. 97), that s has been rubbed away (extritum detritum); cf. Quintilian I.7.29: apud antiquos consules n exempta littera legimus. Conversely, abridging Festus, who had abridged Verrius Flaccus, Paulus (59.4) expresses the same fact of language in the words: antiquissimi interserebant s litteram; cf. Festus (222.25), antiqui Casmenas dicebant pro Camenis. The phraseology, the manner of speaking, of Verrius himself seems to be revealed in Festus 222.6: Orcum quem dicimus ait Verrius ab antiquis dictum Uragum, qui et u litterae sonum per o efferebant et per c litterae formam nihilominus g usurpabant; [and Festus goes on, in sound criticism to say] sed nihil affert exemplorum, ut ita esse credamus: nisi quod is deus nos maxime urgeat. The same mode of statement on the part of Verrius Flaccus is attested by Servius ad Aen. VIII.423: hoc pro huc . . . . sicut in epistulis probat Verrius Flaccus exemplis auctoritate ratione, dicens in adverbiis pro u o plerumque maiores ponere consuetos. To Verrius himself, accordingly, not to Festus, the phraseology of 410.2 is to be ascribed: strenam . . . . a numero, quo significatur alterum tertiumque venturum . . . . veluti trenam, praeposita s littera ut in loco et lite solebant antiqui. Here we may further note in Cicero  p180 (Orator 157), isdem/idem. In the light of these examples we may fairly conclude that Verrius often made his orthographic statements inversely, describing the ancient as expansions of the modern forms; if not Verrius, some other "practical" teacher.

12. It is to the practical teacher that we must charge such misleading forms of statement as the two following, touching "paragogic" d: (1) Quintilian I.7.12: ut a Latinis veteribus d plurimis in verbis ultimam, quod manifestum est etiam ex columna rostrata quae est Duilio in foro posita; (2) Charisius (Keil I.112.9) cites ted from Plautus (Cu. 1), remarking: mos erat d litteram omnibus paene vocibus vocali littera finitis adiungere. Only think that the great teacher Quintilian mentions ancient final d without betraying the least knowledge that it was an ablative ending. Yet the same Quintilian, had he been set to supervise the production of an "archaic" inscription, might so carefully have used extant collections of glossemata as to have employed this -d with none but ablative forms.

13. The "method" of the archaizer would have been, not the method of Varro, historically correct, but the inverted method of the "practical" teacher as exhibited in Paulus, after Festus, after Verrius Flaccus, as e.g., in antiqui u litterae sonum per o efferebant . . . . per c litterae formam nihilominus g usurpabant (§ 11). In fact, but not in form, Verrius here accounts, not for c as an orthographic variant of g, but for g as the phonetic source of c. As we proceed down the history of Latin words we usually find that phonetic change is expressed in reductions, that ei yields ī and sm m, whereas the practical grammarian stated the rules for (ancient) spelling in terms of addition (addes e as Lucilius puts it in 367, 370) and insertion (interserebant s litteram, see § 11). It is by such rules of thumb, such rules as we may legitimately ascribe to Verrius Flaccus and his followers, viz., Quintilian, Velius Longus, Terentius Scaurus, and Festus, that we must test and verify the archaistic orthography of the Elogium Duilianum, as set forth in the following rubric.

a) Examples of e for i: (1) 3d. sg. pf. exemet cepet ornavet; (2) enque; (3) nauebos; (4) naualed. Note Quintilian's examples (I.4.17): quid? non e quoque i loco fuit? ut Menerva et Leber [not leber, as Meister reads] et magester et Diove Victore, non Diovi  p181 Victori. In these examples Quintilian fails to distinguish ĕ and ĭ from ē and ī, and certainly never had it in the back of his head that , and then , was the product of still earlier -ei (-ai); cf. also I.4.12 (pinna: bi-pennis) and I.7.24 (sibe quase, a spelling ascribed to Livy and doubtless representing by -e final , not the original diphthong that preceded it). In our inscription, if marId is actually a correct form, certainly navaled is incorrect. Mommsen was doubtless right in ascribing the e- of enque as archaistic after endo (v. in Neue, Formenl.3 2.907); cf. in, later on in the inscription. As regards nauebos, one may well believe with Lindsay (Lat. Inscript., p39) that in the transition of unaccented e to i genuine i was sometimes written or sounded as e, but that is not to say that the e of such forms, hyperarchaism as it was, was not regarded as genuinely archaic.

b) Examples of ei for i: (1) Ablv. castreis socieis; (2) acc. claseis naueis; (3) nom. numei. This orthographic style was known to any grammaticus after Lucilius. The variation ei/i appears in full vigor as late as 49 B.C. (Lex Iul. Munic.). On our inscription the actual variation of accusatives in -eis -es (clases) and -Is (CartaciniensIs) obtains equally on inscriptions of the time of Cicero; cf. (a) turreis (moiros, cited for archaic oi) in Diehl, No. 292 (turreis murum in 297); but (b) murum turres (No. 293; cf. No. 299, 33 B.C.); (c) turrIs (No. 296). In the sacral formulae of an Augustan inscription (Lindsay, Lat. Inscrip., p103) we have both ei and i in the formula ut(e)i tib(e)i in ill(e)is lib(e)is.

c) Examples of o for u: (1) maximos (cf. maxumas later on with u) primos; (2) poplom captom; (3) olorom; (4) nauebos (o chiseled over original u). Quintilian's examples (I.4.16) are: Hecŏba (ŭ) and nōtrix (ū), but inverted Cŭlcides (ŏ) and Pŭlixena (ŏ). His further instances and, as revealed by the choice of examples, these are certainly taken over from inscriptions, are the verb forms dederont probaveront. [Festus also cites the verb forms nequinont and praedotiont; cf. sont in Diehl, No. 243 (130/90 B.C.); and avonculus as late as Vergil (see in Thes. LL. II.1607.60)]. In I.4.11 Quintilian also furnishes attest for volgus and servos and in I.7.26 cites servom and cervom. Let us beware of attributing to any archaizer so much precision as to cause him to avoid the application of the "rule" for  p182 servos and its like to words like primos (o not preceded by v). In fact, our archaizers overplayed their hand and, as Mommsen correctly observed, substituted (5) o for original (IE) u, both in macistratos and in exfociont; note a like error in erodita (see Lindsay, Lat. Inscrip., p96), in an epitaph of 50 B.C. Further note that Cicero in the Leges (III.7) has soboles for suboles.º (6) The form consol is not to be taken with Professor Frank as proof of a transcript of 150 B.C., intermediate between an original of 260 B.C. and the extant text, but rather to be explained as a grammatical feat; cf. Velius Longus (Keil VII.49.14): . . . . id quod apud nos quoque antiqui ostendunt, qui aeque confusas o et u litteras habuere. nam consol scribebatur per o, cum legeretur per u, consul (cf. data for simol ap. Neue). The restoration of n in consul (cf. § 11), whenever it took place, was a fact of recomposition, and in essence and origin recomposition is a psychological act and beyond all dating. Note that in the Sen. Cons. de Bacch. the superscription has consoluerunt, the decree proper cosoleretur; but in the superscription cos. for consules.

d) The diphthong oe (Poenicas). Seeing that Poenus lived on in Latin, one is not warranted in charging up the diphthong of Poenicas to the genuine orthography of 150 B.C., transcribing correct oi of 260 B.C. With archaizing moiros/moerus for murus on every hand (see § 3), the prompting to Poenicas does not even need to be charged to Poenus.

e) Gemination neglected (clases numeri). For the fact cf. Festus 222.27: quia nondum geminabant antiqui consonantes; Quintilian (I.7.23) notes that, a little before Cicero and Vergil, illud quod nos gemina dicimus "iussi," una dixerunt.

f) Examples of c for g: leciones macistratos exfociont pucnandod cartaciniensis. See above, § 13 and cf. Varro LL V.64: antiqui quod nunc G C — citing Ennius for quod gerit fruges Ceres (!). Mommsen thought that c for g in our Elogium was too uniform for a genuine inscription contemporary with Duilius.

g) Examples of ablatives in -d: praedad; pucnandod altod; marId praesented dictatored [navaled, see § 13a]. This -d would have been known to any reader of early public monuments. Quintilian actually cites for it our Elogium and fails to limit his paragogic -d to the ablative (§ 12). Inscriptions apart, this -d would have  p183 been known to all who knew Livius Andronicus and Naevius. Charisius (cited in § 12) either used his source differently or had a different source from Quintilian, for he took his example from Plautus. One cannot suppose that grammatical comment on -d had failed to emerge in Latin scholar­ship before Quintilian.

h) Lack of aspiration in CartaciniensIs: Cf. Quintilian on triumpis non triumphis (I.5.20).

i) On triresmos: Prima facie, triresmos seems a genuinely archaic form and it has been eagerly exploited by morphologists. For all that, the form is probably a veritable howler. So the inscription of the Faliscan Cooks (see Lindsay, Lat. Inscrip., p67) is full of howlers. One can imagine the solemnly pretentious archaisms and hyperarchaisms — Dogberry elegancies — in the liturgies of such a guild. We saw above (§ 11) that Paulus described the s of dusmo, etc., as an insertion. Besides the examples of Paulus (dusmo cosmittere <cf. dismata in Sen. Cons. de Bacch.> Casmenae), Festus (222.25) cites pesnis for pennis and cesnas for caenas; also, as we saw, strenam (for *trenam) slis and slocus; cf. with a floating s nominatives such as magistreis and hisce; also isdem/îdem. The prosodical neglect of final -s had also been commented on by Cicero — as afterward by Quintilian (see the passages in Lindsay, Lat. Lang., p108, § 126). From such a wealth of instances an earlier and better scholar than Dea. Paulus might easily have drawn the inference of a floating s of ad libitum employment for Old Latin. Accordingly, it is open to belief that, as our archaizers mistakenly employed o for u in macistratos and exfociont, they similarly overplayed their hand by inserting s in triresmos. When we combine such exhibitions of hyperarchaism with the demonstration that the other archaisms in our Elogium all conform to grammatical lore extant in the time of Tiberius, we cannot feel confident that our inscription contains a single word taken from an earlier text that it had replaced.

14. Be it permitted here to note, prior to a subsequent fuller discussion, that I also believe that the epitaphs (but not the superscriptions) on the two oldest Scipio tombs were first chiseled thereon in the age of Cicero. Their promoter was Scipio Metellus, known to have been active in restoring, or in newly creating, memorials of his ancestors (see Cicero ad Att. VI.1.17).

University of Texas

The Author's Note:

1 Involved in the destruction of the Col. Rostr. Aemili mentioned above was the entire loss of any inscription it might have had.

[image ALT: Valid HTML 4.01.]

Page updated: 25 Apr 19