1. AFRANIUS 68 (= NONIUS P. 126.18)

inavdire, audire. M. Tullius ad Pansam lib. I (2) quorum erupit illa uox, de qua ego ex te primum quiddam inaudieram. Nouius Maliuolo (55–6):

Pacuuius Armorum Iudicio (35):

Afranius Diuortio:

The quotation of Afranius’ Diuortium seems perhaps to be an iambic senarius now missing a single anceps at the head of the line. Anon.’s methods of quoting verse suggest that Nonius would have transcribed a complete verse, which implies that the missing anceps has been lost in transmission. There is, therefore, a preliminary argument in favour of a palaeographically simpler supplement than those proposed by Ribbeck and Müller.Footnote ⁵

In comedy the verb in(d)audire is used where a personage has heard or overheard a detail or story that is readily intelligible. There can therefore be no defence of the paradosis, nouis dictis, as an invitation to speak in code. NeukirchFootnote ⁶ and OnionsFootnote ⁷ both attempted to restore a form of dicere but their solutions (~ ‘Tell us, please, so he won't hear’) leave the resulting verse rather flat. Buecheler's ne bis dicatis Footnote ⁸ makes good sense but the phrasing seems too precise for what should be, if that sense is right, a slightly exasperated expression, ‘don't keep harping on it!’ Lipsius’ nobis dictes Footnote ⁹ seems, as it has to many, the best correction offered, although dictare is not without difficulties; the verb is at best doubtful at Lucilius 1139 Marx, and is otherwise not attested before the first century b.c.Footnote ¹⁰ Whatever form of the verb one prefers to restore, the sense of the verse seems clear. Neither Ribbeck's <clam> nor Müller's <ita> harmonizes with the manner of speaking suggested by dictes, and even if a different verb be preferred, the omission of either word could not easily be explained. To complete the verse, I suggest trying:

<non> ṇobis ḍictes uaeso, ṇe ille ĭnạudĭạt.Footnote ¹¹

The speaker tells the interlocutor not to keep repeating him- or herself, to prevent their conversation from being overheard. The omission will have resulted from a lapse over the sequence Diuortio non no- or, if abbreviated, Diuortio no-. For non with a present-tense prohibitive subjunctive, cf. Plaut. fr. 146–7 (= Gell. NA 18.12.4) exi tu, Daue, age sparge, mundum esse hoc uestibulum uolo, | Venus uentura est nostra, non hoc pulueret, and Ter. An. 787 non te credas Dauom ludere.

2. THE FRAGMENT GIVEN AS ACCIUS 95 (= NONIUS P. 39.22)

ordire est ponere (promere Müller), incipere. Accius Amphitryone:

Ribbeck thrice printed Bothe's paru<ol>am Footnote ¹³ without commenting on its difficulties. Diminutive adjectives are not a regular feature of the language of Republican tragedy and their rarity necessarily raises some doubts about Bothe's conjecture, which moves the verse closer to the style of comedy. Furthermore, expressions like ne retice, obsecro (ne with the imperative and a polite modifier) occur only five times in comedy and not otherwise in the remains of Republican tragedy, where hypotaxis is preferred;Footnote ¹⁴ the comparable utterances in comedy do occur, however, in diverse stylistic registers, such that it would be impossible to mark their tone as exclusively comic.Footnote ¹⁵ If the verse belongs to Accius, judgement of its tone and of Bothe's conjecture must be left in suspense, since ne retice, obsecro is just possibly, although quite imperfectly, suited to a tragic treatment of the story of Amphitruo. Müller's <iam> would be a little easier to accept as Accian but it leaves in place the minor obstacle presented by ne retice, obsecro. I suspect that a different, albeit more invasive, explanation is to be preferred.

Nonius took from Anon. perhaps three quotations of Accius’ Amphitruo and three quotations of Plautus’ script of the same title, one of which comes from the extensive lacuna in the Plautine manuscripts following upon Amphitruo 1034. The distribution of utterances like ne retice, obsecro and the frequency of errors over author indications in Nonius raise the question of whether Nonius’ text ought to read, by positing a lacuna, Accius < * * * Plautus> Amphitryone.Footnote ¹⁶ A clear home emerges for this fragment in the Plautine script. It would belong near the beginning of the fragmentary scene between Alcumena and Amphitruo (containing Lindsay's fragments VII–X), probably the only scene in that lacuna in senarii, in which the real Amphitruo accuses Alcumena of infidelity. In that context, this fragment would have Amphitruo stating his intent to voice a serious accusation and a confused Alcumena, who had already been reconciled with her ‘husband’ at Amphitruo 882–955, urging him to speak up.Footnote ¹⁷ Attributing the fragment to Plautus rather than to Accius would diminish the difficulty of ne retice, obsecro. The verse remains metrically incomplete; however, if the attribution to Plautus’ Amphitruo, rather than Accius’, is correct, then Bothe's paru<ol>am seems the most plausible way to restore a complete verse.

3. AFRANIUS 319 (= NONIUS P. 39.22)

ordire est ponere (promere Müller), incipere. Accius Amphitryone (95) … Afranius Suspecta:

The transmitted text of Afranius 319 appears to be a corrupted senarius that is now too long by a syllable. Attempts to recover sense from this verse led editors to rewrite it substantially; their attempts occasionally stretch the limits of credibility and the corruption is perhaps less extensive than has been assumed. The speaker claims that he or she must undertake something in mind (tamen animo ordire oportet me; cf. Pacuvius 379 machinam ordiris nouam); the text and metre of the first five words are unobjectionable. paululum gives a fine cretic ending and as such has a claim to correctness that ought to be entertained. The greatest suspicion attaches to paula: the expression paula paululum is unparalleled and does not make sense,Footnote ¹⁹ while the trochaic word in this position disrupts the metre. I suggest that what is needed is a partitive genitive to depend upon paululum, which would preserve the parts of the fragment which seem secure: ‘I have to undertake, in my mind (i.e., plan), a little something of “X”.’ In place of paula I propose to try operae, so that the speaker will need to come up with something to aid the situation:

tămĕṇ ănĭmo orḍire ŏortet ṃe ŏpĕrae aulŭḷum.Footnote ²⁰

The o- of operae could, under pressure from animo, have been attached to me (animo … meo) and then corrected to me to restore the usual oportet me, leaving perae to be rewritten as paula (no doubt conditioned by paululum). This explanation is speculative, and the conjecture perhaps simply shows a less invasive way to restore the verse, on which others may improve.

4. TURPILIUS 106–7 (= NONIUS P. 75.25)

attigat, contigat (continget LB^A : contingetur (-tig-) C^AD^A). Turpilius Leucadia:

ne me attigat!

atque aufer manum! :: heia, quam ferocula est < × – >

A young woman wards off the unwelcome advances of someone who then speaks of her spunk with some admiration.Footnote ²² While the text of the response is certain, the precise phrasing of her shouts is unclear, and the metre has never been definitively identified. I suspect that the verse is also a complete metrical unit, but that its shape has been obscured by the muddle made of the poor girl's words. Discussion of this fragment must proceed first by examining Nonius’ text of it, before considering the relationship of Nonius’ quotation to what Turpilius is likely to have written.

Nonius’ lemma suggests that the form in the quotation should be attigat; the variant readings continget and contingetur for the definition contigat imply, however, that the form remained obscure and came under considerable pressure from scribes or correctors. The same pressure, I would suggest, gave rise to the reading attiga in Turpilius’ verse in the majority of manuscripts. Setting aside all corrections and conjectures, and considering the reading ne me attiga atque aufer against the lemma attigat, it becomes plain that atque conceals a missing –at, which properly belongs to ne me attigat. The question is then whether atque is correct, and has simply absorbed the ending of the preceding word, or whether atque instead represents an expansion of the dissociated ending. Editors of Turpilius have accepted the first explanation but the second merits some consideration, since the correction of at to atque in a verse that seemed, to a scribe, to offer two imperatives (attiga, wrongly, and aufer) is readily intelligible.

On that explanation, atque would need to be bracketed. The resulting verse, ne me attigat! aufer manum! :: heia, quam ferocula est! has good iambo-trochaic sequences at beginning and end but only a Procrustes could make this text fit into the confines of a single long verse. But Anon. was equally respectful of metrical units when quoting lyric verses. Allowing for hiatus at the change of speaker yields an iambic dimeter joined with a catalectic trochaic dimeter (ia⁴ tr⁴^):

ne ṃe attĭat! auer măṇum! :: ḥeiă, uam fĕṛocŭḷa est!Footnote ²³

Anon.’s rule about quoting complete metrical units supports the view that this fragment should not to be treated as two verses (an iambic dimeter followed by a separate catalectic trochaic dimeter), since Anon. did not need heia, quam ferocula est to illustrate the use of attigat; he could have excluded the second speaker's words, if the utterances had been given as separate verses. When such conversations formed a single verse, however, his methods required that he quote it in its entirety.

This discussion has so far emphasized the restoration of Nonius’ text, which presents a formal expression followed by the more direct and vociferous aufer manum. What relationship that text had to Turpilius’ own words remains to be considered. It is possible to imagine the young woman calling out to another person before she tells her interlocutor to take his hand off her, but Junius’ attigas Footnote ²⁴ perhaps gives more natural dialogue. When a fragment has been quoted with considerable respect for completeness of metre, however, it is dangerous to tinker with its edges, since what strikes us as difficult in isolation would have made good, and often different, sense in its original context. I see no way to solve the problem with certainty and would retain attigat, while acknowledging that attigas has much to recommend it; each form produces the same metrical result.

5. AFRANIUS 279–80 (= NONIUS P. 111.7)

fvam, sim uel fiam. Pacuuius Duloreste (116–17):

Titinius Fullonia (32–3):

Liuius Hermiona (23):

Afranius Promo:

Ennius Hectoris Lytris (150–1 JocelynFootnote ²⁵):

Anon.’s methods are evident throughout this entry, except in the anomalous quotation of Afranius’ Promus and in the difficult fragment of Ennius’ Hectoris Lytra. All editors have set Afranius’ fragment such that si fuas forms the opening cretic of a second, incomplete trochaic septenarius. In the remains of Republican scenic verse, forms such as fŭās, convenient for securing cretic sequences, tend markedly to occur at the end of iambic senarii and at the equivalent metrical positions at the opening of trochaic septenarii and the midpoint of long iambic verses (e.g. Plaut. Amph. 985, Aul. 233, Mil. 299; cf. Aul. 426, at the break in a uersus reizianus). In at least two instances in mixed-metre cantica, however, fuas is set outside those normal positions and must be scanned as a pyrrhic by iambic shortening: Plaut. Persa 51 (ia⁷) sed recipe quam primum potes, caue fuas mi in quaestione; Trin. 268 (in an anapaestic system) Amor, mihi amicus ne fuas umquam.Footnote ²⁶ The similar language of Persa 51 and Afranius 279–80 is perhaps significant. In Afranius’ fragment, scanning fuas as a pyrrhic restores a bacchiac tetrameter, which Anon. will have quoted as a complete metrical unit:

Bacchiac tetrameters with a resolved anceps are admittedly rare in Plautus; such resolutions that do occur, though, are more common in the first and third ancipitia than in the second or fourth.Footnote ²⁷ Tetrameters with two resolved ancipitia are even rarer still, which perhaps leaves room for doubt over the scansion I propose, but we cannot know whether Afranius eschewed such resolutions to the degree that Plautus did. The single parallel for a bacchiac tetrameter with resolved first and third ancipitia is provided by Plaut. Mostell. 89 similem esse arbitrarer simulacrumque habere.

6. AFRANIUS 285–6 (= NONIUS P. 515.8)

svperbiter. Afranius Purgamento (Priuigno ‘amanti Lipsius : Pantaleonte dubitatim Junius):

Naeuius Belli Poenici lib. VI (37 Strzelecki):

Afranius’ fragment is problematic, and certainty about any solution is unattainable. I see three possible solutions to the difficulties it poses, of which the third strikes me as the most probable, when taken from the perspective of Anon.’s methods.

1. (1) Ribbeck printed his conjecture and set the fragment as ita superbiter | imperat, two trochaic snippets, leaving in his apparatus Buecheler's suggested transposition illa imperat superbiter, which restores two iambo-trochaic metra. It was not only the problem of the split and exposed resolutions in illă sŭperbĭtĕr imperat that prompted Buecheler to make this correction, for it also allows superbiter to stand at verse-final position, which, together with the opening three elements of the trochaic septenarius, is the regular position occupied by adverbs in –iter that are formed on adjectives in –us in iambo-trochaic verse.Footnote ²⁸ This solution would nevertheless leave an exception to Anon.’s habit of quoting complete metrical units, which could perhaps be explained with the assumption that the quotation came to Anon. as a marginal annotation in his copy of Naevius’ Bellum Poenicum.

2. (2) The title Purgamentum is nowhere else attributed to Afranius, and has occasionally attracted suspicion since the sixteenth century; Lipsius’ conjecture is just slightly more compelling than that of Junius.Footnote ²⁹ Buecheler's transposition, applied to Lipsius’ guess, restores an iambo-trochaic rhythm but still yields less than a complete metrical unit. A complete senarius could be restored if one assumes more extensive corruption than even Lipsius suggested, and tries: ‘Afranius Priuigno: amenti <amanti> illa imperat superbiter’. It will be obvious that this suggestion is somewhat fanciful; it attempts to explain, with a nod at Plaut. Merc. 82 and Ter. An. 218, the sequence -ament- in the title Purgamentum, if indeed that title is corrupt.

3. (3) Just as was observed to have been the case with a form like fuas, adverbs in –iter formed on adjectives in –us are set with greater freedom in lyric verses and need not be restricted to the metrically significant positions at the end of an iambic senarius and the equivalent BcD positions.Footnote ³⁰ The transmitted words illa superbiter imperat scan without difficulty as a complete glyconic ( × × – – × – ):

Cf. Plaut. Mostell. 721 haud bŏnum tĕnĕo seruum and, for the resolved and split second anceps, Pseud. 603 iam pŏl ĕgo hunc strătĭotĭcum.Footnote ³¹ This setting seems to be the most plausible solution to the problem, for it respects both the transmitted text and Anon.’s methods.

7. LIVIUS, TRAG. 19 (= NONIUS P. 473.26)

minitas, pro minitaris. Liuius Danae:

Editors have treated the fragment of Livius’ Danae as an acephalous trochaic septenarius; in his apparatus Lindsay pondered an anapaestic scansion. The former does not accord with Anon.’s methods, while an anapaestic rhythm is most implausible. Buecheler's uitia is tidy but perhaps goes too far, since the subsequent clause more than adequately explains ea and the conjecture provides no benefit aside from filling out the hypothesized lacuna. Such a correction suits what I have argued were Anon.’s methods but an alternative solution is available, and I suggest trying:

tĭam ṃĭnĭtas? ṃitte ĕạ,

uae tŭă ṣunt măgĭs uam mĕạ.

The second verse could be treated as a tidy glyconic but the first is rather more difficult to accept as the same scheme, and it is simpler to take this fragment as two catalectic trochaic dimeters; cf. the contrasting treatments of Plaut. Men. 985 <eo> ego exemplo seruio, | tergo ut in rem esse arbitror offered by Lindsay (glyconics) and Questa (tr^{4^}).

8. POMPONIUS 10–11 (= NONIUS P. 516.15)

pvriter. Pomponius Aruspice uel Pexore Rustico:

Lindsay set the fragment as two incomplete iambic septenarii, Ribbeck as senarii. The latter's solution complies with Anon.’s methods but his supplement <rem> is not obviously right. BergkFootnote ³² set the verse as a complete trochaic septenarius by deleting uti (which has the look of an insertion to clarify the syntax of the quotation): “bucco, puriter fac tractes.” :: “laui iamdudum manus.” That approach is attractive but a new problem arises in the metrical position of puriter. Pomponius no less than other Republican dramatists seems both to have exploited such adverbs to secure cretic sequences, and, with only slightly less rigour than his predecessors, to have restricted them to metrical boundaries.Footnote ³³ Bergk's scansion is therefore still somewhat problematic, since it puts puriter in an anomalous DaB setting. It seems preferable to transpose the initial words, adopting also Bergk's deletion of uti, to produce the septenarius:

9. ACCIUS 119–21 AND LIVIUS, TRAG. 2–4 (= NONIUS P. 512.31)

aeqviter, pro aeque. Pacuuius Atalanta (73):

Accius Antenoridis:

Liuius Aegistho:

Lindsay assigned the quotation of Pacuvius 73 to Nonius’ ‘Lucilius i’ (9) list, Rychlewska to the Pacuvius (12) list. Both explanations are somewhat problematic and the precise origin of that quotation must remain uncertain.Footnote ³⁴ The added quotations in this lemma come from two scripts that are known to Nonius only through Anon., but the quotations are transmitted in forms that appear to offend against Anon.’s methods, in that both begin and end with incomplete metrical units. They accordingly come under suspicion. It would be possible to consider the pair to have resulted from a momentary lapse in Anon.’s diligence, but the arrangement of Accius’ fragment poses a metrical difficulty that cannot be so easily dismissed.

As Ribbeck printed them, Pacuvius 73 and Livius, trag. 2–4 show aequiter at verse end, in one of the positions expected for similarly formed adverbs. The quotation of Accius does not, and while Pomponius evidently felt justified in setting such adverbs at other BcD positions in the interior of a verse, evidence that Accius was similarly footloose on this point is not forthcoming.Footnote ³⁵ There are, therefore, two problems in the Accian fragment, in the metrical setting of aequiter and in the violation of Anon.’s methods. Both difficulties would be eliminated by transposing the opening of the fragment to produce iambic septenarii:

inṭellĕo rĕfĕṛundum

ad ŏpŭlum, uŏnĭam hoṛum aequĭṭer senṭentĭạe fŭẹre.

On this metrical setting, aequiter would stand in a less common but attested position before the mid-line break of long iambic verses,Footnote ³⁶ and Anon.’s habit of beginning or ending his quotations with a complete verse would be restored.Footnote ³⁷

That conjecture invites reconsideration of the fragment of Livius’ Aegisthus, since the exceptional form of these quotations becomes considerably more suspicious when one of them can be eliminated. In addition to the metrical problem, the transmitted text presents a difficulty in the light of Anon.’s secondary concern for giving complete units of sense. The fragment presents two co-ordinated verbs (accensa [sc. est] and partita est) within a single temporal clause introduced by ut, but, on that analysis, the quotation lacks a principal clause that would have been introduced by nam, which leaves the thought incomplete (‘for, when Pergamon was set ablaze and the booty was distributed equally among the participants …’).Footnote ³⁸ That manner of quotation would be unusual, since, in the transmitted text, ut marks a sense boundary and what follows ut constitutes a complete unit of sense and consists of at least one complete metrical unit; in such a case, rather than including nam without its own complete clause, Anon. is more likely simply to have omitted nam.Footnote ³⁹ These factors suggest that the original main clause, introduced by nam but postponed by the temporal ut clause, has been absorbed into the temporal clause through an error in transmission. The guilty party is perhaps et, which could be a later addition supplied to make sense of a quotation that had lost the crucial word which put accensa properly into its own subordinate clause. It is possible that the -a that Junius had to restore to Pergam<a> in fact took more of the original text with it when it was lost. To solve these difficulties I suggest trying the iambic octonarii:

năm ŭt ergăṃa est acensă, raedă er parṭĭcĭpeṣ aequĭṭer

parṭita est.

This supplement sets ut Pergama est accensa off as its own subordinate clause, dependent upon the main nam-clause whose verb is partita est (‘for, when Troy was set ablaze, the booty was distributed equally among the participants’). Patterns of speech and the metrical positioning of marked forms such as aequiter that are customary on the Republican stage would thereby be restored, along with Anon.’s typically diligent methods. The correction would also further reduce the distance between Livius’ fragment and its echo in Sen. Agam. 421–2, ut Pergamum omne Dorica cecidit face, | diuisa praeda est, maria properantes petunt, on which see Tarrant, ad loc.Footnote ⁴⁰

10. POMPONIUS 65 (= NONIUS P. 508.19)

venibo pro ueniam. Pomponius Maccis Geminis:

On six other occasions Nonius preserves quotations from Pomponius’ Macci Gemini, three of which certainly and the remaining three (which are iterations of the same fragment) plausibly derive from Anon. In this fragment the placement of uenibo, identical with the lemma, at the end of the quotation calls attention to itself, for in such instances of dubious text and metre it is sometimes the case that the lemma word within a quotation has become dislodged from its proper position.Footnote ⁴¹ Lindsay suspected, with a great deal of plausibility, that in cases where the lemma was identical with the form of the word in the quotation, the immediate archetype of our manuscripts of Nonius had indicated the word with its first letter only.Footnote ⁴² Such abbreviations were liable to disappear and, later, to require the quotation to be rewritten to bring it into line with the lemma.

The first hand in L gives an aberrant reading that is perhaps significant, with sib for si. Editors have ignored the intrusion. The frequent interchange of b and v might suggest that this is the missing abbreviation for the lemma, but a verse like sero est sexta si uenibo tibi placet is sheer nonsense. Attempting to make sib yield si uenibo must therefore be abandoned, but it is worth considering whether sib conceals something else. Another way to make sense of the fragment would be to try the senarius:

seṛo est; uĕṇibo ṣexta, ṣi <illud> ṭĭbĭ plăet.

The supplement <illud> is no more than a guess aiming to restore a word that, if abbreviated, may have given rise to the intrusive b of L ¹, and other options are worth considering.Footnote ⁴³

11. AFRANIUS 2 (= NONIUS P. 374.28)

pariter, ex aequo. … Afranius Abducta:

Although early editors realized that pariter had fallen out of the quotation of Afranius 2, they restored it without regard for metre, which habit can be traced even in Bothe's edition of the fragments of Republican drama. Bothe'sFootnote ⁴⁴ second attempt at repairing this fragment provided nineteenth-century students with their vulgate: dispertiantur < – > patris pariter bona. That text, while plausible, ignores the reading of L ¹, which hand is in general remarkably faithful to the archetype and its peculiarities. If pariter is correctly restored before bona, one could conjecture in the archetype the sequence PARTRISPAR or perhaps simply PARTRISP, from which the abbreviation for pariter fell out. Instead of reading patris with the corrector of L I suggest restoring from PARTRISP(AR) the following text, which scans as a complete senarius:

disertĭạntur arta ĭịs părĭṭer bŏṇa.

‘Let their possessions’ (dative iis with participial substantive parta) ‘be divided up equally’ or perhaps ‘let (their) possessions be divided up equally between them’ (dispertiantur … iis, the equivalent of Livius’ praeda per participes aequiter | partita est [see § 9, above] or of Plaut. Asin. 271 praedam pariter cum illis partiam). This conjecture requires an iambic pronunciation for iis in order to restore a full senarius. Iambic scansion of the pronoun is rarer than synizesis, but is attested in earlier dramatic verse at Ter. Eun. 250, Plaut. Curc. 371–3 and Rud. 73.Footnote ⁴⁵

12. CAECILIUS 221–2 (= NONIUS P. 258.38)

contendere significat conparare. Plautus in Vidularia (fr. XII) … Lucilius Satyrarum lib. I (24) … idem lib. VII (4) … Caecilius Titthe:

egon uitam meam

asticam contendam cum istac rusticana Syra? <  – >

Licinius Macer in epistula ad senatum: illi suam uitam mecum contendunt, quorum in corpore ita crebra sunt uulnera †uitae†, nouae cicatrici ut locus non sit?

Naeuius in Lycurgo (36):

Accius in Tereo (647–8) … Varro Γνῶθι σɛαυτόν (210) …

The origins of the quotations from Caecilius and Licinius Macer are obscure; Lindsay, with some hesitation, attributed both to list 27. The order of the quotations in this entry is problematic (apparently from lists 2, 9 bis, 27 bis, 4, 6 and 31) and considerable uncertainty must remain about the degree to which Nonius adhered to his methods here. Elsewhere Nonius preserves five quotations of Caecilius’ Titthe, of which three certainly and possibly all fiveFootnote ⁴⁷ come from the works of Anon. It is plausible but uncertain that the quotation from Caecilius derives from Anon.; much greater doubts must remain about the quotation from Licinius Macer.Footnote ⁴⁸ Some support for the attribution of the quotation of Caecilius to Anon. would be provided by the following conjecture, which would restore Anon.’s normal methods to the fragment.

Licinius Macer's words are vitiated by the intrusive uitae, which is superfluous and even detrimental to the sense; uulnera uitae as ‘the wounds of life’ is scarcely credible, and certainly not the Latin of correspondence with the senate. Müller bracketed the word, suspecting that it had intruded under the influence of uitam in the opening clause of the same quotation. He perhaps did not look far enough afield. Another explanation is that uitae originated as a marginal correction that was wrongly interpreted as belonging to Licinius Macer and was altered to fit that context, when it perhaps belonged instead to Caecilius’ Titthe.Footnote ⁴⁹ That fragment could then be restored to read:

gŏn uiṭam mĕạm

ạstĭam conṭendam um istac ṛustĭa <uiṭa >, Sṛa?

This correction eliminates a number of difficulties beyond explaining the origin of the intrusive uitae in Licinius Macer's words. It replaces the unusual form rusticana with rustica; rusticanus is not otherwise attested before Cicero, and it is attractive to think that the -na of rusticana is the last vestige of the original uita. It also aligns the quotation with the form of expression evident in iram … cum ira in Naevius, trag. 36, quoted above. Finally, the conjecture restores a complete trochaic septenarius at the end of the quotation. The metre is not itself a decisive point, given the lingering doubts about the origin of this quotation, but since Nonius’ other glossaries quote units of sense without regard for metre, it removes an obstacle to the attribution of this quotation to Anon.

13. TITINIUS 56–7 (= NONIUS P. 256.10 AND P. 510.5)

p. 256.10:

conparare ueteres confirmare et constituere dixerunt. Terentius in Eunucho (47–8):

idem in Andria (627–8):

Titinius Gemina:

Accius Eurysace (371–2) … idem in Tereo (634–5) … Turpilius in Lemniis (99) … M. Tullius de Republica lib. VI (1) … Afranius Epistula (116–7) … Cicero de Finibus Bonorum et Malorum lib. V (86) …

p. 510.5:

blanditer pro blande. Titinius Gemina:

In the present and subsequent sections of this paper, my attributions of quotations to Anon. occasionally depart from the lists of Lindsay and Rychlewska.

The quotation of the second half of the Titinian fragment at p. 510.5 occurs in a series of lemmata that both Lindsay and Rychlewska attributed to ‘Gloss. i’.Footnote ⁵⁰ Many entries in that sequence can scarcely be assigned to ‘Gloss. i’ on grounds of both the particular literary works citedFootnote ⁵¹ and the manner in which those quotations are given,Footnote ⁵² and I here propose a different analysis of the opening of Book 10. There are clear and predictable signs of Nonius’ usual methodology and of his usual errors at the outset of this book. In the following sequence (p. 509.18–p. 510.32), Nonius has committed a typical error, where he continues to use a list that has just given him an added quotation, to write new lemmata, instead of returning to the list that had provided the previous lemmaFootnote ⁵³ (I give my own attribution in the third column, Rychlewska's in the fourth):

Nonius, taking up ‘Gloss. iii’ to provide an added quotation of Livius that illustrated disertim, then continued to use that list for a new lemma, instead of returning to list 1. When he reached mvnditer and its quotation of the Poenulus (given in complete metrical units, in Anon.’s style), Nonius realized his ‘mistake’ and returned to his ‘Plautus i’ list, picking up, not from the beginning of the ‘Plautus i’ list, but from the Poenulus.Footnote ⁵⁴ The quotation of Poen. 335, along with the snippet of Trin. 1060, is given without regard for completeness of the metrical units, as is more characteristic of Nonius’ own habits.Footnote ⁵⁵ There would therefore be no difficulty in assigning the quotation of Titinius illustrating blanditer to a work of Anon. here; and a considerable number of peculiarities would be eliminated on this hypothesis.

The case is clearer still at p. 256.10. Lindsay assigned the quotations of Terence and Titinius to ‘Gloss. i’, and the remaining quotations to the ‘Accius i’ (5), Turpilius (11), ‘Cicero i’ (13), Afranius (19?) and ‘Cicero vi’ (34??) lists, although the last two items are problematic, for the Afranius volume did not certainly contain the Epistula and the ‘Cicero vi’ list always gives its author's name as ‘M. Tullius’, not as ‘Cicero.’ The C-section of Book 4 begins with a long sequence of lemmata, all assigned to ‘Gloss. i’ by Lindsay, in which Nonius seems rather to have made extensive use of the Virgil (22), Terence (23), ‘Cicero iv’ (24), ‘Lucilius ii’ (25), ‘Cicero v’ (29), ‘Plautus ii’ (30) and ‘Varro ii’ (31) lists. Many of the quotations in those pages that cannot reasonably be attributed to those lists (since they derive from other authorsFootnote ⁵⁶) suit the interests and methods of Anon., whose works fill the gap in the sequence of lists just enumerated. More immediately locally, at several points in the broader discussion of conparare Nonius has exploited the scripts of Terence (p. 255.34, p. 256.34) and it is reasonable to conclude that the two quotations of Terence here derive from list 23, not the first glossary; and that at least the Titinian fragment and perhaps others of the remaining seven quotations derive from the works of Anon., for they are completely of a piece with his interests and methods.Footnote ⁵⁷ (For a comparably developed entry, cf. § 15 below.)

Allowing the probability that this quotation of Titinius’ Gemina has come to Nonius from Anon., and not from ‘Gloss. i’, raises the question of whether the quotation, which editors have tried to stretch or supplement in various ways to produce iambic octonarii or impossible trochaic septenarii,Footnote ⁵⁸ is in fact to be scanned differently. The placement of blanditer presents a neglected difficulty, since its position even at the mid-line break in a long iambic verse is slightly anomalous and would perhaps require that this fragment be treated as part of a polymetric canticum, as with Accius 119–21.Footnote ⁵⁹ Even allowing for such a treatment, the fragment does not preserve perfect octonarii: the missing element in verse 57 is problematic, and it is worth trying other scansions. The stumbling block is dōna (56), which, if correct, must be set metrically as Da. But reading bŏna, suggested first by Hermann as part of a different restoration of the text,Footnote ⁶⁰ would give four catalectic trochaic dimeters requiring no further supplementation, completely of a piece with the methods and habits of Anon.:

ẹrgo ut ṣŭă serụet bŏṇa

ṇobis ăcĭunḍum est ĭṭa,

ụt cum rimum ḅlandĭṭer

ompăṛemus onlŏui.

14. AFRANIUS 100–1 (= NONIUS P. 517.10)

desvbito. Titinius Prilia (80) … Ennius Pancratiaste (3) … Afranius Emancipato (100–1):

Caecilius Asoto (9) … Quadrigarius Annali lib. II (38) … Novius Tripertita (89) … Lucretius lib. II (265) …

Lindsay and Rychlewska both attributed the impetus for Nonius’ entries on desvbito and derepente, at the opening of Book 12, to ‘Gloss. i’.Footnote ⁶¹ This explanation cannot be accepted. Book 12 bears the title De doctorum indagine, itself more suggestive about Nonius’ methods in this book than has been acknowledged. The two entries are introduced with Nonius’ preening claim to be offering his readers the truth: omnes artem secuti negant aduerbiis praepositionem addi oportere; sed auctoritas ueterum praeponi iubet. The standard account of Nonius’ use of his lists requires that we accept that Nonius, aiming to show that a preposition could be prefixed to an adverb, took no material from the alphabetical list of adverbs to write the entry for desvbito and only used that list for derepente when he had already written most of the entry.Footnote ⁶²

Despite that claim, Nonius’ entry for desvbito reveals the familiar material and habits of Anon., whose quotations have been augmented with some that were drawn from Nonius’ other lists. The entry gives quotations from Titinius’ Prilia, Ennius’ Pancratiastes, Afranius’ Emancipatus, Caecilius’ Asotus, Quadrigarius’ Annales and Novius’ Tripertita, which set of source texts seems far closer to Anon. than to ‘Gloss. i’, before added quotations taken, in predictable order, from the Lucretius (3), ‘Accius i’ (5), Pomponius (6), ‘Lucilius i’ (9), ‘Cicero i’ (13) and Naevius (21) lists.

Nonius did not himself make use of a complete script of Afranius’ Emancipatus. He gives three other quotations from the script, two of which certainly (p. 89.5, p. 109.10) and the third plausibly (p. 496.31) are taken from Anon. There is therefore cause to reconsider the text and metre of this fragment. It has consistently vexed editors, who have reshuffled and supplemented its words to produce a sensible and metrical text.Footnote ⁶³ The most conservative, albeit still unmetrical, text is understandably that of Lindsay, whose setting of the fragment as pieces of trochaic septenarii clung, mostly desperately, to the paradosis. To the problematic state of the text it may now be added that his scansion is contradicted by Anon.’s method of quoting verse. A different scansion is necessary but it is far from clear that reshuffling the words will yield the palmary solution.

The manuscripts give: quam beatae scenicae (scentce L : scente A) uidentur mihi (mihi uidentur BrugmanFootnote ⁶⁴) mulieres quae iurgio et beniuolentia terrent desubito uiros. Brugman's transposition seems essential, in order to avoid a number of metrical barbarisms. Not much else has been considered sacrosanct. But the text of the fragment gives sequences at its beginning and end that could be scanned without difficulty as catalectic trochaic dimeters: uam bĕạtae ṣcenĭc.ae on the one hand, and ṭerrent ḍesŭbĭtọ uĭṛos on the other. Given the methods of Nonius’ source, it is possible that the fragment belongs to a polymetric canticum and that the dimeters flank a single long verse.

Difficulties remain in mihi uidentur mulieres, quae iurgio et beniuolentia. The least invasive solution that has been offered is Passerat's uiolentia for beniuolentia, although it is somewhat difficult to see by what intermediate steps that change could have been effected. I suggest trying uolentia, deleting beni- as an explanatory annotation of a rare word, which note was then promoted into the Afranian quotation; the corruption could have happened at a number of points in the transmission of Afranius’ words. The result is a trochaic septenarius flanked by two catalectic trochaic dimeters:

uam bĕạtae ṣcenĭae

ṃĭhĭ uĭḍentur ṃŭlĭĕṛes, quae ịurgĭọ et uŏḷentĭạ

ṭerrent ḍesŭbĭṭo uĭṛos!

For just such a metrical run, cf. Plaut. Pseud. 222–4. The guess about uolentia is obviously uncertain. The word is not otherwise attested before Fronto (p. 112.22 van den Hout) and Apuleius (Met. 5.31.3, by emendation; 11.6.1), but it is as possible that they were reviving an old word as coining a new one. The change to uolentia would also diminish the awkwardness that some readers have felt in the juxtaposition of iurgium and beniuolentia,Footnote ⁶⁵ since uolentia comes across considerably more forcefully as ‘will’ or ‘intent.’

15. AFRANIUS 203 (= NONIUS P. 306.26)

facessere est facere. Ennius Annalibus lib. I (57 Skutsch):

Afranius Inimicis (202):

Vergilius Aen. lib. IV (295)

facessere significat recedere. Ennius Eumenidibus (136):

Titinius Gemina (51–2):

Pacuuius Teucro (342):

item (326):

Cicero de Legibus lib. I (39): paulisper facessant rogemus. Afranius Liberto (203):

Plautus Rudente (1061–2):

M. Tullius in Hortensio (49): facessant igitur omnes qui docere nihil possunt, quo melius sapientiusque uiuamus.

In the discussion of facessere Lindsay marked as primary quotations those of Ennius’ Annales and Eumenides and of Titinius’ Gemina, which he attributed to ‘Gloss. i’. The added quotations from Plautus’ Rudens and from Cicero's Hortensius were attributed to lists 2 and 29, respectively. Lindsay's silence on the sources of the intervening quotations implied that they derive similarly from the first glossary. Much of that analysis is implausible. The works of Republican literature cited and the manner in which they are quoted point to a work of Anon., rather than to ‘Gloss. i’, as the source of the quotations.Footnote ⁶⁶ There remains room for doubt over the quotation from the Rudens Footnote ⁶⁷ while that from the Hortensius is surely from list 29.

The quotation from Afranius’ Libertus therefore attracts suspicion since it is, as students since Hermann have printed it, an iambic verse that is deficient at line end, in contravention of Anon.’s methods. If my claim that this quotation was taken from Anon.’s compendia is correct, it is essential to set the nine words as a complete verse. They can be treated as a uersus reizianus, with a characteristic strong break between the iambic dimeter and the colon reizianum:

16. AFRANIUS 223 AND COM. INC. 22 (= NONIUS P. 518.3)

derepente. Turpilius Boethuntibus (3) … Ennius Phoenice (268) … Afranius Incendio (198–9) … idem in Omine (223 and Com. inc. 22):

In the remaining sections of this paper I take up certain of Nonius’ attributions of metrically complete verses to particular dramatic scripts, which attributions have sometimes come under suspicion.

As in the case of Afranius 100–1 (§ 14, above), the primary quotations in the entry derepente seem to derive rather from Anon. than from ‘Gloss. i’. Among those quotations are the two verses printed above, which are given in the manuscripts as a single fragment from Afranius’ Omen. Lipsius saw that the two verses were scarcely intelligible as a single utterance, and since both preserve the target word derepente he sought to insert < idem> between them. That conjecture still attributes the verse contendit oculos, derepente abiit †colo† to Afranius. Ribbeck, preferring <item>, printed the second verse among the incerta incertorum, as Com. inc. 22. In the previous paper I suggested that the quotations of Afranius’ Omen were not set down by Anon. himself but by a later reader, who added quotations in the margins of Anon.’s works but did not adhere to his method of quoting verse.Footnote ⁶⁸ On that view one must assume that both verses came to Nonius from marginal annotations. It would have been easy for Nonius to overlook an indication that the second verse was not connected with the first, if the marginal annotator had even set down that information. I suspect, therefore, that the error is not owed to the transmission of Nonius’ text but to his source, and that Nonius accurately, but wrongly, transcribed the two verses as one fragment.Footnote ⁶⁹ The second verse could as easily belong to any other dramatist as to Afranius, and it must remain in the ranks of the incerta incertorum.

17. FRAGMENTS ATTRIBUTED TO CAECILIUS’ VENATOR

Nonius twice attributes quotations to a script of Caecilius entitled Venator. At p. 483.18, the quotation comes from either list 26 or list 38.Footnote ⁷⁰ At p. 42.19, the quotation has been dislodged from its proper home but appears to conclude a sequence of entries taken from source 28. The evidence is therefore uncertain but there is some probability that the quotations, which are given in complete metrical units in accordance with Anon.’s habits, belong to lists 26 and 28.

Spengel, doubting the form of the title Venator, assigned the two quotations instead to the Obolostates siue Faenerator; on that view, Venator would represent a corruption of the simple title Faenerator. This proposal was adopted by Ribbeck and given tacit approval by Lindsay in the index to his edition of Nonius, but it cannot be accepted so easily. Two quotations in Nonius are given to a Caecilian script entitled Faenerator, which perhaps both come from ‘Gloss. v’ (p. 149.27 and p. 543.20, both metrically incomplete). By contrast, Anon. knew that script only under the single title Obolostates (p. 98.4, 154.10, 277.28, 279.24, and probably p. 508.7), and shows no signs of having known the title Faenerator. If the two quotations attributed by the manuscripts to Venator are rightly thought to have descended from Anon., they cannot belong to the script we know as the Obolostates siue Faenerator. Guardì preserved the title Venator in his edition of Caecilius for other reasons, but that conservatism is in all likelihood warranted.Footnote ⁷¹ However implausible one considers Venator to be as a Caecilian title, stronger arguments than have been adduced, and a script other than the Obolostates siue Faenerator to which these fragments should be attached, will be needed before we may dislodge it.

18. POMPONIUS SECUNDUS 1–2 (= NONIUS P. 144.21)

notificem, notam faciam. Pomponius Atreo (Pomponius * * * <Accius> Atreo Müller : Pomponius Patruo dubitanter Lindsay):

Doubts have often been raised about Nonius’ single quotation attributed to the Neronian tragedian Pomponius Secundus, and attempts have sometimes been made to attribute the fragment to the author of Atellanae or to Accius.Footnote ⁷² The fragment and its attribution are of considerable importance for a variety of reasons, not least of which, in the present context, are their implications for the period in which Anon. wrote. In the previous paper I noted that it did not seem possible to specify the period of Anon.’s activity more precisely than to note the termini established on one end by the quotations of Varro, Cicero and their contemporaries, and by Nonius himself on the other end. This fragment is obviously weak evidence for that argument, but it is worth indicating the possible explanations of it, not least since the investigation of Anon.’s methods brings new information to bear on its possible attributions.

Three approaches to this fragment present themselves. First, it is possible that Anon. included a fragment of the Neronian tragedian,Footnote ⁷³ and that Nonius, who usually ignores later material in his glossaries,Footnote ⁷⁴ did not realize that the later tragedian, and not the author of Atellanae, was meant. Students of Nonius have placed a little too much trust in his abilities to separate wheat from chaff, and it would seem easy for Nonius either not to have known or not to have noticed that this Pomponius did not belong in his canon of ueteres.

Resistance to attributing this fragment to Pomponius Secundus led students of Nonius to attribute the fragment instead to a Republican dramatist, as on the conjectures of Müller and Lindsay recorded above. Both are possible, but Lindsay's has a slight edge over Müller's. The reason for that preference has to do with Anon.’s other quotations, for Anon. quotes from Pomponius’ Patruus once (p. 474.26) but never, it would seem, from Accius’ Atreus; Nonius’ five quotations from the Accian script all seem certainly or probably to derive from the ‘Accius ii’ (8) list.Footnote ⁷⁵ Lindsay's conjecture is therefore a little easier to accept, but there are enough examples of scripts quoted only once in what remains of Anon.’s work that Müller's suggestion cannot be excluded entirely, nor need Lindsay's be the only satisfactory conjecture that would restore this fragment to Pomponius.

One further possibility, although never much in favour, is that the text is correct and refers not to the Neronian tragedian, but to the author of farces, as Hertz suggested.Footnote ⁷⁶ The indirect tradition preserves references to several scripts of Atellan farce that bore titles redolent of tragedy. Pomponius is credited with an Agamemno Suppositus, an Armorum Iudicium, a Marsya and, if Porphyrio (on Hor. Ars P. 221) is to be believed, an Ariadne, an Atalante and a Sisyphus. Novius is given an Andromache and a Phoenissae, and Juvenal (6.71) speaks of an Atellan Autonoe. There is therefore no external obstacle to making this fragment derive from an otherwise unattested Atellan Atreus written by Pomponius.

The most significant argument adduced in favour of attributing this fragment to Pomponius Secundus or Accius, and not to the author of Atellan farce, is that the tone of the fragment suits tragedy better than it does farce. The loftier tone implicit in the syntax of nunc te obsecro | … ut has already been pointed out elsewhere (cf. n. 14), and there is indeed a certain pomposity in the language that is redolent of the language employed by Republican tragedians. But it should also be noted that the tone of this fragment finds a counterpart in a fragment of Pomponius’ Armorum Iudicium (9^1–2): tum prae se portant ascendibilem semitam | quam scalam uocitant. Making allowances for farce's ability to prick the balloon of its own paratragedy, one may also point to the language of fragments of Novius’ Andromache (4^a–b), quod tu, mi gnate, quaeso ut in pectus tuum | demittas tamquam in fiscinam uindemitor, and of Pomponius’ Agamemno Suppositus (4–5) ne quis miretur, cum tam clare tonuerit, | ut, si quis dormitaret, expergisceret. Since Atellan farce no less than the other comic genres could ape the particular style of tragedy, it seems worth asking again why these words could not belong to a farcical script entitled Atreus.

There seems to be no way to judge reliably between these possibilities, and the question of the attribution of this fragment must be left open, even as closer attention to Nonius’ sources provides more information about these options than we had before. It would, in any case, be hazardous to use this fragment to refine our sense of the period when Anon. was working.