The Need for a New Text of Catullus *It might seem surprising that I should wish to argue that there is a need for a new text of Catullus. He is, after all, not an obscure author who has been unjustly ignored or infrequently edited. Currently there are (amongst others) two eubner editions of Catullus, one by !ardon from "#$% and one by &isenhut from "#'%, though only the first is still in print, an (xford Classical ext of "#)' by *ynors, a brief but important edition by +oold from "#'%, and in "##$ the second edition of homson,s major text (the first appeared in "#$') ". !ut there is, I thin-, a need for a new critical edition. o explain why, I shall di.ide this paper into three sections. /irst, I shall summarise what is now -nown about the transmission of the text, which shows what -ind of textual tradition we are dealing with and suggests in general terms how itmight be approached by an editor. 0econd, I shall loo- at some of the most important modern editions of Catullus, and explain where they are open to impro.ement. 1nd third, I shall gi.e some examples of the -ind of progress that can still be made in trying to reconstitute the text of Catullus through emendation and conjecture, despite the fi.e centuries and more of continuous scholarly attention which ha.e been paid to it. 1 : The Transmission of the Text of Catullus : from antiquity to Scaliger 2espite the complex appearance of modern stemmata (e.g. homson ("##$) #%), there are basically four extant manuscripts of real importance for the text of Catullus 3. (ne, , from the ninth century, is a florilegium which contains only one poem of Catullus, 43, where it pro.ides a .ery helpful early witness, but of course it does nothing for the tradition of the other poems. hese other poems are preser.ed in * I am most grateful to 52 2r. Christiane 6eit7 for her -ind in.itation to spea- at the *annheim conference and for her editorial indulgence, and to my (xford colleague 2r. 0tephen Heyworth, both for commenting on this paper and for our long collaboration on Catullus 8cf. Harrison and Heyworth ("##')9. han-s are also due to 5rof.6.+.*.:isbet for much discussion of problems in the text of Catullus.1 !ardon ("#$%), &isenhut ("#'%), *ynors ("#)'), +oold ("#'%), homson ("#$') and ("##$).2 his account of the manuscript tradition of Catullus is hea.ily dependent on the .eryfull and learned account in homson ("##$) 33;#3< for a good summary of the facts see arrant ("#'%).many manuscripts, but the earliest and most important are all from the last third of thefourteenth century, (, + and 6. (, + and 6 plainly descend ultimately from a single archetype, but the greater textual proximity of + and 6 ma-e it clear that these two descend together from an intermediate copy of the archetype which was the direct source of (, a copy usually referred to as =. >ntil recently, the archetype which is the common source of (+ and 6 was thought to be the lost manuscript called ? by editors. his manuscript seems to emerge (appropriately enough) in the late part of thethirteenth or the early part of the fourteenth century in ?erona, the poet,s birthplace, and is clearly a.ailable to .arious 5aduan and ?eronese humanists in the period "3#@;"%"@. !ut only twenty years ago, *cAie pro.ed conclusi.ely that ( and =, the common source of +6, were not copied direct from ? but must ha.e together deri.ed from a lost intermediate source, which editors now call 1< this conclusion is deduced fromthe titles and di.isions of the poems in these .arious manuscripts %.he common ancestor of (, + and 6, the three extant *00 which contain the whole of Catullus, whether it was ?, probably Carolingian in date, or 1, probably copied when ? emerged in ?erona, was e.idently full of corruption. he *0 + preser.es a complaint which may go bac- to the scribe of 1, but is in any case from the fourteenth century, which holds good today and is worth repeating B CTu lector quicumque ad cuius manus hic libellus obvenerit Scriptori da veniam si tibi cor[r]uptus videtur. Quoniam a corruptissimo exemplari transcripsit. Non enim quodpiam aliud extabat, unde posset libelli huius habere copiam exemplandi. Et ut ex ipso salebroso aliquid tamen sugge[re]ret decrevit pocius tamen cor[r]uptum habere quam omnino carere. Sperans adhuc ab aliquo alio fortuito emergente hunc posse cor[r]igere. alebis si ei imprecatus non fueris.,Dou, reader, whoe.er you are to whose hands this boo- may find its way, grant pardon to the scribe if you thin- it corrupt. /or he transcribed it from an exemplar which was itself .ery corrupt. Indeed, there was nothing else a.ailable, fromwhich he could ha.e the opportunity of copying this boo-< and in order to assemble something from this rough and ready source, he decided that it was better to ha.e it in 3 0ee *cAie ("#$$), discussed and appro.ed by homson ("##$) 3$;'.4 (n this subscriptio see *acAie ("#$$) and homson ("##$) %";%.3a corrupt state than not to ha.e it at all, while hoping still to be able to correct it from another copy which might happen to emerge. /are you well, if you do not curse him,.Here is the fundamental problem of the textual tradition of Catullus. he whole of our manuscript tradition, outside the fortuitous Carolingian transmission of poem 43, is descended from a late and corrupt copy which was already the despair of its earliest scribes. his is radically different from the textual tradition of Catullus, contemporary Eucretius, preser.ed in two excellent ninth;century manuscripts, fi.e hundred years earlier than the extant manuscripts of Catullus, or from that of ?ergil, with its magnificent collection of manuscripts from the fifth and sixth centuries, eight hundred years earlier ). he -ey to reco.ering what Catullus wrote lies not in disco.ering more about his manuscripts, but in attempting to emend and elucidate the particularly corrupt manuscript tradition which has come down to us.his was ob.ious .ery soon after the redisco.ery of Catullus in the fourteenth century, an age of considerable scholarly acti.ity on Eatin texts in Italy. In about "%#"the lost *0 =, already mentioned as the common ancestor of + and 6, was copied forthe great humanist Coluccio 0alutati, the chancellor of /lorence. his copy is the manuscript we -now as 6. Coluccio himself added a number of important marginal readings, -nown to modern editors as 63< some of these plainly deri.e from =, since they are also present in +, while others seem to be Coluccio,s own conjectures, though it is difficult to be certain of this 4. &mendation of Catullus became one of the sports of &urope once the text reached printed form in "B$3, in ?enice, at the hand of the printer Fendelin .on 0peyer, by which time there were clearly many manuscripts in circulation. In its first century after printing, the text of Catullus, now e.en more widely a.ailable, was wor-ed on by humanists from 5olitianus to 0caliger and dramatically impro.ed< the apparatus criticus of any modern edition bears eloquent witness to the acti.ities of these fifteenth and sixteenth;century scholars. $ 5erhaps most interesting is the grouping of items now thought to be separate poems in different bloc-s with collecti.e titles C it is clear from modern in.estigations ' that in the editions of the fifteenth and sixteenth centuries the text gradually came closer and 5 (n these two traditions see most con.eniently 6eynolds ("#'%b) and ("#'%c).6 /or a list of these readings and their li-ely origins see homson ("##$) %';B%.7 /or an excellent account see Haig +aisser ("##%).8 0ee the useful analyses ofdi.isions in early editions by Haig +aisser ("##%) 8collected in the index, p.BB@, s... dispositio carminum9. %closer to obser.ing the di.isions of poems generally mar-ed in modern editions, especially in the ")$$ edition of 0caliger # . he issue ofwhether the modern di.isions of poems are in e.ery case correct, and the related issue of whether the collection of Catullus transmitted to us is a single ancient boo-, are not questions I will debate here< but they are of course issues on which any edition of Catullus needs to ta-e a position "@.2 : The Modern Text of Catullus : from Haupt to Thomson 11Fhen *orit7 Haupt, one of the best Eatinists of his time, edited Catullus for the first time in "')% "3, he produced a text which -new only one of the three *00 on which modern texts are based (+) "%. Its .alue today is for se.eral important emendations, not for its account of the history of the text. Fithin fifty years, howe.er,the true relationship of (+6 emerged and the basis of the modern text of Catullus was born. + was used properly for the first time by Eudwig 0chwabe in his text of "'44, while ( was presented first by 6obinson &llis in his first edition of "'4$ "B. !oth these texts still relied too hea.ily on other minor *00. In "'$4, &mil !aehrens brough out the first .ersion of his edition, which basically constituted the text from + and ( alone (and also contained a number of emendations by this brilliant and sometimes wayward critic) "). 1ll that was needed now was 6< this manuscript, lost through a cataloguing error, was dramatically redisco.ered in a dusty corner of the ?atican Eibrary by the 1merican scholar F.+.Hale in "'#4 "4 and utilised by &llis in his (xford Classical ext of "#@B, the first edition in which (, + and 6 are deployed together for the constitution of the text "$. It is interesting to note that 6 did not gain 9 (n this important edition see +rafton ("#$)), Haig +aisser ("##%) "$';#3. 10 (n the issue of di.iding poems see recently Heyworth ("##))< on the issue of whether the extant ordering of the Catullan collection is authorial (I thin- not) see the copious literature gathered by Harrauer ("#$#) "3@;" and Holo-a ("#')) 3B#;)", and the recent treatment by 0cherf ("##4).11 In this section I am once again indebted to homson ("##$) )$;4@.12 Haupt ("')%). /or an e.aluation of Haupt by an important &nglish scholar of the next generation, cf. :ettleship ("'')) ";3313 + itself had been redisco.ered only by 0illig in "'%@ ; cf. homson ("##$) )$.14 0chwabe ("'44), &llis ("'4$).15 !aehrens ("'$4), with a commentary< for a modern account of !aehrens cf. 0hac-leton !ailey ("##@). 16 /or the details see Hale ("#@').17 &llis ("#@B)< &llis does not use 6 as fully as he might owing to his desire not to Bimmediate authority, perhaps because Hale himself did not pro.ide a full collation, which was first generally a.ailable in homson,s facsimile, published in "#$@ "'. Filhelm Aroll, in the preface to his classic commentary on Catullus of "#3% which is still deser.edly in print, mentions + and ( as witnesses to the ?eronese tradition of the text but not 6 "#, and Aroll,s edition is indeed not noted for its contribution to textual criticism.1fter &llis, we may leap half a century into the modern era of editions of the text. In "#)' &isenhut produced his first eubner text, a re.ision of that of 0chuster of"#B#< in the same year *ynors, (xford Classical ext was published. &isenhut,s first edition was superseded by his second of "#'%, which I will discuss below 3@. *ynors, edition, partly because of its wide a.ailability, has become the standard text, at least inthe &nglish;spea-ing world. It is clear about the relationship of the main manuscripts (, + and 6, as set out abo.e, calling their agreement ?, and gi.es clear information inthe apparatus criticus about the secondary manuscript tradition< it is helpful in ma-ing a clear distinction between the four major manuscripts ((,+,6 and ), cited by capital letters in the apparatus criticus, and the minor sources, which are grouped under minuscule +ree- letters. his is a clear statement of the relati.e importance of the two groups. It has two main problems C its general conser.atism and insufficient willingness to admit conjectural solutions to e.ident corruptions, and its consequent omission of many important conjectures from the text and apparatus criticus. *ynorswas a brilliant palaeographer and a first;class dater and sorter of manuscripts, but, as we shall see in some of the textual examples I will discuss in the second half of my paper, he could be deficient in the diagnosis of textual problems, and was sometimes content to lea.e dubious readings in the text.!ardon,s 0tuttgart eubner edition of "#$%, which followed his edition of "#$@in the Collection Eatomus series 3", is in many ways an eccentric text. Its main suggestion on textual transmission is that some readings deri.e from a source other than the common source of (,+ and 6, generally called ?. !ardon produces two typesof argument for this proposition C the assertions of humanists that they had found a reading in an old manuscript, and the indirect tradition, where quotations of Catullus steal Hale,s thunder ; cf. &llis ("#@B) ix.18 homson ("#$@)19 Aroll ("#33) III;I?.20 /or these texts see n." abo.e, adding &isenhut ("#)').21 !ardon ("#$@) and ("#$%).)in later ancient authors di.erge from the text generally transmitted in the manuscripts ofCatullus. 1s &.G.Aenney pointed out in a forceful re.iew of the "#$@ .ersion of this text 33, these are both .ery sha-y arguments C humanists are notoriously unreliablein reporting supposedly old manuscript readings which are in fact recent conjectures3%, while the indirect tradition is of course .ery unreliable and much more li-ely to ma-e mista-es than a copyist, gi.en that ancient quotation wor-ed largely from memory. !ardon,s -ey example here is a passage of 1puleius (!pologia 4) which citesa line ("#) from the famous poem on &gnatius, dubious habits of dental hygiene, poem%#. his line contains the phrase russam defricare gingivam, ,rub clean your red gums,< 1puleius quotes the phrase as russam pumicare gingivam, substituting a different but metrically equi.alent .erb of .ery similar meaning, ,scrub your red gumswith pumice,. his is surely a memory error, or e.en a deliberate change by 1puleius, as ?incent Hunin-, the most recent commentator on the 1puleius passage, suggests 3B.1 memory error is made more li-ely by the fact that Catullus uses the .erb defricare in the other poem he de.otes to &gnatius, %$ (%$.3@ dens "ibera defricatus urina), suggesting that the same .erb should be read in %# too. !ardon, howe.er, thin-s that 1puleius preser.es the genuine text of Catullus and puts pumicare in his text. his seems to me per.erse and to o.erpri.ilege the indirect tradition< !ardon,s e.idence foran ancient tradition fundamentally different from ? amounts to nothing. His text does pro.ide more conjectures in the text and apparatus than that of *ynors, and is generally less conser.ati.e, though many of the new readings he admits are dubious (Irefer to Aenney,s re.iew for examples).In "#$' homson published his important first edition of Catullus< this is now superseded by his "##$ edition, which I will discuss below 3). 5erhaps the single most important contribution to the textual criticism of Catullus in the "#$@,s other than homson,s edition was 6.+.*.:isbet,s article ,:otes on the text and interpretation of Catullus, 34. In this article :isbet, beginning from *ynors, text, suggested his own conjectural solutions to more than twenty problematic passages of Catullus and also re.i.ed a number of older conjectures which editors had ignored. *any of his suggestions were indeed ta-en up by later editors, but what is important is his 22 Aenney ("#$3)23 Cf. e.g. 5asquali ("#)3) 4';$" and '4;$.24 Hunin- ("##$) II.%".25 cf. n." abo.e.26 :isbet ("#$')4approach C he gi.es proper weight to the highly corrupt nature of the transmitted text of Catullus, and is not afraid to try conjectural solutions, both his own and those of other scholars. his in my .iew is the right approach< where a text is poorly transmitted, extensi.e conjecture is legitimate. 0imilar in approach, and perhaps the best text of Catullus to date, is that of +eorge +oold, published in "#'% 3$. +oold is well -nown as a searching and often radical critic of Eatin texts, and his talent is well exercised on Catullus. His text incorporates a number of important conjectures of his own, and also pic-s up on the conjectures of other recent scholars, especially those made in :isbet,s article, to produce a text which is perhaps closest than any other modern edition to what Catullus actually wrote. he only (mild) complaint I ha.e about +oold,s text is his practice of putting his own Eatin .erses in the text whene.er the manuscripts ha.e a lacuna C this of course useful in that it shows what the editor thought might ha.e stoodin the lacuna, but it is a little distracting for the reader. It should also be said that his text does not pro.ide a full apparatus criticus, only a section of critical notes indicating departure from the usual text. Fhat we need is a text li-e +oold,s with a full and effecti.e apparatus.his need is not met by either of the other two main modern editions. &isenhut,s eubner edition of "#'% has a number of .irtues C it has a sound grasp of the textual tradition,s essential dependence on (, + and 6, though it does not -now *cAie,s dissertation, and cites a reasonably wide range of modern conjectures, though it does not -now the wor- of :isbet. It does -now the useful text;critical wor-of the /innish scholar (-sala 3', not -nown to +oold. *y main criticisms of &isenhut are two C first, his apparatus criticus, unli-e that of *ynors, does not differentiate sufficiently between the major and the minor manuscripts, so that 2 and *, the 2atanus and *arcianus, which deri.e from + and 6 respecti.ely, are gi.en apparent equi.alence with (,+ and 6 by being cited quite often and in capital letters. he system of *ynors, in which only the major manuscripts are capitalised in the apparatus and the agreement of their descendants is usually not noted, is much better. 0econdly, &isenhut,s text shares the major problem of most texts apart from that of 27 cf. n." abo.e.28 (-sala ("#4)), more useful for its diagnosis of problems and re.i.al of older conjectures than for the author,s own conjectural solutions.$+oold C it is too conser.ati.e, content to lea.e clearly unsatisfactory readings in the text, often without much indication in the apparatus criticus. he most recent and the most extensi.e modern edition, the second edition by2./.0.homson, containing a Eatin text with accompanying commentary, is homson ("##$). here is no doubt that homson has studied the manuscript tradition in greaterdetail that any pre.ious editor, and his long catalogue of manuscripts and his elaboratestemma is certainly the fullest a.ailable. His interest in the more intricate details of the manuscripts perhaps leads to some of the problems in his edition. (ne of these problems is the excessi.e detail of his apparatus criticus C he often pro.ides too muchinformation about minor orthographical .ariants in the major manuscripts when the reading is perfectly clear and uni.ersally agreed 3#. Con.ersely, li-e most modern editors of Catullus, he is often content to lea.e se.ere difficulties in the text without attempting to emend them, in contrast to :isbet and +oold, and though he lists a fair number of conjectures in the apparatus, the number could be increased. Fhen he does.enture a conjecture of his own, it is generally unpersuasi.e C for example, at "3.';# est enim leporum # differtus puer ac facetiarum, differtus is a uni.ersally;accepted conjecture %@ for the transmitted dis$s%ertus of (+6 C the error is a simple one. homson, wishing to retain something closer to the reading of (+6, conjectures est enim leporum # diserte pater ac facetiarum, ,for he is explicitly the father of charm and wit,C in this pater is fine, gi.en that another friend of Catullus is called pater esuritionum, ,father of famines,, at 3".", but the use of the ad.erb diserte with pater seems dubious in both sense and construction. homson,s objection to the uni.ersally accepted differtus is that one would expect an ablati.e rather than a geniti.e after this participle< this is true, but as /ordyce argues in his commentary, differtus is here simply ta-ing on the construction of plenus by a natural analogy %", and differtus with the geniti.e seems better Eatin than the ad.erb diserte as homson wants to use it.In summary, then, none of the current editions of Catullus is entirely satisfactory in either text or apparatus criticus.How would an ideal text of Catullus then loo- in my .iew H /irst, it would ha.e an apparatus criticus which was free of minor orthographical .ariants< I would also probably cite the three main manuscripts 29 0ee for example his apparatus criticus on "%.4;"% or 4B.$$;#3.30 5re.iously ascribed to 5asserat, but now traced bac- to +io.anni di 5astrengo (5astreng-ius), friend of 5etrarch (on whom see homson ("##$) 3)).31 /ordyce ("#4") "%".'(+6 singly rather than using the sigla ? or 1 for the ?eronese tradition common to them all, or the siglum = for the agreement of + and 6< the manuscripts .ary sufficiently at crucial points to ma-e this worthwhile throughout. 0econd, both text and apparatus would contain more conjectural solutions to textual problems than in other editions< in my .iew, the textual tradition of Catullus is so poor that e.en after fi.e centuries in print there are still many unsol.ed problems, and many good conjectures to be made. 3 : Corruption and Conjectural Solution he final section of my paper will discuss a few problematic passages of Catullus by way of methodological example, suggesting some new conjectural approaches to sol.ing long;recognised difficulties, re.i.ing conjectures which ha.e been neglected, and pointing out some hitherto unnoticed problems %3. In Catullus, as in other ,canonical, ancient authors whose wor-s ha.e been studied for centuries, we are often too hesitant to posit corruption rather than accepting anomalies which wouldnot be tolerated in a less familiar text< indeed, as already argued, there is especially little reason to do this in the case of Catullus, since his text is so poorly preser.ed in comparison with those of many other classical writers. /or the same reason, we should also be much more ready to consider and re.i.e earlier conjectures which ha.ebeen too lightly rejected by pre.ious editors.%3 he material that follows is selected from my own contributions to Harrison and Heyworth ("##')< there the material was presented in a brief format, and here I ta-e the opportunity of more space to pro.ide fuller linguistic and stylistic argumentation and methodological justification. #37.11-14puella nam mi, quae meo sinu fugit,amata tantum quantum amabitur nulla,pro qua mihi sunt magna bella pugnata,consedit istic.In line "" puella nam me ((+6) ma-es no sense< me is clearly wrong, ha.ing no grammatical function in the sentence. mea is printed in some early editions, but Catullus ne.er in his choliambics allows a resolution for a syllable which is properly short< 0chwabe,s mei would require an unparalleled syni7esis of this geniti.e pronoun,which in any case hardly seems natural here.mi (Heinsius, anticipated by an anonymous 6enaissance critic %%) at least scans, but is highly unsatisfactory, though it is printed by most modern editors, e.en +oold< *ynors, who also prints it, rightly comments uix latine in his apparatus (a peculiar remar- on a conjecture which the editor has chosen to put into his text).*odern scholars tend to defend the conjecture mi as a dati.e of agent with the participle amata %B, but this seems difficult. hough such dati.es of agent can be separated from their passi.e participles in Catullus, as in $$." &ufe mihi frustra ac nequiquam credite amice, such separation by an inter.ening relati.e clause which lea.es mi so mar-edly hanging is unparalleled and stylistically undesirable, as *ynors, comment owns. homson in his recent edition interprets mi as equi.alent to possessi.e meus, comparing 3".""< but there the *00 read the e.idently corrupt me me puer et sitire discet, and homson,s own reading a te mi puer et sitire discet incorporates parts of two pre.ious conjectures, a te from *unro,s a te mei, and mi from ?enator,s meus mi, a fragile base for a linguistic parallel %). he main problem with homson,s interpretation is that puella ' mi simply cannot mean ,my girl, as a self;contained phrase, as he requires< mihi#mi as an ethic dati.e or dati.e of ad.antagecan indeed be used quasi;possessi.ely in colloquial contexts, but it is always closely juxtaposed with the action of a main .erb as well as with the noun to which it relatesC 33 he reading is found in the second hand of 5arisinus lat.$#'#, written in "B3%, as first noted by homson ("#$')< on Heinsius see n.%), below.34 0o (-sala ("#4)) B# ((-sala himself suggests enim mi, no better).35 /urthermore, the case made in his commentary on 3"."" relies on reading mi at %$."" C ,the colloquial use of mi for meus, etc., may be supported by the now uni.ersally accepted Humanistic emendation mi at %$."", (homson ("##$) 3)')."@compare e.g. 5lautus (acch. "B' iam excessit mi aetas ex magisterio tuo, )as.4") nunc tu mi amicus es in germanum modum, *seud."@)3 nunc demum mi animus in tuto locost. he case in Catullus is plainly different.Here, I would argue, is an instance where a relati.ely early conjecture attached to the name of a respected scholar %4, apparently sol.ing the problem of the corrupt transmitted text, has been too easily accepted by modern editions. 1n e.en earlier conjecture, 1.antiusI namque for nam me, recorded by most editors, is for me more attracti.e C +que would be lost by haplography before qu$a%e and then the gap filled with a con.enient monosyllable. he only objection would be the presence of namquetwo lines before, in line #. If this is felt to pre.ent namque in "", I suggest nam illa fornam mi, easily pic-ed up by quae (for illa quae of the lost Eesbia cf. )'.3)< for the elision of nam compare "%."" and 3".$.38.6-8irascor tibi. sic meos amores ,paulum quid lubet allocutionis,maestius lacrimis SimonideisIn line 4, I would follow +oold in reading !aehrens, sic tuos amores. Here, as often, editors ha.e paid insufficient attention to Catullus, linguistic usage. In the diction of the colloquial poems of Catullus, the plural amores when used with a possessi.e pronoun always has its 5lautine concrete sense of ,the object of lo.e, (cf. 4."4, "@.", ").", 3".B, B@.$, B)." and T-- "."#$@.#ff )< the transmitted sic meos amores would naturally mean ,is that how you treat my darling,, in a context where a reference to Catullus, belo.ed seems wholly out of place< the reference really ought tobe to Catullus himself, who is complaining of being neglected by Cornificius, the poem,s addressee. /ordyce suggests a possible exclamatory sense (,that you, my darling, should beha.e li-e thisJ,), but in such an ellipse the accusati.e is more naturally the object of the omitted .erb than its subject. Fith tuos the phrase is more 36 his is of course :icolaus Heinsius ("43@;'"), whose textual notes on Catullus were published posthumously in Heinsius ("$BB)< for his high reputation as an emender of Eatin texts cf. e.g. 0andys ("#@') %3B;4.""playful than indignant (,is that how you treat your darlingH,), using the same witty and quasi;erotic tone to a male poetic friend as in poem )@ to Cal.us.Eines $;', a (comic) request to a fellow;poet for consolatory .erse in the manner of 0imonides, again contain no main .erb. /ordyce holds that this is satisfactory, citing two parallels from Cicero,s letters, which both understandimperati.es C Cicero !tt."@."4.4 et litterarum aliquid interea 8sc. ,send,9, .am."4.3B.%3 de publicis omnia mihi certa 8sc. ,write,9. In fa.our of his .iew is the fact that the context in Catullus is similarly colloquialKepistolary, and that line 4 also omits a main .erb, the two constructions supporting each other. Howe.er, there is a problem. In contrast with the Ciceronian passages, it is not entirely clear what we are to understand as the omitted main .erb here C paulum quid lubet is clearly the object of a request from the poet, but that request might be expressed either in a first person .erb with the poet as subject, or (as in Cicero) in an imperati.e aimed at his addressee. here is also the fact that the phrasing of paulum quid lubet allocutionis is unparalleled in classical Eatin %$< /ordyce compares paulum nescioquid (Cicero &osc.!m."")) and paulum aliquid (/e 0r.".#)), but these phrases are both used absolutely, not with a following defining geniti.e, and quidlibet ne.er has an adjecti.eattached to it elsewhere in classical Eatin %'.!oth these factors suggest to me that the passage is corrupt, and that editors need to supply a .erb of request or command here. /or lubet Heinsius (again) suggested iuvet, ,would please (me),, which only &isenhut mentions among modern editors. his is not unattracti.e in itself, but the resultingpaulum quid is again without 5HI parallel, and I feel the need for a stronger .erb, replacing not lubet butpaulum ; e.g. praebe, Lgi.e meI, or posco, LI demandI. 55.6-14in 1agni simul ambulationefemellas omnes, amice, prendi,quas uultu uidi tamen sereno.2auelte2, sic ipse flagitabam,)amerium mihi pessimae puellae.quaedam inquit, nudum reduc37 Information from 5HI C2;6(*.38 I.e. in the B" instances to be found on the 5HI C2;6(*."33en hic in roseis latet papillis.4sed te iam ferre "erculei labos est 5tanto te in fastu negas, amice.his poem contains some of the most acute textual problems in the whole of Catullus, which scholars ha.e not yet sol.ed< as often, textual difficulties are grouped together in particular parts of texts which ha.e somehow been especially liable to corruption inthe copying process. I will not here discuss its metrical difficulties or its relationship to poem )'b %#, but merely loo- at a few e.ident corruptions. he meaningless auelte (#) is clearly corrupt, perhaps influenced by uultu in the pre.ious line.Fhat should stand hereH*ost editors belie.e that this corrupt word should be the beginning of CatullusI speech to the femellae, rebu-ing them for hiding Camerius, and pro.iding a .erb to go.ern )amerium. I agree with this approach.0uggestions ha.e been made asfollows, not an exhausti.e list B@ (some remo.ing sic and ipse as well as auelte)Cauertistis saepe 1.6iese, N6bb. #"("'4)) 3#'C Lyou ha.e stolen him away oftenIauelli sinite 1.antius C Lallow him to be torn from youIreddatis G./oster, )Q 3" ("#$"), '4;$ Lgi.e him bac-Ia cette (from cedo) huc F.1.Camps, !6* #B ("#$%), "%";3 LheyJ gi.e him hereIaudite en homson CLhey, listen to meI (with ellipse of .erb go.erning )amerium)aufertis+ooldC Lare you ta-ing him awayHI+ooldIs is the most attracti.e of these suggestions, but the present tense seems odd (one expects a perfect, gi.en that the implied action has plainly happened already) andthe question is perhaps too polite gi.en the roughness of prendi (the same could be said for /osterIs subjuncti.e reddatis). hough I would reject homsonIs assumption that the .erb go.erning )amerium is omitted and hence his conjecture en audite, thereare some attractions in en, often found with imperati.es (cf. (.id, 1et."%.34B aspicite en ?ergil Ecl.4.4# en accipe, 7eorg.%.B3 en age, 0ilius B.3'" disce en nunc, T-- ).)B$.34ff.), and pic-ed up with the en in the reply at "3. I suggest efferte en, Lhey, bring me out Camerius (from hiding)I. 39 /or discussion of these issues see Harrison and Heyworth ("##') #';"@@, and for a unified .ersion of )) M )'! see the edition of +oold ("#'%). 40 /or further lists of conjectures see +iri ("'#B) I."#% and (-sala ("#4)) 4@."%1nother ob.ious problem are the three syllables clearly lost at the end of line "". /or the missing last word, &llis plausibly restores pectus, gi.en nudum and the sense of line "3< but his reducta, the most ob.ious participle to complete the space, does not ha.e the required sense (see /ordyce)< /oster,s interpretation of &llis, nudum reducta pectus as ,with her bare breasts drawn apart, ()Q n.s. 3" ("#$") '4;$) seems forced, and is offered without a parallel.nudum sinum recludens (1.antius) pro.ides a .erb easily corrupted to reduc+ and assumes that sinum has dropped out before it< the difficulty then is nudum, which ma-es no sense with sinum, Lbosom of garmentI, and the sense of recludere is not quite appropriate (it means Lopen upI in the literal sense ; see /ordyce here).+ooldIs rewriting, 3en4 inquit quaedam, sinum reducens, a.oids this and another difficulty, inquit occurring before speech (see abo.e), but is too radical. /riedrichIs nudum reclude pectus, beginning the girlIs speech, is attracti.eexcept for reclude (see abo.e).I suggest nudum retecta pectus, Lwith her bare breast unco.eredI, i.e. already showing her charms for sale (or perhaps supposedly fresh from entertaining Camerius)< if nudum retecta is thought too pleonastic, try niueum retecta, pro.iding a neat colour contrast between niueum and roseis./or retecta cf. (.id, 1et."%.B)' (5olyxena) iugulumque simul pectusque retexit , for the type of phrase Catullus 4B.4B contecta leui uelatum pectus amictu at 4B.4B, and for niueum cf. "$ lacteolae, again of the white s-in of the girls.In "% the transmitted iam ferre has not been questioned, but perhaps editors should loo- harder. .erre can be doubted C though the fact that Catullus cannot endureCamerius, beha.iour might be a reasonable conclusion from the poem, CatullusI main problem is not enduring Camerius, self;concealment but finding him. he temporal iam also seems wea- here (what does it signify H). I suggest intercipere, Lcut offI, LcatchI, an appropriately military metaphor for a Eabour of Hercules (cf. T-- $.".3"4B.%Bff). 57*ulchre conuenit improbis cinaedis,1amurrae pathicoque )aesarique.nec mirum8 maculae pares utrisque,urbana altera et illa .ormiana,impressae resident nec eluentur8 )"Bmorbosi pariter, gemelli utrique,uno in lecticulo erudituli ambo,non hic quam ille magis uorax adulter,riuales socii puellularum.pulchre conuenit improbis cinaedis. "@his attac- on Caesar and *amurra consists in lines %;# of a list in the nominati.e case of insults applicable to the pair, nominati.es apart from maculae pares utrisque in line %.his phrase, though transmitted without comment by editors, seems strange in any case C what precisely are the maculae deri.ing from their patriae of 6ome and /ormiae (cf. line B) which affect the pairH1 pro.incial origin might be unfortunate, but not birth in the 9rbs itself.0urely the two should themsel.es constitute irremo.able stains on their respecti.e patriae, rather than ha.ing mysterious stains from their places of birth. !oth these considerations suggest the conjecture utraeque for utrisque B"C for macula used directly of persons in a similar and closely contemporary in.ecti.e context cf. Cicero, *rou.)ons."% (describing 5iso and +abinius) has duplicis pestes sociorum, militum cladis, publicanorum ruinas, provinciarum uastitates, imperi maculas. he point is that the reputations of 6ome and /ormiae will ne.er reco.er from these two blots on their respecti.e landscapes C ,the two match each other as stains, one on the City, one on /ormiae, remaining deep;stamped and ne.er to be washed out,.63.52-4 ad :dae tetuli nemora pedem,ut aput niuem et ferarum gelida stabula forem,et earum omnia adirem furibunda latibula '5oem 4% still has a number of problems, partly because it has not been as much studied by scholars as some other poems, owing perhaps to its colourful subject matter and difficult metre. Here earum omniain line 4B is the main problem. :isbetpoints out that it constitutes a metrical oddity B3, since this is the only .erse in the 41 he corruption to the easy possessi.e dati.e utrisque may also ha.e been influenced by false assimilation to the neighbouring paris, transmitted in (+6 for pares.42 :isbet ("#$') "@@.")poem where the initial ionics do not undergo anaclasis. He further argues that omnia is pointless and that , earum is a.oided in .erse (except in Eucretius and HoraceIs satires), and here seems too emphaticI.He goes on to conjecture harundinosa (which Catullus uses also at %4."%), before expressing concern at the lac- of a geniti.e to define latibula.If a geniti.e is felt to be needed, we might combine aprorum (Heyworth,s suggestion B%) with Eucian *NllerIs opaca for omnia BB< boars are appropriate to this landscape ; cf. $3 ubi aper nemorivagus. 63.63-4ego mulier, ego adolescens, ego ephebus, ego puer,ego g;mnasi fui flos, ego eram decus olei.4% puber 0caligerC iuuenis 0chwabe4B g;mnasi C gimnasti (+6 fui (C sui +6homson has followed +oold in rejecting the transmitted mulier, which is an ob.ious interpolation by a reader who could not wait for the contrasting account of the presentin 4'C ego nunc deum ministra etc.< in lines 43;4$ 1ttis is loo-ing bac- at his male past before going on to consider his female present at 4'. homson rightly says that ,itis better to emend, in the interests of consistency in 1ttis, list, B). !ut 0caligerIs suggestion puber, fa.oured with a place in the text by both homson and +oold, is morphologically and lexicographically dubious. his nominati.e form of the adjecti.ewhose geniti.e is puberis is not otherwise found in classical Eatin, which uses the nominati.e pubes (Cicero &ab.*erd.%", :epos /i.B.B), as in.ariably with its opposite impubes, +beris< puber is a post;classical form, with its earliest occurrence in 0er.ius (on !eneid ).)B4), just as the parallel form impuber is -nown only from late glossaries (T-- $.".$@).$@). /urthermore, adjecti.al pubes is strongly prosaic and -nown to poetry first in ?ergil, where it refers only to .egetation (!eneid "3.B"%). /orthese reasons I would incline to 0chwabeIs iuuenis, unnoticed by editors of Catullus 43 Harrison and Heyworth ("##') "@B.44 In his edition, *ller ("'$@).45 homson ("##$) %'%."4this century B4, gi.ing four of the ages of man in re.erse order. (nce again the .alue ofnot neglecting older conjectures is clear.In 4B the perfect fui and the imperfect eram cannot stand together, as :isbet rightly argues B$< homson thin-s the two can coexist, but this combination of tenses referring to the same time is surely intolerable B'. Conjecture is needed. (f :isbet,s own two conjectures, suus (for fui) upsets the balance of the two hal.es by gi.ing an adjecti.e to flos but not to decus< I prefer prius, which would match eram, with .erb and ad.erb shared out between the clauses, and loo- ahead to nunc in 4#.1n alternati.e would be to read Heyworth,s ego eram g;minasii flos B#C the euphonic i is undesirably lost at least twice elsewhere in the transmission of this poem, once in this same word (cf. 4@ and $").84.5-8credo, sic mater, sic liber auunculus eius,sic maternus auus dixerat atque auia.hoc misso in S;riam requierant omnibus aures8audibant eadem haec leniter et leuiter.) eius C eius est (+6:isbet argues forcefully in his article )@ that liber cannot stand in ) C the poem is not atall interested in the status or name (Eiber) of 1rrius, uncle, who is so insignifican as tobe necessarily anonymous. :isbetsuggests the ad.erb semper.!ut we should note that there is another word in the line that deser.es suspicionC eius.It is transmitted only here in CatullusIs manuscripts< and the form is rare enough in poetry for doubt toarise )".In this case, moreo.er, the pronoun is redundant. /or the reasons :isbet has gi.en in his attac- on liber an ad.erb is the most attracti.e substitute, and Heyworth has suggested olim (Lin the pastI).46 0chwabe ("'44)< the last to mention it was 5almer ("'#4) xl.47 :isbet ("#$') "@@.48 homson,s parallel, !eneid "3."B$ qua visa est fortuna pati *arcaeque sinebant (homson ("##$) %'%) is .ulnerable to :isbet,s objection ("@@)C ,it is no use quoting instances of tense;.ariation where the .erb also changes,.49 Cf. Harrison and Heyworth ("##') "@).50 :isbet ("#$') ""@.51 Cf. the statistics cited by 1xelson ("#B)) $3."$!ut if olim were right here, semper can hardly be the word we are loo-ing for earlier.In its place, one might consider Fic-Is libere, re.i.ed by (-sala )3 and &isenhut, an ad.erb ha.ing something of the force of quantum poterat (B)./or libere of untrammelled spea-ing, cf. T-- $.3."3'#.33ff (Catullus uses the ad.erb in a different sense at 4%.'@)< it indicates freedom from normal constraints, here those of polite pronunciation (cf. T-- loc.cit. Lsignificatur fere oratio audaciorI). :isbet objects to the elision of libere avunculus, but such long final .owels of ad.erbs are similarly elided in Catullus by naturally short syllables at B).% perdite amo and "@B.% perdite amarem.! : ConclusionI hope that in what I ha.e said I ha.e made a case for a need for a new text of Catullus. His text remains extensi.ely corrupt despite the attentions of generations of scholars, and needs to be questioned and scrutinised more intensely than that of better;preser.ed Eatin poets< consequently, conjectural solutions to its percei.ed problems can still be newly made and re.i.ed from pre.ious scholarship in the editions of the twenty;first century. he existing editions of Catullus, with the noble exception of that of +oold, which does not pro.ide a full scholarly apparatus, are too conser.ati.e in the inclusion and citation of conjecture. 1 repertory of conjectures would itself be a .ery useful tool, but it would be better to ha.e all the good conjectures in the apparatus criticus of a widely;a.ailable modern text. I hope that someone will soon produce such a wor-.52 (-sala ("#4)) #B;), not realising that Fic- had anticipated him. I ha.e been unableto find the original place of publication of Fic-,s conjecture, but it is clearly pre;"#4); cf. Oicri ("#$4) 34B."'Bibliora!h"1xelson, !. ("#B)), 9npoetische ??+=@??. +roningen.Holo-a, G.5. ("#')), 7aius alerius )atullus 8 ! S;stematic (ibliograph;. :ew Dor-.Hunin-, ?. ("##$), !puleius 8 *ro Se /e 1agia 83 .ols.9. 1msterdam.Aenney, &.G. ("#$3), re.iew of !ardon ("#$@), )& n.s. 33 C 3"3;"%.Aroll, F. ("#3%), )atull. Eeip7ig.*cAie, 2.0. ("#$$), The 1anuscripts of )atullus 8 &ecension in a )losed Tradition. 2iss., Cambridge.*ynors, 6.1.!. ("#)'), )atulli )armina. (xford 8(C9*ller, E. ("'$@), )atulli Tibulli *ropertii )armina. Eeip7ig.:ettleship, H. ("'')), -ectures and Essa;s. (xford.:isbet, 6.+.*. ("#$'), ,:otes on the text and interpretation of Catullus,, *)*hS 3B C #3;""),reprintedin :isbet ("##)), $4;"@@.:isbet, 6.+.*. ("##)), )ollected *apers on -atin -iterature. (xford.(-sala, 5. ("#4)), !dnotationes )riticae ad )atulli )armina. Helsin-i.5almer, 1. ("'#4), )atulli )armina. Eondon.5asquali, +. ("#)3), Storia della tradiAione e critica del testo. /lorence.5ecere, (. and 6ee.e, *.2., ("##)), eds., .ormative Stages of )lassical Traditions. 0poleto.6eynolds, E.2. ("#'%a), ed., Texts and Transmission. (xford6eynolds, E.2. ("#'%b), ,Eucretius, in 6eynolds ("#'%a) 3"';33.6eynolds, E.2. ("#'%c), ,?irgil, in 6eynolds ("#'%a) B%%;4.0andys, G.&. ("#@'), ! "istor; of )lassical Scholarship 8 olume ::. Cambridge.0caliger, G.G. (")$$), )atulli *roperti Tibulli nova editio. 5aris.0cherf, G. ("##4), 9ntersuchungen Aur antiBen erCffentlichung der )atullgedichte 80pudasmata 4"9. Hildesheim.0chwabe, E. ("'44), ).alerii )atulli -iber. +iessen.0chuster, *. ("#B#), )atulli )armina. Eeip7ig 8!9.0hac-leton !ailey, 2.0. ("##@), ,&mil !hrens, in Hofmann ("##@), 3);%$.arrant, 6.G. ("#'%), ,Catullus,, in 6eynolds ("#'%a), B%;).homson, 2./.0. ("#$@), ,he Codex 6omanus of Catullus C 1 Collation of the ext,, &h1 ""%C #$;""@homson, 2./.0. ("#$'), )atullus 8 ! )ritical Edition. oronto.homson, 2./.0. ("##$), )atullus 8 Edited Dith a Textual and :nterpretative )ommentar; 8*hoenix 0uppl.%B9. oronto.Oicri, *. ("#$4), Scritti )atulliani. >rbino."#