A note on Catullus 63.5.
deuoluit iletas acuto sibi pondere silices. The gist of this (`he rolled down from (?) himself flints with sharp mass') is improbable, and in particular the form iletas is a non-existent word and two syllables too long for the metre.
The sense requires that v.5 tell of Attis' self-castration; the most conservative attempt to extract this from the paradosis is:
deuoluit ile acuto sibi pondere silicis.(2) This involves: (i) taking ile to mean `the genitals', and (ii) taking acuto ... pondere silicis to mean with a sharp mass of flint'. However, ile in the singular is without a convincing parallel(3) and the signification `genitals' (rather than `groin') is also unparalleled.
Avantius changed (i) silices to silice, ablative in agreement with acuto, `with sharp flint'; and (ii) pondere to pondera. Bergk proposed (iii) ili (or its orthographical variant ilei) for the corrupt iletas, this genitive is then made to depend on pondera, `the burdens of his groins', i.e. his genitals. Haupt further proposed (iv) deuolsit for deuoluit, giving a more graphic verb (`tore down') to suit the activity. The line so reconstructed reads:
deuolsit il(e)i acuto sibi pondera silice. Despite being almost universally adopted by editors, the supplement ili is subject to some powerful objections. Not only is the word for groin normally plural, as we have seen, it is also third declension. The genitive is elsewhere always ilium, the dative and ablative almost invariably(4) ilibus. Yet the interpretation in question requires a form which is properly that of the second declension singular.
There are several parallels for the expression ilium pondera (or an equivalent),(5) most notably Ovid, Fasti 4.241 onus inguinis, also describing the castration of Attis, and closely resembling the Catullan passage in its wording.(6) But there are also examples of pondus (and its synonyms) being used of the genitals without any dependent genitive: Lucilius 534-6 ibat forte aries, inquit, iam quod genus! quantis testibus! uix uno filo hosce haerere putares \ pellicula extrema exaptum pendere onus ingens, Martial 7.35.3-4 sed meus, ut de me taceam, Laecania, seruus \ Iudaeum nulla sub cute pondus habet. These parallels show that the genitive with pondera at Catullus 63.5 is not indispensible: the fact that Ovid chose to make the expression explicit with a genitive need not imply that Catullus did so. In Lucilius and in Martial, as at Petronius 92.9 habebat enim inguinum pondus tam grande, ut ipsum hominem laciniam fascini crederes, the metaphor of the genitals as `weight' serves to convey the impressive size of the membra. At Catullus 63.5, on the other hand, they are seen by Attis not as a valuable asset but rather as burdensome, undesirable baggage. At v.6 Attis feels his limbs at last unencumbered (relicta, 6) by his unwanted manhood; only later (v.46) does he see this `burden' for what it really was: liquidaque mente vidit sine quis ... foret. The metaphor in v.5 is still unspecific, the nature of the burden, and the horror of the castration become progressively clearer in the subsequent lines: 6 relicta ... membra sine viro, 7 terrae sola sanguine maculans, 17 corpus evirastis. The recognition that iletas need not have been a corruption of inguinum (vel sim.) now opens new avenues for emendation.
Statius' suggestion, for iletas is icta: `he tore off with sharp flint the weights (i.e. genitals) which he had struck':
deuolsit icta acuto sibi pondera silice. But how appropriate is icta really? Are we to suppose that Attis not merely cut off his genitals, but beat them first as well? An act of tearing does not require an additional act of striking.
I suggest a different track. An obvious candidate to fill the metrical gap in the place of iletas is the pronoun ille, but this gives an inappropriate emphasis: the subject is still the same as it was in the previous four lines. Better is ipse,(8) which is `used to lay stress on the reflexive relation':(9)
deuolsit ipse acuto sibi pondera silice. The combination of ipse with forms of the reflexive pronouns (me, te, se, etc.) and of the reflexive possessive adjective (meus, tuus, suus, etc.) is well attested in expressions of violence directed against the self, and especially suicide: Virgil, Aen. 12.659f. dextra \ occidit ipsa sua and Tacitus, Ann. 6.14 suam ipse ceruicem perfregit. Catullus himself uses this expression to emphasize the reflexive relation at 22.17 se ipse miratur, 31.5 vix mi ipse credens, 67.30 qui ipse sui gnati minxerit in gremium (emphasizing the incestuous relationship), 76.11 istinc te <ipse> reducis,(10) 88.8 non si demisso se ipse uoret capite (emphasizing the self-directed sexual act) and 107.5-6 ipsa refers te \ nobis. The expression is used of self-induced harm at 64.81-2 ipse suum Theseus pro caris corpus Athenis \ proicere optauit, 75.2 se officio perdidit ipsa suo. Most tellingly, the expression occurs three times in poem 63, viz.: 45 simul ipsa pectore Attis sua facta recoluit, 56 cupit ipsa pupula ad te sibi derigere aciem and 85 ferus ipse sese adhortans ra<b>idum incitat anim<um>. In line 5 the expression would serve to emphasize the horror and unnaturalness of Attis' self-mutilating act, providing not only an idiomatic Latin expression, but a powerful and effective sense in context.
The conjecture is admittedly far from the paradosis. But it may be a mistaken enterprise to try to reconstruct the true reading from the hopelessly corrupt iletas. iletas is extremely close to the sequence -lictas- which occurs in the line below; it is even conceivable that in some lost exemplar the former appeared directly above the latter.(11) The visual similarity between lines 5 and 6 is further increased by the occurrence of sibi at the beginning of the second half of each line:
5 deuoluitiletasacutosibi ...
6 itaq;utrelictasensitsibi ... The garbled sequence iletas in line 5 may therefore have arisen through parablepsy, a misreading of the letters directly below. In that case, the true reading which it ousted need not have any paleographical resemblance to it.(12) This would entitle us to emend the text purely on grounds of sense and metre.
The conjecture ipse improves on previous conjectures by offering an expression which is excellent Latin and common in Catullus. Emphasizing the reflexive relation in Attis' horrific and unnatural act, it is entirely appropriate in context. Moreover, there is a ready explanation of the corruption in the supposition of parablepsy.
(1) I am much endebted to Oliver Lyne for his advice and various suggestions which I have incorporated.
(2) So R. Ellis, Catulli Veronensis liber(2) (London, 1878).
(3) There are only two doubtful occurrences of ile in Latin literature. (i) Servius says of Virgil's practice ad A. 7.499 and Ecl. 7.26: `hoc ile' et `haec ilia' facit; but the singular form does not in fact occur in Virgil (A.7.499 and Ecl. 7.26 both have ilia). (ii) It is the conjecture of H. W. Garrod (P. Papini Stati Thebais et Achilleis [Oxford, 1906]) at Statius, Theb. 9.766 figitur <ile> Lamus, flet saucius inguina Lygdus. The manuscript P reads ilia, most editors follow the vulgate in reading ora.
(4) Latin does have a number of metaplasts. But for ilia a second declension form is attested only once: iliis (dative) at Celsus, Med. 4.1.12 iliis (ablative) is a conjecture at Seneca, Dial. 7.7.1 (MSS illis). Perhaps ilibus should be read in both these places? The only place where a form of ilia occurs which is both second declension and singular is Servius, A. 7.499 ab ilio, this, however, is a pun on ab Ilio (`from Troy / from the groin'), cited from the mimographus Marullus. Servius himself condemns the forth: `ab ili' debuit dicere (cf. Servius, Ecl. 7.26: est autem hoc dictum per wnaritudinem rusticam).
(5) Cf. J. N. Adams The Latin Sexual Vocabulary (London, 1982), 71.
(6) With Catullus 63.5 acuto ... silice cf. Ovid, Fast. 4.237 saxo ... acuto and with Catullus 63.6 itaque ut relicta sensit sibi membra sine viro cf. Ovid, Fast. 4.242 nullaque sunt subito signa relicta viri.
(7) Followed by R. G. M. Nisbet, Collected Papers on Latin Literature (S. J. Harrison [ed.], Oxford, 1959), 99 = PCPS 24 (1978), 112.
(8) The same conjecture was also arrived at independently by Dr S. J. Harrison.
(9) B. L. Gildersleeve & G. Lodge, Latin Grammar (London, 1985), [sections] 311, R.2.
(10) Read thus by Ellis, Thomson and Goold. OGR have istinc teque reducis.
(11) Particularly if the -que of itaque were written compendiously (as my representation below supposes).
(12) For the parablepsy, cf. Catullus 95.5, where the reading of OGR, Zinirna canas, is probably influenced by Zinirna cana in the line below, and cf. Nisbet (op. cit. n. 7), 98 ad loc.: `canas has clearly jumped from the line below, and in such circumstances palaeographical considerations should not be given undue weight.'
|Printer friendly Cite/link Email Feedback|
|Publication:||The Classical Quarterly|
|Date:||Jul 1, 1996|
|Previous Article:||On the construction of the 'Syracusia' (Athenaeus V.207-A-B).|
|Next Article:||Mythological incest: Catullus 88.|