MORE FALSA GELLIANA.
1. Hertz, in his editio maior of Gellius (vol. ii, p. xxvii), cites from Peter Damian (1007-72), Opusculum tricesimum tertium, de bono suffragiorum et variis miraculis, praesertim B. Virginis, caput primum Quod somniis non sit credendum, in Opera, ed. C(onstantinus) Cajetanus [= Gaetani] (Bassano del Grappa, 1783), iii.573-4,(2) the words:
Hic ad memoriam redit, quod, sicut Gellius ait, Alexander somniavit, ne somniis crederet. Vbi quidquid eligat Alexander, hoc disceptatio fine concluditur: ut illi, quod viderat, somnio non credatur. Nam si somniis lure creditur, somnium illud, quod asserit, non credendum esse mentitur. Quod si nequaquam debet somniis credi, consequitur etiam, ut nec illi fides debeat adhiberi.
The opusculum is in fact a letter written in the spring of 1064 to Abbot Desiderius of Montecassino; it has now been critically edited by Kurt Reindel,(3) who reveals the name to be not Gellius but Grillius.(4) The editor asserts: `Bei Grillius, Commentum in Ciceronis rhetorica, einem Grammatiker des 4. Jh., la[Beta]t sich dieses Zitat nicht nachweisen';(5) to be sure, the passage is not among the extracts printed in Carolus Halm, Rhetores Latini Minores (Leipzig, 1863), pp. 596-606, but in the full text as edited by Josef Martin, Grillius: Ein Beitrag zur Geschichte der Rhetorik (Studien zur Geschichte und Kultur des Altertums 14/2-3 [Paderborn, 1927]), p. 7, ll. 14-18 we read, as the eighth and last of the faults that deprive a controversia of coherence:(6)
"[GREEK TEXT NOT REPRODUCIBLE IN ASCII] inops, qui quid agat non invenit, sed etiam contrarium facit quam cupit, ut illud est: `Alexander somnio monitus est, ne somniis crederet. Deliberat, quid agat.' Quicquid enim persuaseris, in contrarium venit; si ut non credat somniis, huic credit; si ut credat, huic ergo non credit.
The example is repeated at p. 53, ll.25-7:
`"[GREEK WORD NOT REPRODUCIBLE IN ASCII], id est inops, cum dissuasione favente laborat persuasio, ut puta Alexander somniavit, ne somniis crederet; deliberat credere.(7)
The eight vices, in a different order, had already appeared at Hermogenes, [GREEK TEXT NOT REPRODUCIBLE IN ASCII], 1, this case being no. 4 (p. 33, 11.3-7 Rabe): [GREEK TEXT NOT REPRODUCIBLE IN ASCII].
2. In 1193 or shortly afterwards, Gerald of Wales, in an invective against William Longchamp, bishop of Ely, papal legate and virtual viceroy of England, declared:(8)
Necesse est ut multos timeat quem multi timent, et qui multos offendit a multis exosus habeatur. Quoniam ut ait Agellius cunctorum meretur odium qui omnium in commune se approbat inimicum.
Gellius says nothing of the kind; but at Jordanes, Getica 187, Valentinian III, seeking the support of the Visigothic king Theuderid of Toulouse against Attila in 451, urges the argument:
cunctorum etenim meretur hic odium, qui in commune omnium se adprobat inimicum.
How Gerald came to know Jordanes I leave for others to determine;(9) but the ascription to Gellius admits instructive explanation. The verse `necesse est multos timeat quem multi timent' (Laberius 126 [R.sup.2] ap. Sen. De ira 2.11.3) is appended--with intrusive ut--by John of Salisbury, Policraticus 8.14,(10) to a sequence of sententiae ascribed to Publius Clodius; they are in fact the sententiae of Publilius Syrus recorded by Macrobius, Saturnalia 2.7.10, who himself had them from Gellius 17.14.(11) Gerald, recognizing them from the Gellian section of the Florilegium Angelicum,(12) associated `Agellius' with the extra verse he had found in John, but assigned his name to the next quotation instead.(13)
Such confusions are not exclusive to medieval authors, but were facilitated, on the one hand by the use of florilegia,(14) on the other by the ascription of authority rather to antiquity (especially as preserved in such anthologies) rather than to individual ancients, which made the anxious assignment of suum cuique seem less important than it does to us. The point is not to rebuke them, but to learn from their dealings with known texts how little to trust them on unknown ones.
(1) See LCM 9.10 (December 1984), 151; 10.1 (January 1985), 16; 15.10 (December 1990), 150-1, 18.8 (October 1993), 126-7; [CQ.sup.2] 44 (1994), 486.
(2) Reprinted from Opera (Pads, 1642), iii.251; in Gaetani's first edition, Opera (Rome, 1606-15), i.103-4 and the separate edition of Peter's letters (Pads, 1610) this was epist. 2.14. cunctorum etenim meretur hic odium, qui in commune omnium se adprobat inimicum.
(3) Monumenta Germaniae Historica, Epistolae 2: Die Briefe der deutschen Kaiserzeit, 4: Die Briefe des Petrus Damiani, iii (Munich, 1989), no. 106, pp. 168-85. Our passage appears on p. 169.
(4) For another instance of Grillius wrongly emended to Gellius see Hertz, loc. cit.
(5) P. 169 n. 3; the statement is not corrected in the supplementary note at vol. iv (Munich, 1993), p. 547.
(6) Quorum si aliquid inciderit, manifestum est controversiam stare non posse; that is to say, it is [GREEK WORD NOT REPRODUCIBLE IN ASCII] (parum consistens Gell. 9.15.6), a term for which there is no recognized English equivalent. Malcolm Heath, in his translation of Hermogenes, On Issues (Oxford, 1995), p. 30, makes such questions `lack issue'.
(7) Read deliberat <an debeat> credere?
(8) De vita Galfridi archiepiscopi Eboracensis 2.19 (`Descriptio beluae multiformis'), J. S. Brewer (ed.), Giraldi Cambrensis Opera iv (London, 1873), p. 425.
(9) The same passage of the Getica was exploited soon afterwards in a letter ostensibly written to Philip of Dreux, bishop of Beauvais, who had been taken prisoner by Richard's brother and subsequent successor John on 19 May 1196, by Pope Celestine III (d. 8 January 1198), reproduced in Roger of Howden (d. 12017), W. Stubbs (ed.), Chronica, s.a. 1197, iv (London, 1871), 23: `Sinistre licet tibi euenit, nec mirum: cunctorum enim meretur odium qui omnium se in commune approbat inimicum.' In a footnote, Stubbs observes `It is hardly necessary to remark that this letter is a fabrication'.
(10) Ed. C. C. J. Webb (Oxford, 1909), ii.335.
(11) Publilius is Publius in most MSS of Macrobius as of Gellius; P. Clodius has just been mentioned at Sat. 2.6.6. That Macrobius, not Gellius, is John's source appears from the most cursory comparison.
(12) On which he largely relied for his classical authorities: see A. A. Goddu and R. H. Rouse, `Gerald of Wales and the Florilegium Angelicum', Speculum 52 (1977), 488-521.
(13) One may envisage a commonplace book in which Gerald, having ascribed the verse to Agellius in the margin, followed it with the unattributed passage from Jordanes.
(14) As when Gerald assigns to Gellius a passage of Ennodius also known to him from the Florilegium Angelicum (Goddu and Rouse, art. cit., pp. 512-13) or Petrus Cantor a passage of Valerius Maximus from the Valerio-Gellian florilegium (Holford-Strevens, [CQ.sup.2] 44 , 486).
email@example.com firstname.lastname@example.org 67 St Bernard's Road, Oxford