When Evander conducts Aeneas around the future site of Rome, the objects of interest he points out to his guest include the following: "hinc lucum ingentem, quem Romulus acer asylum / rettulit, et gelida monstrat sub rupe Lupercal" (8.342-43). John Conington (1883, 119) complained that rettulit had "not been satisfactorily explained." A century later, Gerhard Binder (1971, 118 with n. 27) repeated the complaint: "Über Vergils rettulit ist viel gerätselt worden." Binder's study was then followed by the unprecedented appearance of three different commentaries on this book of the Aeneid in three successive years. 1 None of them was however able to offer a satisfactory explanation of rettulit: the latest discussion of this passage is obliged to reiterate Conington's original complaint. 2 The purpose of the present article is to suggest a possible solution to this vexatious and long-standing crux. Servius Danielis was himself perplexed by this Virgilian passage, for which he felt the need to offer two alternative explanations: "retulit aut restituit aut nominavit" (Aen. 8.342). The second of these was accepted by Heyne (1833, 221: "retulit est appellavit"); however Conington rightly objected that there is no evidence for such a meaning of referre. The first exegesis proposed by Servius has on the other hand found widespread support. 3 It must however be pointed out that those who accept the equation of referre with restituere appear to have forgotten that the object of rettulit is not asylum, but lucum: while the "re-establishment" of an asylum makes excellent sense, the same can hardly be said of "re-establishing" a grove of trees. Servius himself detected an allusion to the asylum founded at Athens by the Heraclidae: "ideo ergo ait 'quod Romulus acer asylum retulit,' hoc est fecit ad imitationem [End Page 527] Atheniensis asyli" (Aen. 8.342). This interpretation of referre as signifying "to reproduce from a model" has similarly been widely accepted. 4 Such a construe is however open to the same objection that can be raised against restituere: while this rendering would have been admirably suited to asylum as direct object, the notion of "reproducing" a grove "from a model" is likewise far less satisfactory. 5 In support of both these explications it has been customary to adduce Aeneid 5.596-98: "hunc morem cursus atque haec certamina primus / Ascanius, Longam muris cum cingeret Albam, / rettulit." The supposed parallel is however illusory, since here referre simply governs a direct object, whereas in book 8 the same verb is instead employed in conjunction with the predicative apposition asylum. K. W. Gransden's commentary on this latter book (1976, 129) accordingly proposed the novel rendering "instituted as," which was however rejected by P. T. Eden's postscript (1986, 439) to his own commentary as "emphatically not a meaning of refero." Eden's commentary itself had taken rettulit as equivalent to reddidit. 6 This view that referre is here being used in the sense of "to render" had already been tentatively advanced by Conington himself; it too has received a certain measure of support. 7 In his latest contribution to the debate, however, Eden (1986, 440) changes his mind and instead adopts the unsatisfactory notion of Fordyce's commentary that rettulit signifies "reproduced from a model." Eden's reason for this volte-face is the following: "even if . . . the construction here of referre with a double accusative is an analogical imitation of reddere with the same construction, it cannot therefore also be presumed to share its meaning." Perhaps it is nonetheless possible to indicate how this objection of Eden's can be circumvented by a novel hypothesis which would at last appear to provide a key to this troublesomely enigmatic text. [End Page 528] Quintilian (Inst. 8.6.37) illustrates the rhetorical figure of metalempsis or transumptio with a reference to Odyssey 15.299 where thoos ("quick") is employed in the sense of oxus ("sharp"). Heinrich Lausberg (1990, 295) explains the figure as follows: "Die...
Read full abstract