The conscientious commentator will offer his work to the public in a mood of doubt and self-questioning. As an Editor of the series in which this edition appears I have felt a special duty to keep in the forefront of my mind its declared aim: ‘to provide the student with the guidance that he needs for the interpretation and understanding of the book as a work of literature’. The amount of guidance here provided may, however, strike some readers as excessive. If so, it is because it has seemed to me that in the past Lucretius’ interpreters have not always taken enough pains to disentangle and follow his argument as he intended it to be followed, and this, whatever shortcomings may be found in the execution, is what I have attempted to do. The De Rerum Natura, in spite of the lucid style of which the poet was rightly proud, is a difficult book, and I have often preferred the risk of telling the reader what he already knows to that of leaving him in the lurch – the besetting sin of commentators. It may also be felt that there is here too much expatiation on the poetical techniques of Lucretius. In this department the existing commentaries seem to leave much to be desired. In spite of the lead given by H. Sykes Davies in his Criterion article of 1931–2 and in spite of more recent contributions in this field such as Professor David West’s excellent The imagery and poetry of Lucretius (1969), the conventional idea of Lucretius’ art still persists: ingenio maximus, arte rudis. Cicero knew otherwise; but posterity has yet to be convinced. The student who finds some notes inordinately long may care to note that an effort has been made to, so to say, ‘grade’ their contents so that the essential information is usually presented at the beginning.

Two considerations have prompted the thought that a second edition of this book may not be unwelcome. As informed interest in Lucretius has continued to grow – a fact strikingly illustrated by the number of translations that continue to appear – so it must be accepted that students now come to him less well prepared linguistically than was the case in the 1970s. Accordingly the Commentary has been extensively revised and enlarged, with, it is hoped, due account taken of the comments of reviewers – though my old friend and critic Professor David West would now miss in the notes the ‘brevity that comes close to wit’ that he admired in the first edition.

Three passages in the Introduction to the first edition have called for reconsideration: what was said there about the ‘middle’ or ‘florid’ style of oratory, the discussion of the diatribe and what is said about the spelling of seorsum. These points are dealt with at p. 13 n. 51, p. 14 n. 55 and p. 18 n. 73 respectively. The section on the text has been rewritten in the light of subsequent work in that field, especially that of Professor Michael Reeve and Dr David Butterfield, to both of whom I am greatly indebted for help and advice generously given. The apparatus criticus has also been revised in accordance with Dr Butterfield’s advice. Otherwise the Introduction is reprinted unaltered apart from a handful of additions to the footnotes in addition to the three noted above, and adjustment of the references to the secondary literature in conformity with current series style. In the Supplementary Introduction I have confined myself for the most part to comments on such post-1971 contributions to Lucretian studies (some of which are in any case noticed in the revised Commentary) as seem likely to be useful to readers of this book of the De Rerum Natura. I regret that it has not been possible to include references to all the important work of Professor Ferguson Smith on the Oenoanda inscription.

Introduction

1–30

Prooemium. Lucretius begins each book of the poem in the high ‘pathetic’ style (genus grande, amplum, acre (Kenney 2007: 93)); here and in Books i, v and vi, it takes the form of a panegyric on Epicurus (Introd. 10–11). Only here is he addressed in the second person. For a sensitive analysis of the verbal structure and imagery of these lines see Stokes 1975. L. may intend to suggest a contrast with the false revelations of mystery religions, also expressed through images of light and darkness (Richardson 1974: 26–9). Whereas initiates in those cults were sworn to silence about what they had seen, the revelation offered by Epicurus is freely displayed in his teaching and writings and in L.’s verse.

1-4 These opening lines are carefully structured to create and then fulfil expectation. The apostrophe announced by O is deferred by the intervening qui-clause and so gains in weight; the grammatical and rhetorical structure is articulated by the sequence of monosyllables o – qui – te, each beginning a verse. This simple structure (characteristic of L.’s technique: Kenney 1977b/1995: 29; 2007: 103–4) is complicated and enriched by anticipation and enjambment (extollere…potuisti, tuis…signis) and by ‘theme and variation’, sequor being expanded in what follows. As throughout the Prooemium and indeed the whole poem the emotional impact is reinforced by alliteration. The whole style and feeling of L.’s address to Epicurus is hymnic, implicitly anticipating the explicit identification of 5.8 deus ille fuit, deus. Cf. 9–10nn. O: this, the reading of OV, and not the humanist conjecture e, is certainly what L. wrote. The sonorous interjection matches the emotional tone of the passage better than the prosaic (indeed superfluous) preposition (Timpanaro 1960/1978) and imparts ‘an elevated note appropriate to prayer’ (N–H 364 on Hor. C. 1.32.13).

During the forty-odd years since the publication of the first edition of this book, informed interest in Lucretius has continued to grow: a process reflected both by the continuing increase in the number of translations (Kenney 1985: i, x) and the proliferation of the scholarly literature (Gale, ‘Addenda’ in Kenney 1977b/1995; list of works cited in Gillespie and Hardie 2007). The volume on Lucretius in the series Oxford Readings in Classical Studies (Gale 2007a) has made a number of influential articles conveniently available.

That, ‘as weapons in outer space threaten to move from science fiction to fact, Lucretius…seems more tragically contemporary than ever’ (Segal 1990: 227), and that he has much of importance to say to today’s politicians and scientists is strikingly argued in the concluding chapter, ‘Wizards in bondage’ (135−55), of W. R. Johnson’s reading of the DRN as ‘a powerful representation of classical antiquity’s best and most influential effort to speculate on the material universe and the place and conditions of human beings in it’ (Johnson 2000: ix). This is a welcome reminder of the fact that the DRN ‘is the one didactic poem, ancient or modern which is both great poetry and good science’ (Kenney 1998−9: 418).

Introduction

An alternative model of competitive markets in an infinite-time economy is the Arrow–Debreu model of contingent commodity markets. In the Arrow-Debreu model it is assumed that there exist at date 0 markets for consumption at every future date contingent on every event. Agents trade only at date 0 facing a single budget constraint that restricts transactions in all contingent commodity markets. Time plays no explicit role in market transactions. Transactions are agreed on at date 0 and executed at their respective dates. There is no reason for further trade at future dates. This market structure is distinguished from the security markets of Chapter 29 where agents trade sequentially at all dates and face separate budget constraints for every date and every event.

The Arrow–Debreu model of contingent commodity markets is hardly realistic. Yet, it serves as an important tool for the analysis of infinite-time security markets. Extending the results of Chapter 23 for multidate markets, we show that Arrow–Debreu equilibria in contingent commodity markets and equilibria in security markets under debt constraints have the same consumption allocations when security markets are dynamically complete and debt bounds are nonbinding. However, the specifications of primitives under which equilibria exist are not the same in the two cases, as we see. The substantial body of knowledge about the properties of Arrow–Debreu equilibria can be used at least for a subclass of models of security markets. In particular, it can be used for the study of representative-agent models.