The admiration many readers feel for Virgil's word order was expressed by J. Dryden in the preface of his translation of C.-A. du Fresnoy's work on the art of painting: ‘Virgil is so exact in every word, that none can be changed but for a worse; nor any one removed from its place, but the harmony will be altered’ (The Art of Painting [1695], p. xlix). D.’s work is the first book-length study of word order in Virgil, and of any Latin poet since H.D. Naylor's Horace Odes and Epodes: a Study in Poetic Word-Order (1922). It is a revised and expanded version of D.’s doctoral thesis (Università degli Studi di Salerno, 2012), produced partly under the guidance of G.B. Conte, and it has been translated into English by A. Campbell. Its focus is on the Aeneid, though quotations from other Virgilian works are abundant and helpful, and it concentrates particularly on aesthetic phenomena, especially that work mimetically or rather, in D.’s more precise Peircean terminology, iconically.

Surprisingly perhaps, many of these effects are not conventionally harmonious at all, but work by being ostentatiously off-kilter or deviating from an expected pattern of some kind. For instance, enjambement, the phenomenon to which D. gives the most attention, works by dislocating a sense unit across a verse break. Other prosodic and metrical effects that D. considers work similarly, for example hiatus and hypermeter, by breaking established norms. Even a golden line (A B C B A), which seems visually so harmonious, violates norms that operate at a different linguistic layer, namely the syntactic constraints that structure typical Latin word order.

D.’s attention to the aesthetic dimension of word order distinguishes it from recent Anglophone work on Latin (and ancient Greek) word order, for instance, by A.M. Devine and L.D. Stephens, B.L.M. Bauer, and A. Ledgeway. Instead, D. works largely within a framework and terminology that should be familiar to most Classicists: ‘hyperbaton’, ‘anastrophe’ and ‘emphasis’ instead of, for example, ‘discontinuity’, ‘long-distance fronting’ and ‘focus’. This makes the work accessible to a wide range of Classicists and demonstrates the continued hermeneutic value of these concepts in the hands of a sensitive reader. One advantage of the familiar terminology is its ancient pedigree: through Servius and Quintilian it takes us back perhaps as close as we can get to the vocabulary in which Virgil himself probably understood his poetic creation.

For describing enjambement and other effects that cluster around verse breaks, D. persuasively advocates adopting two concepts from M. Grammont (Le vers français [2nd edn 1913]): rejet and contre-rejet. Both help to describe the aesthetic tension between the metrical unit of the verse and the syntactic unit of the sentence. Rejet occurs when a single word has overstepped the boundary of the metrical verse and been projected, or enjambed, onto the following verse, and is then followed by a pause (p. 58). Contre-rejet describes the mirror image of this, when the sentence ends before the verse does, and a new sentence begins shortly before verse-end. D. also introduces a third term of his own, après-rejet, which describes the start of a new sentence immediately after rejet. This toolkit of vocabulary helps D. to break new ground in his description of enjambement and related phenomena and their interaction with metrical structure and semantic content.

Another strength of D.’s approach, which makes this book a valuable reference work on Virgil's language in general, is his willingness to treat word order in relation to a whole range of prosodical and rhetorical effects. There is no better place to turn for concise and well-documented discussions of figures such as: dicolon abundans (p. 35 with n. 118), ‘arching alliteration’ (p. 52), ‘apposition of pathos’ (p. 45 n. 159), ‘grammatical rhyme’ (p. 51 n. 2), ‘ribattuta’ (p. 77 n. 254), ‘poetry of names’ (p. 79) and schema Cornelianum (p. 102 n. 330). D. is at his best when describing how several of these effects work together in structuring and propelling forward complex narrative passages. Most of these are found in the handy index rerum. In assembling this material, D. has drawn on the full range of modern commentaries and secondary literature, providing clear formulations that are reliable and informative.

All the same, the absence of contemporary linguistics, especially pragmatic approaches to word order, narrows the range of questions that D. pursues. One very large question that needs more attention for Latin poetry is the one posed by H. Dik in Word Order in Greek Tragic Dialogue (2007): namely, to what extent the normal ‘prose’ rules of Latin word order continue to operate in poetry. It is possible that some of the salience that D. attributes to metrical placement can actually be explained by typical syntactic operations, for instance the fronting of verbs under focus (e.g. Aen. 2.52 stetit in what D. calls après-rejet, p. 131). Answering this question would require more attention to mundane clause structure in the Aeneid and its articulation by postpositives, which are not given much attention here. The absence of a theory of focus also reduces D.’s ability to describe precisely how the information content is affected by word placement. D. often falls back on saying that the ‘semantics’ of a word or phrase is ‘emphasized’ or ‘highlighted’. A more nuanced approach might describe the information as new, corrective, counterpresuppositional, contingent or resumptive, among the many kinds of attitudes that can be expressed through focus.

After an introduction with a useful typology of iconicity, the first and longest chapter addresses enjambement, in particular what D. calls ‘interlinear juxtaposition’ (a noun-adjective phrase that is interrupted by a verse boundary), along with rejet and contre-rejet. Within these categories, the material is arranged by both grammatical and semantic categories. The semantic groupings bring together parallel passages and lead to valuable generalisations; for instance, a verb in rejet tends to describe falling, throwing, violence or momentary actions. Likewise, apres-rejet often contains the effect after a cause, the reaction to an action (especially a command) or a psychological response. These conclusions are based on an exhaustive collection of evidence for the effects discussed, an enormous and praiseworthy labour.

The second chapter discusses a grab bag of effects: elision (called synaloepha here), hypermetric lines and various Graecising devices (hiatus, ‘irrational’ lengthening, polysyllabic and spondaic line ends), caesurae and monosyllabic line endings. Although the caesura is discussed many times throughout the book, the eight pages of discussion here, though useful and convincing, are too brief to do justice to its importance as a structural device. By contrast, the other devices are given excellent coverage. Especially noteworthy is the observation that elision occurs with verbs of fusing or conflicting, semantic categories that are effectively antonyms. This seems to imply that the effect had no semantic or mimetic content in itself but merely strengthened whatever it was applied to. The discussion of various Greek-style effects is very useful and raises broader questions about bilingual interaction.

The final chapter examines particular forms of marked word order: such as the inversion of the position of noun and adjective, the placement of crucial words at the beginning and end of lines, surprising juxtapositions and various kinds of hyperbaton. Some of these effects could receive more extensive treatment, but D. picks out valuable examples and lays the groundwork for future work. The monograph has been translated into faultless, idiomatic English and maintains praiseworthy standards of accuracy.