What is linguistic naturalism? Some readers will think of the etymologies in Plato's Cratylus and the discussion of whether they reveal the true natures of the things denoted by the words. For other readers the phrase will evoke Lucretius’ description of early humans creating language out of natural cries. But neither Lucretius nor the Cratylus make much of an appearance in this work, which focuses primarily on Varro and also on Cicero, Lucilius, Nigidius Figulus, Posidonius and Dionysius of Halicarnassus. Of course, the views of ‘nature’ in these different writers were not the same; often multiple views can be found even within a single text. But the volume embraces this diversity and uses a maximally inclusive definition of ‘nature’ in order to explore the phenomenon to the full.

Does it succeed? That would depend on one's definition of ‘success’. Wolfgang De Melo (‘Naturalism in morphology: Varro on derivation and inflection’, 103–20) offers a detailed study of Varro's use of the word natura in linguistic contexts, providing a persuasive explanation of the term's different meanings and uses, and of Varro's understanding of morphology. David Blank (‘What's Hecuba to him? Varro on the natural kinship of things and of words’, 121–52) argues that Varro's treatment of nature as a linguistic force reveals his debt to the Stoics and their ideas of oikeiosis (accommodation and appropriation). But then in the last chapter James Zetzel (‘Natural law and natural language in the first century b.c.e.’, 191–211) offers a completely different perspective: ‘Varro does not make coherent sense … “Nature” is not a term with a single meaning, and therefore claims to rely on nature are necessarily inconsistent and incoherent; naturalism approaches meaninglessness … “Nature” is what we invoke in order to explain what we do not understand and therefore cannot explain’ (194). This statement challenges the basic premise on which De Melo's and Blank's chapters were based, namely that ancient texts do make sense. If you do not start from that premise, then neither their arguments nor any of the other chapters are convincing. But is that premise as weak as Zetzel claims? Does polysemy really render a word useless? Perusal of the Oxford Latin Dictionary suggests that most Latin words have multiple meanings: are all such words meaningless?

Readers would be in a better position to engage properly with this challenge if the authors and editors of this book had prioritised clarity more highly: confidence that one has fully grasped all the arguments earlier in the book and understands the evidence on which they are based is a pre-requisite for tackling Zetzel's chapter, and few readers are likely to reach it with that confidence. This unclarity has multiple causes. Sometimes it is not made clear what the basis for an argument is, as when we are told that an ‘extreme naturalistic doctrine’ of spelling long i with ei in forms used for senses that are naturally fuller ‘underlies Lucilius’ guideline (358–61 M.; 364–6 M.) for the choice between ei and i both in the body of a word (mille, meilia) and in the case endings (e.g. gen. sing. pueri, dat. sing. puerei)’ (58). But when the relevant passages of Lucilius were quoted (51–2), a normal reader would have thought that Lucilius was advocating the ei spelling in both mille and milia (‘mille hominum, duo milia’ item: huc ‘e’ utroque opus: ‘meille’) and that puerei was a nominative plural (iam ‘puerei uenere’ ‘e’ postremum facito atque ‘i’, ut ‘puerei’ plures fiant). How exactly did the author get from those quotations to that interpretation?

Other sources of unclarity include terminology that will be unfamiliar to many readers (e.g. ‘morpholexical’, 97), distinctions that could have done with more explanation (e.g. ‘he goes beyond our concept of naturalness and reaches the higher concept of naturalism’, 119), expressions that do not seem to make any sense (e.g. ‘φιλήτης < *ὑφειλέτης, with the first letter “taken away”, ὑφεῖλον FOR steals the < ὑφείλω’, 59), and expressions that do make sense, but not the right sense (e.g. ‘contrary to morality’ for ‘unlike morality’, 22; ‘substituting e for i’ for ‘substituting i for e’, 100). This last category is particularly dangerous and is probably connected to the fact that on the whole the standard of English is high, despite most contributors not being native speakers: it looks as though a veneer of good English has sometimes been imposed without full appreciation of what the authors were trying to say.

Fortunately, one good clear section is the introduction (Giuseppe Pezzini and Barnaby Taylor, ‘Introduction: first thoughts on language and nature’, 1–14), which offers an excellent concise summary of the various forms of linguistic naturalism in ancient thought, with useful references to further discussion. Alexander Verlinsky's chapter (‘Posidonius’ linguistic naturalism and its philosophical pedigree’, 15–45) then sets the scene by arguing that the lost works of the Stoic Posidonius, influential on Roman thought, combined two distinct forms of naturalism, the kind in the Cratylus and the kind in Lucretius. Anna Chahoud's ‘Lucilius on Latin spelling, grammar, and usage’ (46–78) offers a close study of fragments that seem to advocate some startling connections between meaning and linguistic form. Then Alessandro Garcea (‘Nigidius Figulus’ naturalism: between grammar and philosophy’, 79–102) bravely tries to reconstruct the thoughts of an even more fragmentary writer and to compare those thoughts with Varro's. One feels on firmer ground with texts that actually survive, which fortunately are treated in the following chapters. ‘Linguistic naturalism in Cicero's Academica’ (Tobias Reinhardt, 153–70) examines a debate in which Stoics and Academics each try to categorise their own linguistic theories as natural and their opponents’ as unnatural, revealing what such arguments show about the strength of linguistic naturalism in Cicero's day. Then Casper de Jonge (‘Linguistic naturalism and natural style: from Varro and Cicero to Dionysius of Halicarnassus’, 171–90) examines the characterisation of rhetorical Atticism (i.e. writing like Lysias rather than like Thucydides) as ‘natural’ style, and how Dionysius’ ideas about this draw on Cicero and Varro.

The book arose from a conference on linguistic naturalism, hence the dissonance between the focus on ‘linguistic naturalism’ inside the work and the vaguer ‘language and nature’ in the title. Some effort has been made to draw out connections between the different chapters, but there is no real unity and little connection even on a superficial level: sometimes the same passages are discussed at length in different chapters without even a cross-reference (e.g. 106 and 131). Nevertheless, some chapters constitute interesting and useful contributions to scholarship. The book is to be welcomed both for those contributions and for the debate it may spark on linguistic naturalism, on the value of words with multiple meanings, and on whether it is legitimate to assume that ancient texts make sense.