This book addresses possibly the most troublesome aspect of Latin word order: the ordering of words within the noun phrase. As Olga Spevak herself points out, ‘the internal ordering of Latin noun phrases and its variability is a very complicated topic. This book by no means pretends to explain everything’ (337). As in her earlier book, Constituent Order in Classical Latin Prose ((2010): see BMCR 2011.06.30), S. adopts the theoretical framework of Functional Grammar. Latinists without specialist training in linguistics will find S.'s treatment easier to follow than, for example, the generativist treatment in Devine and Stephens (Latin Word Order: Structured Meaning and Information (2006)), and a glossary of technical terms is included.

S.'s main contention, if it may be summarized briefly, is that order within the noun phrase cannot be fully accounted for either by syntactical rules or by logical and pragmatic ones; the main emphasis in her investigation is in fact on semantics, and in this respect her approach appears as a continuation and refinement of that of Marouzeau, the most influential (though not always the most systematic) researcher on Latin word order in the last century. Marouzeau famously posited a distinction between adjectives with ‘subjective’ meaning, which precede the noun, and ‘objective’ ones which follow; thus, for example, bonus vir expresses a subjective judgement about the man while navis triremis conveys an objective classification of the ship. It is easy to find examples where this neat distinction appears to break down; nevertheless, S.'s rather more complicated table of factors involved in the position of modifiers (335), while based on a detailed study of a large number of examples, still looks something like an expansion of Marouzeau's principle, with the addition also of logical factors (generic versus specific, contextually given versus contextually new) and pragmatic ones (emphasis and contrast; though bewilderingly, ‘contrast’ appears as a factor making for both anteposition and postposition).

The main material of the book is a series of case studies of the way modifiers of different types are ordered with certain specified nouns. S. starts with a typology of nouns and their modifiers based ultimately on that of John Lyons, Semantics (1977). Among the modifiers, some are given less attention as they are ‘not very problematic’ (for example, demonstrative and indefinite pronouns and possessives) but other categories are treated fully. Quantifiers (omnis/nullus, multus, magnus etc.) are rightly distinguished from ordinary adjectives, although not all their peculiarities are highlighted (for example, their tendency to appear widely separated from their nouns). Ch. 2 covers the most common types of single-word modifiers (for example, quantifying, classifying, descriptive, evaluative, possessive, and ‘valency complements’ such as objective genitives). Ch. 3 covers prepositional phrases, both in their own right and in terms of their integration into a larger noun phrase, while ch. 4 covers apposition of various kinds. The examples are largely taken from a defined corpus of pre-Augustan prose texts (selected texts of Cicero, Caesar and Sallust) which, though limited, is not inadequate for the purpose; as in the earlier book, comparatively little attention is given to differences of style and register.

S.'s consideration of the examples chosen is alert and sensitive to nuances which are not always immediately obvious, for example, the distinction between attributive and predicative adjectives (does milites omnes occisi sunt correspond to ‘all the soldiers were killed’ or ‘the soldiers were all killed’?), and, as one would expect in a functionalist account, to the various pragmatic functions such as topic, focus, contrast etc. Even so, the linguist's almost inevitable convention of taking examples out of context makes it not always easy to check up (without turning away to look up individual passages) on what wider contextual or rhetorical factors may lie behind the word order in any particular case (take as a typical example the variation between dies comitiorum and comitiorum dies (204–5)). Occasionally S. resorts to explanations that do not convince fully; for example, it seems hard to see a valid semantic distinction between dies + numeral expressing ‘how many days?’ and numeral + dies answering the question ‘how much time?’. However, even where the hypotheses advanced seem uncertain or speculative, the questions are always interesting and will provide material for further research.

The overall conclusions are perhaps more tentative than some might expect, but the absence of clear-cut rules is itself to some extent a salutary conclusion, and the greatest value of the exercise undoubtedly lies in the detail. Commentators on Latin texts will find the book particularly helpful, as it is a mine of information (not always obvious or well-known) on Latin usage, including facts about other things than word order. For example, S. was the first to bring it clearly before this reader's mind that ‘a lot of money’ is virtually never *multa pecunia but rather magna pecunia. The grammarian's myth that Latin prepositional phrases do not typically function as modifiers of nouns is decisively seen off in ch. 3; as in other areas, it depends on the semantics of the noun. And some editors of Latin texts will echo, while others may learn from, S.'s plea: ‘there are close and free appositions in Latin, so please punctuate them properly!’ (330). The only major regret is that there is no index verborum, which makes the book unnecessarily difficult to use for reference; could Brill be persuaded to include one if they reprint?