Waltereit's work explores the nature of French third-person non-clitic (disjunctive) reflexive pronouns and the syntactic changes they have undergone from Old French to the present day. While the inventory of relevant forms has not evolved greatly during this period, significant developments have nonetheless occurred, in particular with soi, whose evolution commands particular attention. Findings from recent theoretical work and data from various French corpora are used throughout to inform the discussion, yielding a presentation which is both clear and instructive, although it may perhaps prove somewhat challenging in places for the more traditionally trained historian of the French language.

The text falls broadly into three parts. The first occupies the opening two chapters, where basic theoretical aspects underpinning the later detailed treatment of historical data are introduced. An initial chapter (1–16) considers anaphoric systems in a general way and presents the key distinction between binding (an anaphoric relation within a construction, hence grammatical) and co-reference (an anaphoric relation across constructions, hence discourse-based). Binding is investigated further in the following chapter (17–75), where other fundamental considerations (such as grammatical vs. discourse anaphora, and specificity and definiteness) are also examined. It is concluded inter alia that soi displays grammatical anaphora and hence expresses a binding relation, whereas anaphoric lui and elle express a co-reference relation and thus relate to discourse; and also, crucially, that a non-specific anaphor of a clause-mate antecedent forms part of a predicate, whereas a specific anaphor constitutes an argument. Admittedly (31–39), the dividing line between specific and non-specific NPs may not be absolute, as the latter can become specific in the flow of discourse. On the other hand, certain problems, duly acknowledged, remain unresolved, in particular the distinction between argument and adjunct status in prepositional phrases, e.g. parler de soi/lui vs. voir devant soi/lui (70–74).

The second part of the work addresses diachronic issues, with three chapters examining, respectively, the changing balance of usage between reflexive soi and personal pronoun lui/elle (77–132), the interplay between simplex reflexives and their suffixal counterparts in -même (133–160), and evolution in intensifier forms containing -même (161–175). Extensive data are used, drawn from three corpora of contemporary spoken French and four of written French covering the period from the earliest texts onward. The analysis leads to some interesting observations on details of change. For example, the claim that soi serves as an anaphor for non-specific indefinite pronouns only (e.g. on, chacun) is shown to be untenable (89) as full non-specific NPs may also occasionally trigger soi, as in Ils (= Les enfants) arrivent chez soi, an example contained in one of the contemporary spoken corpora. Also, the general decline in the use of anaphoric soi has been offset in recent times by the rise of soi in idioms and other non-anaphoric functions, notably soi-disant and aller de soi; the latter is only attested from the 19^th century (95). The investigation of usage with reflexival pairs (lui/lui-même etc.) prompts further proposals on theoretical issues, e.g. that suffixal -même serves as a predicate focus marker whilst the non-suffixal anaphor indicates argument focus (153–160). However, diachronically, there appears to have been no significant functional change from Old French onward in the distribution of the two types of anaphor in individual predicates. Lastly, intensifiers containing suffixal -même are shown to fall into three types, viz. adnominal modifying an NP (le produit lui-même), adverbial-exclusive modifying a VP (il construit lui-même sa maison) and adverbial-inclusive modifying a VP as well but carrying a repeatable/transferable value paraphrasable with ‘also’ (je l’aiderai à perdre du poids, j’en ai moi-même perdu 5 kilos). Diachronically, the first is attested from the 13^th century, the second from the 16^th, and the third from the 17^th.

The final part, which consists of a chapter on the etiology of the diachronic changes observed (177–204), is less successful. After a tidy summary of generative and functional approaches to explaining language change, it is proposed that grammatical change is motivated by speakers seeking to produce ‘maximally noteworthy and relevant contributions to conversation’, where ‘noteworthy’ implies forcefulness or rhetorical over-use. Resulting deviant or marked forms may subsequently, through frequent use, become unmarked. A sketch is then offered of two diachronic test-cases, French negation and the rise of compound past tenses, which purportedly illustrate rhetorical over-use. Turning finally to the increased use of the personal pronoun over reflexive soi in clause-mate anaphora, the author suggests that this can be analysed as ‘over-use of the pronoun’, the pronoun being more speaker-centred and expressive than the reflexive, ‘thus creating a choice between a more vs. a less speaker-centred form’ (197). As it stands, the proposal, though interesting, seems unlikely to convince all historians of the language.

Overall, the volume, which comes complete with a brief concluding chapter largely summarising the results, a bibliography, an appendix with tables detailing statistics of anaphoric soi/lui in individual contexts, and a brief index, offers a focused and often penetrating study of an intricate area of French syntax. The author is to be thanked for a valuable contribution to the field.