Contrast: Adversative and Concessive Expressions on Sentence and Text Level (review)

The inner page subtitle is more informative: ‘Adversative and concessive relations and their expressions in English, German, Spanish, Portuguese on sentence and text level’. With occasional brief references to Latin and French in addition to the four title languages, the study displays great descriptive energy and is intricately subclassified, as reflected in the book’s chapter, section, and subsection organization. As a whole, the book is peppered with highly insightful observations, most densely so in the ‘Survey of research’ (Chs. 3–5; 57–237).

The theoretical highpoint of the volume is Rudolph’s concept of ‘contrast’ as whatever the adversative and the concessive relations have in common, essentially allowing the near paraphrasability ‘X but Y’ as ‘(al)though X, Y’, and equivalents in the other title languages. This commonality is semantically characterized as marking the contradiction of a pragmatically assumed ‘causal chain’ of the type ‘if X, (then) NOT Y’ (31ff.). Since the order of the propositions X and Y (clauses or other-sized discourse constituents) is inherent in this characterization of contrast, R dilutes her generalization in considering postposed though-clauses and concludes that these are of two types, one maintaining her characterization of contrast and another in which the second proposition expresses a ‘restriction’ on interpretation of the first proposition (e.g. 411ff.). Setting aside the problematic relationship of ‘contrast’ to ‘restriction’, her semantic characterization of contrast as a ‘broken [i.e. contradicted] causal chain’ allows her to recognize a great many other expressions of contrast in the title languages in addition to the major (‘most frequent’) adversative and concessive markers, for example, English however, nevertheless, yet, still, anyway, at least, after all, and clause-final though. This leads to a highly complex but well-organized and richly exemplified discussion which takes up the main body of the book. The greatest value of the book is in its generous and impressively varied sets of written examples in all four languages and in the book’s organization and index, which allow specific points of interest to be easily located. The ‘connective index’ (534–39) is particularly well-suited to this purpose, separated into the four title languages and Latin, with a final section on ‘other languages’, all of which are of European origin and some of which are Slavic languages referred to in the text only once in discussing the specialized ‘corrective’ use of but (as in English not (only) X but (also) Y), equivalent to German sondern/Spanish sino (144). The book is the fruit of extensive research, involving selection and organization of the written examples. Both features are invaluable as a resource for further and more adequate descriptive and theoretical studies.

Despite the generally carefully reasoned discussion in the book, R introduces some highly problematic notions that need to be addressed, necessarily briefly in this notice. In my view, the most contentious one is her suggestion that contrast reflects ‘the underlying European view . . . that the world is structured according to regular and predictable causal laws’ (237), repeated regularly in other summary passages (e.g. 9, 437). Swahili, among other non-European languages, immediately refutes this restrictive claim by having an adversative lakini, linking coordinate clauses and various concessives, for example ingawa, subordinating and allowing postposition to a main clause. Moreover, both colloquial speech and idiomatic writing exhibit a correlative construction ‘ingawa X lakini Y’ which supports R’s point about a common semantic element to both types of connectives but, clearly, not her association of this means of expression (let alone ‘thinking’) with ‘the European way of thinking’ (9). It falls to further research to establish how common specialized markers of contrast are among the world’s languages before drawing conclusions about why languages may differ in this respect, to the extent that they do.

Another problem involves her dismissal of spoken data on the grounds that spoken language (dialects?) has no ‘standard’ (11). This stance prevents her from recognizing, among other...