Nølke’s book introduces English-speaking audiences to the so-called théorie SCAndinave de la POlyphonie LINguistiquE, or ScaPoLine. Polyphony is a key topic in what is known as linguistique de l’énonciation, a long-standing French and French-speaking perspective on semantics. It refers to the idea that constructions often evoke ‘voices’ (standpoints) other than that of the speaker: thus, apart from what is actually said, an utterance such as I didn’t want ice cream also conveys, in the background, the notion that I did want or could have wanted some.
Chapter 1 of the book espouses the view (credited to Charles Bally) that dictum (proposition) and modus (speaker stance) are intertwined, and that aspects of language use are conventionally enshrined in language structure. There is no meaningful semantics-pragmatics interface in either ScaPoLine or, more broadly, linguistique de l’énonciation (a label Nølke translates as utterance act linguistics). Given the author’s audience, more care should perhaps have been taken to highlight the importance of this tenet, which stands in stark contrast with the widely practiced Anglo-American division of labour between semantics (language structure) and pragmatics (language use). The chapter goes on to cover various strands of linguistique de l’énonciation spanning much of the second half of the 20th century. The focus is on Émile Benveniste, Jean-Pierre Desclés, Antoine Culioli, Oswald Ducrot, Jean-Claude Anscombre, Nølke’s own work, and Jacques Bres. The author’s failure to clarify how these mostly French scholars build on each other’s work somewhat reduces the overview’s usefulness; fortunately, a subsequent section titled ‘Conceptual background’ provides an overview of the key tenets of linguistique de l’énonciation as Nølke understands it, without much differentiation between the individuals previously discussed. One widely shared tenet is that utterances (énoncés) are types, whereas utterance acts (énonciations) are tokens. In keeping with this, a distinction is put in place between sentence meaning, utterance-type meaning and utterance-token meaning. Equally important is the concept of ‘ideal discourse’, which can be seen as complementing the concept of the utterance-type, whose default context is a form of ideal discourse.
Chapters 2 and 3 focus on the key issue of the book, i.e. polyphony, defined as ‘an aspect of utterance meaning likely to be encoded in the linguistic form’ (39), where there is ‘always a dominant voice (a hierarchy of voices)’. An essential tenet of polyphony theory is that there is no such thing as a unitary speaker; rather, there are several ‘voices’ arranged by the speaker in an utterance-type. Sentential negation provides a straightforward example: a speaker who says The wall is not white sets up two points of view (‘the wall is white’ and ‘the wall is not white’) and mediates between the two. This obviously amounts to saying that negation often is not pragmatically felicitous unless the corresponding affirmative proposition has been activated or is present in the Common Ground, something that is presumably uncontroversial in semantics. For Nølke, and for polyphony theory more widely, it is crucial, though, that the various ‘voices’ are ultimately anchored in the speaker. Four layers are distinguished in the arrangement of voices: (i) the speaker, (ii) points of view, (iii) discourse entities, and (iv) utterance links. The ‘speaker’ splits into the utterance speaker and the text speaker. The utterance speaker is associated with the utterance-token, the text speaker with the utterance-type. Points of view are (potentially controversial) judgments that have a ‘source’. The example The wall is not white incorporates the two points of view mentioned above. A presupposition, too, can be a point of view. Discourse entities can be sources of points of view. Speaker, addressee, and third persons are discourse entities. They are somehow closer to the speaker than the points of view: ‘The LOC [speaker] constructs [discourse entities] as representations of the various “persons” who inhabit the discourse’ (61). Finally, utterance links ‘express the [discourse entities’] attitudes towards [points of views] in those of “their” utterances where they are not themselves sources’ (81). Key for ‘attitude’ is the notion of responsibility. In The wall is not white, the point of view ‘the wall is not white’ is linked to the speaker via a link of responsibility, whereas the point of view ‘the wall is white’ is linked to the speaker with a link of non-responsibility, more precisely of rejection.
Chapter 4 puts the ScaPoLine apparatus to the test by examining a number of constructions in French. Not unexpectedly, the first of these is sentential negation. Like Ducrot, Nølke distinguishes three types: polemic (see above), descriptive (e.g. There isn’t a cloud in the sky; only one point of view) and metalinguistic (e.g. Paul isn’t tall, he’s a giant). The default is polemic negation; descriptive and metalinguistic negation are derivative. An interesting concept is that of ‘triggers’ vs. ‘blockers’ of descriptive negation. Non-gradable predicates like being white are blockers, gradable ones are triggers. Thus, Marie is not quite 40 years old can hardly be interpreted as polemic. Similarly, metalinguistic negation is dependent on certain contextual conditions, and hence derivative from the default. One question that may arise at this point is how the notions of triggers, blockers and derivation fit in with the concept of ideal discourse and utterance-types. Also investigated in ScaPoLine are French subjunctive constructions. The key assumption is that they embody two points of view (one encoded in the matrix clause, the other in the subordinate clause) and that these are ascribed to different sources. The ScaPoLine stance differs from the prescriptive tradition, which locates the core function of the subjunctive in an indication of uncertainty or subjectivity. Other constructions analysed involve modal adverbials (peut-être, probablement), evidentiality (il paraît que vs. il semble que), reportative conditionals (Le ministre serait malade), ‘announcers’ (certes), speech act modifiers (sérieusement, entre nous), connectors (donc, puisque, mais) and clefts.
Chapter 5 tackles the text level, discussing how polyphony works in short passages (two or three sentences). The issues dealt with include the potential problem of multiple mentions of the same discourse entities within and across clauses, and the potential proliferation of ‘voices’ that may arise in a sequence of constructions each of which has its own set of points of view. To deal with the problem, Nølke develops rules that build on the distinction between utterance speaker and text speaker. Another interesting issue at text level is reported discourse, arguably the most obvious instance of polyphony. A straightforward four-way distinction is made between direct vs. indirect speech, and discourse with vs. without an introductory expression. All are described in terms of speaker vs. points of view. Another subsection is dedicated to irony. Nølke does not aim for a unitary analysis of its many forms and manifestations, but discusses ways in which ScaPoLine could make a contribution, in particular with the notion of a new ‘speaker version’ and by marshalling the notions of point of view, text speaker and utterance speaker.
A final chapter briefly outlines possibilities for further research. The desirability of more cross-linguistic studies and cross-disciplinary work involving cognitive science and psychology is highlighted, as well as the modularity of polyphony. The idea that polyphony operates in conjunction with other linguistic modules is not really pursued in the book under review, but is asserted as early as chapter 1.
By and large, Nølke’s book fulfils its aim of presenting ScaPoLine and polyphony theory to an English-speaking audience. The author strives to be maximally clear in his terminology, providing tables that help the reader along. One question that arises, not just with respect to ScaPoLine but also to other branches of linguistique de l’énonciation, is about the relationship with formal semantics. Since the early 1970s, protagonists of linguistique de l’énonciation like Ducrot and eventually Nølke himself have explicitly construed their framework as an alternative to logical semantics. Nølke stresses his indebtedness to Ducrot and, while placing greater emphasis on formalization, sees his own work as entirely compatible with Ducrot’s. He talks about his frustration with logical semantics and how much of a revelation reading Ducrot has been for him in the 1970s. There has been relatively little dialogue since that time between the two traditions, possibly because linguistique de l’énonciation is mostly published in French. This book intends to at least partly remedy this situation. But would a young semanticist interested in phenomena like reported speech or the semantics of mood feel the same today? Is it not rather the case that in today’s research landscape there is a wealth of work on precisely those phenomena, whereas linguistique de l’énonciation has moved a little more slowly? I do not wish to downplay the insights of linguistique de l’énonciation (including ScaPoLine) or their continued relevance; however, more dialogue between linguistique de l’énonciation and formal semantics, or at least some mutual reception, would be more than welcome.
