Scalarity has enjoyed long-standing attention in the semantic and pragmatic literature partly because the notion is both abstract and flexible: it could refer to measurement along different dimensions, but it can also be used for pragmatic purposes, including the expression of noteworthiness, importance, politeness, and (speaker) commitment. Against this background, I was drawn to Osamu Sawada’s recent monograph, Pragmatic Aspects of Scalar Modifiers, which stands out as a piece of work that fits neatly in a line of research exploring the notion of scalarity at the semantics–pragmatics interface.

The main, general phenomena that Sawada investigates are presented in the ‘Introduction’ chapter. In a concise and organized fashion, Sawada introduces what he calls the ‘dual-use phenomenon’ of pragmatic scalar modifiers, which is crucially based on the observation that degree morphology (i.e. morphological markings of degree expressions) used to express at-issue scalar meanings can also be used to express a not-at-issue scalar meaning. A sentence from Japanese (from page 3) serves as an example:

Here, the sentence has a reading which is at-issue, expressing that the degree of fun of tennis is greater than that of alternative individuals/entities. But there is also another reading, which has a not-at-issue scalar meaning, expressing that the utterance that tennis is fun is more noteworthy than any alternative utterance. This not-at-issue meaning, which is speaker-oriented and thus, at discourse level, is pervasive in scalar expressions of natural language. The ‘dual-use’ phenomenon, which Sawada speculates to be due to the abstractness and flexibility of the notion of scalarity, provides motivation for Sawada’s approach to separate the semantic and pragmatic dimensions of scalar modifiers.

To foreshadow the forthcoming chapters, in Chapter 1, Sawada also outlines his multidimensional analysis of scalar modifiers, examining the systematic similarities as well as differences between the semantic aspects and the pragmatic ones. Chapter 2 then presents the landscape of scalar meanings in the literature, beginning with the classical Gricean approach (Grice Reference Grice, Cole and Morgan1975) of distinguishing the at-issue meaning and conventional implicature (CI). For the purpose of the current study, Sawada introduces four relevant types of scalar meanings: at-issue meanings; conversational scalar meanings; presuppositional scalar meanings, defined according to Karttunen (Reference Karttunen1974); and conventional implicatures (CIs), which are the main focus of the book. While I find this part of the discussion very detailed, at various points later on in the book, the author devotes much space carefully applying the classical tests to defend his position that a certain meaning is CI rather than other not-at-issue meanings, which, to me, may feel a bit repetitive at times.

The formal mechanism which Sawada uses to analyze the dual-use phenomenon and pragmatic properties of scalar modifiers, i.e. the logic of CIs, is laid out in Chapter 3. This formal system consists of a type-driven system, which assumes two basic types: an at-issue type and a CI type in natural language, each used for different meanings while ensuring that the two kinds of scalar meanings are logically and dimensionally independent of each other. Although I was previously unfamiliar with the logic of CIs, Sawada makes various interpretive rules comprehensible to the reader, such as CI application as inherited from Potts (Reference Potts2005), mixed application, and expressive application (e.g. McCready Reference McCready2010, Sawada Reference Sawada2013).

Having presented the theoretical landscape and the formal tools, the next four chapters each offer an in-depth case study of pragmatic scalar modifiers. At this stage, Sawada makes clear a distinction in terms of the varieties of pragmatic scalar modifiers: some scalar modifiers are ‘high-level’ and some are ‘low-level’, as distinguished by their pragmatic functions. High-level pragmatic scalar modifiers convey how the speakers construe the at-issue proposition or speech act, with two case studies presented in Chapter 4 and Chapter 5. Chapter 4’s case study is of comparative constructions involving the Japanese indeterminate pronoun nani ‘what’, while Chapter 5 zooms in onto minimizer positive polarity items (PPIs), focusing on the Japanese minimizer chotto. By contrast, low-level pragmatic scalar modifiers express a speaker’s evaluation of the utterance situation, with the relevant cases discussed in Chapters 6 and 7. The organization of these four chapters is thoughtful, but is unfortunately not made clear to the reader until much later in the book.

Turning to the case study presented in Chapter 4, in Japanese, comparison with indeterminateness is expressed by an ‘indeterminate pronoun’ such as nani ‘what’ plus yori-mo ‘than-MO’; see example (1) above. An ambiguity arises if the indeterminate pronoun nani is used: as mentioned before, the sentence obtains two scalar meanings, one at-issue (individual reading), which compares the individual with alternative individuals, and one not-at-issue/CI (noteworthy reading), which compares the proposition with alternative propositions in terms of noteworthiness. Sawada’s analysis reduces these readings to a difference in the modification structures: nani-yori-mo attaches to the DegP in the individual reading as a result of scrambling, but it is base-generated in the sentence-initial adverb position and attaches to the entire sentence (i.e. a SpeechActP) in the noteworthy reading. Such an analysis offers a natural explanation for certain distributional facts: (i) the intensifier totemo ‘very/really’ and certain idiomatic expressions can appear only in the noteworthy reading, since they can only appear in an adjectival environment; and (ii) the individual reading cannot appear arise in negative sentences, which is naturally expected since nani-yori-mo as a free choice item cannot appear below negation.

What, then, is the relevant notion for encoding noteworthiness? This is a crucial question raised throughout the book, to which Sawada provides his tentative – and I think reasonably successful – proposal here. Sawada posits a scale of preference, with the meaning of noteworthiness derived from the interaction with the general pragmatic principle of ‘relevance to a goal’ (Roberts Reference Roberts2012); the comparative meaning is calculated at the level of CI, which is then combined with a speech act. The same degree morphology is being used to express two layers of meanings. Treating the at-issue scalar meaning as semantic and the CI reading as discourse-pragmatic, Sawada draws a parallel between the two meanings in terms of dimensionality and compositionality, i.e. that of scale structure and the role of scalarity, which derives the individual and noteworthy comparisons at semantic and discourse-pragmatic levels respectively – a very desirable outcome.

Turning now to low-level pragmatic scalar modifiers, Chapter 6 is a detailed investigation of two intensifiers, totemo ‘very’ and motto ‘still much more’; an illustrating example for each of them is provided below. Totemo also has a degree use, seen in (2a) below, plus an expressive use, in (2b): the degree use of totemo combines with a gradable predicate and intensifies the degree of the predicate at the at-issue level, while the expressive use of totemo, which Sawada argues to be a CI, combines with a negative (either epistemic or ability) modal predicate and intensifies the degree of unlikelihood or inability.

Similarly, the intensifier motto ‘still much more’ also has a degree reading in which the sentence in (3a) is interpreted as an ‘elliptical’ comparison, which involves a comparison between two individuals (and indicates that there is a large gap between the two) at the at-issue level, but at the same time, motto also has a ‘negative’ reading in (3b) which conveys the speaker’s attitude about the utterance situation.

These somewhat parallel observations suggest an interesting pattern: the not-at-issue uses of totemo and motto have to appear in negative environments, making them what Sawada calls ‘discourse-oriented negative polarity items’ (NPIs) – a concept that is both empirically and theoretically interesting. Sawada proposes that totemo can appear only in a negative environment because it presupposes that the maximum probability degree of a gradable modal predicate (e.g. can) is 0. Semantically, totemo combines with a gradable modal predicate and refuses to update the common ground with the at-issue proposition. Meanwhile, motto combines with an at-issue gradable predicate and conventionally implicates that the degree of the target in an expected situation is much greater than the target’s current degree. The negative implication triggered by motto, then, comes from the large gap between a current degree and an expected degree.

The last case studies of low-level pragmatic scalar modifiers are the counter-expectational intensifier yoppodo and the scale-reversal adverb kaette, discussed in Chapter 7.

While the preceding chapters investigate the interpretation of pragmatic scalar modifiers in matrix sentences, Chapter 8 examines how pragmatic scalar modifiers are interpreted in embedded environments. The most important characteristics of pragmatic scalar modifiers, as Sawada has shown so far, are that they are speaker-oriented and not-at-issue, logically independent of any semantic operators. Interestingly, however, in an embedded environment, a pragmatic scalar modifier can become subject-oriented and thus receive an at-issue meaning. This means that the CI scalar meanings can ‘project’ out of the complement of an attitude predicate (e.g. want), contrasting with presuppositions, which will typically be blocked in similar environments. Notably, when a pragmatic scalar modifier is embedded under an attitude predicate, the distinction between high-level and low-level scalar modifiers becomes immediately relevant: unlike high-level scalar modifiers, low-level scalar modifiers can project only when there is a speaker-oriented modal in the main clause. This shift of the level of scalar meaning, contrary to Potts’ (Reference Potts2005) claim about appositives and expressives being invariably speaker-oriented, is not just a matter of indexicality as Sawada argues, but a shift from a CI to a secondary at-issue entailment through what he calls ‘projection via modal support’. Compositionally, it is achieved via type shifting of a clause that is embedded under an attitude predicate, applied at the root level of the embedded clause. The clarity of this chapter depends on the technical analyses much more heavily than the clarity of the previous chapters.

From the synchronic typology, Sawada extends his investigation to the diachronic typology of pragmatic scalar modifiers in Chapter 9, by examining the relationship between at-issue and not-at-issue scalar meanings from a historical point of view. Finally, Chapter 10 concludes, with a comparison between the current analysis and alternative views, and outlining directions for future research.

Overall, Sawada’s work makes substantial empirical contribution to the understanding of the semantic and pragmatic meanings of scalar modifiers. Analyzed in a theoretical framework at the semantics–pragmatics interface, the proposal carefully separates the semantic and pragmatic dimensions of scalar modifiers, and draws a parallel between the two dimensions on the basis of synchronic and diachronic typology. The analytical aspect of the book is detailed and especially useful for readers who are familiar with the logic of CIs. Future editions should be free of typos and mistakes found in various examples and minimal pairs throughout the key chapters. Such errors can be confusing for readers (like myself) who are not familiar with the Japanese language.