The Grammar of Emphasis

From Information Structure to the Expressive Dimension

Andreas Trotzke

• Publisher: De Gruyter Mouton
• Year: 2017
• Print ISBN: 9781501515033
• eBook ISBN: 9781501505867
• Edition: 1
• Topic: Languages & Linguistics
• Subtopic: Grammar & Punctuation
• Index: Languages & Linguistics

This book reconsiders the linguistic notion of emphasis. For many, the concept of emphasis is confined to information structure. However, our understanding of the grammatical reflexes of emphasis is only partial as long as the expressive side of utterances is not taken into account. The book explores similarities, differences, and interactions between information structure and the expressive dimension of language in the domain of natural language grammar.

Specifically, this monograph demonstrates that specific word order options, sometimes in combination with discourse particles, yield meaning effects that are typical for the expressive side of utterances and endow them with an exclamative flavor. Approaching this issue from a syntactic point of view, the book shows that there are syntactic categories (e.g., a certain class of particle verbs) and word orders (e.g., certain fronting patterns involving discourse particles) that directly connect to expressive meaning components. The work presented in this monograph combines theoretical analysis with experimental evidence from both perception and production studies.

IIEmpirical domains of emphasis

3Particle topicalization and emphasis for intensity

In the preceding chapter, we saw that the pragmatic effects of marked word order can sometimes be analyzed as conventional implicatures expressing a certain type of emphasis that is distinguished from information structural meaning. The relevant cases of marked word order can be used in contexts where the speaker wants to express his excitement (be it based on surprise or on annoyance as in the reprimand use). I argued that this kind of emphasis can be captured by the notion of emphasis for intensity, which refers to the speaker’s pointing to an extreme end on a scale that is not based on truthfulness. By this pointing, the speaker conveys his excitement concerning the situation that p conveys.

In this section, I provide a comprehensive account of the phenomenon of topicalizing verb particles in German. Section 3.1 critiques previous accounts of particle topicalization and discusses some properties of particles that are potentially relevant to topicalization. I conclude that a particle’s ability to be topicalized depends more on its ability to be contrasted with other particles than on the semantic autonomy of the particle per se. Section 3.2 discusses unexpected cases of non-contrastable particles in the left periphery. In this context, I introduce a previously unnoticed phenomenon in which particle verbs denoting strongly emotionally evaluated situations allow their particles to be topicalized, even if the particle does not receive a contrastive interpretation. Section 3.3 explores to what extent all the distinctions between different particle verb classes discussed in the previous sections must be represented in the grammar.

3.1Particle verbs: Topicalization and semantic transparency

In the literature on particle verb constructions, it is generally claimed that at least two classes of particle verbs exist: transparent (i.e., compositional) and non-transparent (i.e., non-compositional, ‘idiomatic’) particle verbs (see McIntyre 2015 for a recent overview). It is not clear whether the class often referred to as ‘aspectual particle verbs’ can be subsumed under one of the two classes (cf. Jackendoff 2002a for treating aspectual cases as transparent; and Wurmbrand 2000 for a different proposal). In what follows, I propose a new classification of particle verbs that provides a more fine-grained notion of semantic transparency. I argue that the structural analyses given in the literature cannot account for the cases I will point out (Section 3.1.1). After establishing this new classification, I will report a questionnaire study that tested the acceptability of topicalization patterns of particle verb constructions discussed in the literature (Section 3.1.2).

3.1.1Semantic transparency: A new classification

When looking at the transparency of particle verb constructions from a syntactic perspective, we observe that some rules of syntax are not sensitive to the divide between transparent and non-transparent particle verbs, e.g., particle shift in English. In the following examples, the particles out and up differ regarding their semantic autonomy. While out in (1) has a clear spatial-directional content and thus forms a transparent configuration with the verb to kick, the combination of the verb to call and the particle up in (2) results in a non-transparent particle verb: the particle up does not bear any meaning of its own, and hence the meaning of the particle verb is not composed of the meanings of the verb and the particle.

However, other processes are sensitive to the transparency distinction. Observe the following cases of particle fronting in English (cf. Jackendoff 2002a: 75):

(3)	a.	Up marched the sergeant.
	b.	*Up blew the building.

In (3a), according to Jackendoff, the directional particle up is part of a non-idiomatic, transparent configuration and is thus licit in the locative inversion construction. In contrast, the idiomatic particle up in (3b) lacks the directional semantics and cannot be fronted.

As has been pointed out notably by both Wurmbrand (2000) and Zeller (2001), a similar situation obtains in German. In particular, while V2 is not sensitive to the transparency divide (4), topicalization of only the particle is.14 Compare the paradigms in (4) and (5) featuring the transparent aufmachen and the non-transparent aufhören (cf. Zeller 2001: 89–90).

While in both aufmachen and aufhören the verb can move to C (stranding the particle and thus yielding a discontinuous structure, cf. [4b,d]), topicalization may only target auf in aufmachen (5a) and not auf in aufhören (5b).

The phenomenon of particle topicalization has been extensively discussed in the literature on present-day Germanic, including English (cf. Dehé 2015 for an overview; and Müller 2002: 263–280; Zeller 2003 for relevant proposals). The possibility of fronting the particle alone plays a central role in the long-standing debate on the component responsible for particle-verb formation. Although particles are separable and thus have phrasal properties in the syntax, they feature properties of morphological units at the same time (e.g., Olsen 1997). Therefore, some accounts treat particle verbs as compound-like entities, given phenomena showing that particle verbs can feed morphological processes (e.g., taker-out-er); see McIntyre (2015) for an overview. In what follows, I focus on the syntactic mobility of the particle and, assuming that sub-constituents of syntactic atoms cannot be targeted by rules of syntax, I adopt an analysis above the word level.

Crucially, topicalization patterns make clear that a distinction between transparent and non-transparent particle verbs has important reflexes in the grammar. The theories of particle verbs elaborated by Wurmbrand (2000) and Zeller (2001) place a premium on this observation. However, they differ in the way they represent the transparency spectrum in the grammar. On the one hand, Wurmbrand (2000) argues that the two classes are licensed in different structural configurations. On the other hand, Zeller (2001) postulates a uniform syntactic structure, relegating the transparency divide to constraints located at L(ogical)F(orm), the component of semantic interpretation.

According to Zeller (2001: 127), particles are heads of non-functional phrasal complements. That is, the particle projects a phrase (and is hence in principle mobile), as can be seen in (6).

Under this approach, both transparent and non-transparent particle verbs are represented in the same way.

Whereas Zeller accounts for the transparency divide by postulating a uniform phrasal representation in combination with a constraint that is relegated to the semantic component (i.e., LF), Wurmbrand (2000) proposes two different syntactic structures for transparent and non-transparent particle verbs. She argues that not all particle verbs instantiate the same structure: ‘transparent’ particle verbs are structurally different from ‘(semi)idiomatic’ ones, according to her terminology. In particular, she proposes a structure such as (7a) for particle verbs like aufmachen and a configuration such as (7b) for cases like aufhören....

Table of contents

• Cover
• Title
• Copyright
• Preface
• Contents
• I The notion of emphasis
• II Empirical domains of emphasis
• III A unified account of emphasis
• References
• Appendix A
• Appendix B
• Appendix C
• Appendix D
• Index