Universität zu Köln

Meijer, Anna Marlijn (2022). Propositional Anaphora: The case of embedded polar responses in Dutch and English. PhD thesis, Universität zu Köln.

Preview

PDF meijer_diss_2022.pdf Download (1MB) | Preview

Abstract

This dissertation investigates the embedded polar response paradigms of Dutch, English and to some extent of German. These are responses that involve an agent, a propositional attitude verb and an anaphor, such as the affirmative response ‘I think so’ to a question like ‘Did John feed the dog?’ (cf. Sailor 2012). In such responses, anaphors, like ‘so’, have been argued to refer to the proposition introduced in the preceding questions (see, e.g., Cushing 1972; Cornish 1992; Asher 1993; Needham 2012; Krifka 2013; Snider 2017). Unlike anaphors in for instance the nominal domain, propositional anaphors have not been studied extensively, with the exceptions of Cushing (1972); Webber (1991); Cornish (1992); Asher (1993); Hegarty et al. (2002), Snider (2017), Pasquereau (2018, 2022) and recent studies of response particles ‘yes’ and ‘no’ (see, e.g., Krifka, 2013; Brasoveanu et al., 2013; Roelofsen and Farkas, 2015; Claus et al., 2017; Goodhue and Wagner, 2018). The present work aims to fill this gap from the perspective of embedded polar responses. Note that the English embedded polar response paradigm consists of responses like ‘I think so’, ‘I think’ (using a null proform), or ‘I think not’ in combination with the predicate ‘think’. The response ‘I think it’ is considered infelicitous. In combination with, e.g., ‘doubt’, one can only response ‘I doubt it’ in response to a question, whereas ‘I doubt so’ or ‘I doubt’ would be infelicitous responses. This pattern raises the question why there are multiple embedded polar responses in the first place and why they are restricted to certain predicates. Furthermore, it raises the question if embedded polar response paradigms in other languages are similar. To shed light on these questions, the present dissertation investigates responses in Dutch, English and to some extent in German. One of the main conclusions of this dissertation is that there is no uniform class of anaphoric items used in embedded polar responses. I show that these anaphors, like ‘it’, ‘so’, or the null proform, are very different from one another in terms of their semantic contribution and their pragmatic role in their paradigms. Despite these differences, there are also similarities between the different responses considered and the different languages under consideration. This dissertation considers two categories: Type I and Type II responses. This dissertation first argues that the category of Type I responses consists of responses involving items like Dutch polar ‘van’ and English ‘so’. It argues that these responses signal that the proposition under reference has not been settled yet - because either the speaker is uncertain about it his/herself or other interlocutors disagree about the status of the proposition. As a consequence, both polar ‘van’ and ‘so’ cannot generally occur with factive predicates. Note that polar ‘van’ and ‘so’ signal their Type I meanings in different ways. Chapter 3 argues that the ‘uncertainty’ signaled by polar ‘van’ is due to the similative meaning hardwired into its semantics. In contrast, Chapter 4 argues that ‘so’ bears a presupposition with respect to the common ground status of its referent. More specifically, Chapter 4 argues that ‘so’ presupposes that its referent is still under discussion, i.e. on the Table in terms of Farkas and Bruce (2009), thereby building on Needham’s (2012) account. In addition, this dissertation shows that ‘so’ is in fact an adverb. These two properties set ‘so’ apart from the other anaphors considered in this thesis, although this anphor is often considered an exemplary propositional anaphor. In the first part of this dissertation we thus saw that embedded polar responses with polar ‘van’ and ‘so’ thus have rather similar functions, but a very different underlying semantics and pragmatics. This dissertation argues that the second category, Type II responses, consists of embedded polar responses that lack a presupposition or implication that signals uncertainty or non-settledness. This dissertation shows that this category involves embedded polar responses containing for instance weak pronouns, like ‘het’ or ‘it’, or the null proform. As a consequence, these kinds of responses compete with Type I responses. Whenever a speaker wishes to express that the proposition under reference is not yet settled, s/he will choose a Type I response over a Type II response. That is, an English speaker would prefer a response involving ‘so’ over a response involving a null proform in such a scenario. The Dutch counterpart would involve polar ‘van’ instead of ‘het’. Furthermore, Chapters 5 and Chapter 6 argue that there are differences between English and German, on the one hand, and Dutch, on the other, where it comes to Type II responses. Both English and German are able to form embedded polar responses with predicates like ‘think’ or ‘guess’ and the null complement anaphor (NCA) and are unable to form such responses with weak pronouns. For Dutch, this works the other way around: Dutch can form embedded polar responses with, e.g., ‘denken’ (think) and ‘het’ (it), but not with an NCA. Moreover, English and German may use propositional attitude verbs in combination with the negative adverbs ‘nicht’ and ‘not’ in embedded polar responses, whereas Dutch cannot. Chapter 5 shows that such responses with negative operators are most parsimoniously analyzed as involving NCA, just like their non-negative counterparts. On the basis of the languages under consideration this dissertation argues there seems to exist a correlation between (i) being able to form embedded polar responses with NCA and a negative adverb and (ii) being unable to form one with a weak pronoun and a predicate like think. Chapter 6 investigates Dutch embedded polar responses with weak pronouns in more detail. The main question of this chapter is why Dutch features such responses whereas English does not. Chapter 6 argues that this is the case, because ‘het’ is the phonologically weakest proform available in Dutch. Dutch does not feature an NCA. The competitors of ‘het’ are, e.g., the phonologically heavier demonstratives ‘dat’ (= that) and ‘dit’ (= this). These proforms, on their turn, compete with one another in terms of proxomity. Since the demonstratives are phonologically heavier than the weak proform, they are dispreferred for embedded polar responses, which target the most salient proposition in the discourse at the moment of answering. In addition, ‘het’ competes with polar ‘van’. As mentioned above, polar ‘van’ is chosen over ‘het’ if the speaker wishes to express a more subjective meaning. There are thus different types of competition at play: Competition in terms of subjectivity, phonological strength and proximity. The same competition applies to the Dutch forms. This dissertation shows that we cannot simply compare the licensing of one propositional anaphor with another without taking into consideration the individual meanings and uses of these proforms. In discussing different kinds of responses across languages, this dissertation also provides insight into the different discourse moves that constitute answers to polar questions. It shows that responses with for instance polar ‘van’ in Dutch or ‘so’ in English convey more and different information than those with for instance ‘het’ in Dutch or NCA in English. The former bear more information than just the information that the proposition under reference is compatible or not with the attitude holder’s information state, as they also provide information on the unsettledness or uncertainty of the proposition under reference. In addition, the present work sheds more light on propositional attitude verbs. It shows that Anand and Hacquard’s (2008, 2013) distinction between doxastic and assertive predicates is highly relevant when considering embedded polar responses (cf. Scheffler 2008) and that these predicates behave differently when occurring with different items in embedded polar responses. We saw that the use of polar ‘van’ in Dutch and ‘so’ in English is more flexible than that of other anaphors. Again, this can be attributed to the ‘special’ meaning of these anaphors and their evidential uses. This illustrates once more that the items used in embedded polar responses are not a uniform set.

Item Type:	Thesis (PhD thesis)
Creators:	Creators Email ORCID ORCID Put Code Meijer, Anna Marlijn UNSPECIFIED UNSPECIFIED UNSPECIFIED
URN:	urn:nbn:de:hbz:38-624467
Date:	2022
Language:	English
Faculty:	Faculty of Arts and Humanities
Divisions:	Faculty of Arts and Humanities > Fächergruppe 3: Deutsche Sprache und Literatur > Institut für Deutsche Sprache und Literatur I
Subjects:	Language, Linguistics
Uncontrolled Keywords:	Keywords Language linguistics UNSPECIFIED language UNSPECIFIED answers UNSPECIFIED embedded polar repsonses UNSPECIFIED anaphora UNSPECIFIED response paradigms UNSPECIFIED propositional anaphora UNSPECIFIED
Date of oral exam:	17 July 2020
Referee:	Name Academic Title Repp, Sophie Prof. dr. Krifka, Manfred Prof. dr. Hinterwimmer, Stefan PD dr.
Refereed:	Yes
URI:	http://kups.ub.uni-koeln.de/id/eprint/62446