December 8, 2014

Research Groups: Friday, December 12

9:30 AM - 11:00 AM
Language Variation and Change Group
Marisa Brook: "Syntactic categories informing variationist analysis: The case of copy-raising."

11:00 AM - 12:30 PM
Syntax/Semantics Group
Rebecca Woods (University of York): "Embedded inverted interrogatives as embedded speech acts."

Interest in the syntacticization of discourse and perspective has been reignited in recent years (Speas and Tenny 2003, Miyagawa 2012, Sundaresan 2012, Krifka 2014 inter alia). Much of the literature has focused on indexical shifting and other speaker-orientation phenomena. This paper offers a different perspective, examining Embedded Inverted Interrogatives (henceforth EIIs), a form of presenting reported speech common in some English dialects. In proposing a Speech Act projection above the embedded CP in EIIs, predictions about how such structures affect syntactic dependencies will be explored and issues regarding the Question Under Discussion will be touched upon. 

EIIs are present in a number of English dialects, notably Belfast English (Henry 1995), other Irish Englishes (McCloskey 1992, 2006) and African American English (Green 2002). Examples are as below:

1. I asked Jack was she in his (Jack's) class (McCloskey 2006, Irish English)
2. They don't ask you did you sit on the choir (Green 2002, African American English)
3. Do you remember did she(i) say could she(i) come? (Attested, North West English)

They are used to report speech and share certain characteristics with both direct and indirect speech. It is clear to see that EIIs are embedded within matrix sentences in that person and time is anchored to the utterance context rather than the reported context. Semantically they also play similar roles to indirect speech as they often represent speech acts that have not actually occurred – they can 'pre'-present as well as represent speech. However, there are also clear signs that they are in some way "sectioned off" from the matrix clause in a similar way to direct speech, in particular because they form strong islands:

4. a. *[Which book](i) did Dave ask "Should I read [t](i)?"
b. ? [Which book](i) did Dave ask if he should read [t](i)?
c. * [Which book](i) did Dave ask should he read [t](i)?

Furthermore, their use is more similar to direct speech as they are used to convey commitments to propositions and emotions as expressed by the original speaker, but without any commitment by the reporting speaker to convey a full precise quotation; that is, some of the original structure (such as wh- and T-to-C movement) and lexical items (such as epithets) may remain, but not all will.

Following the spirit of Speas and Tenny (2003) and Sundaresan (2012), it is proposed that the embedding verb in EIIs selects a Speech Act Phrase (SAP) which is the highest projection in the embedded clause. The SAP head specifies the illocutionary force of the embedded clause and interacts with the embedded Force head. In the specifier of the SAP is a DP which is minimally specified in terms of its features and is controlled by the matrix addressee, as this is the 'centre of evaluation' for the question contained in the EII. This DP constitutes the perspective being evaluated in the embedded speech act. This structure, for the sentence "Mary asked John did he dance", is in (5) (featural specifications are in < >, underlining marks values determined by Agree, and Control is indicated by {}):

5. [TP [DP Mary] [T° ] [vP ask [DP John] [VP task [SAP [DP [D,{JOHN}]] [SA° ] [ForceP [Force , [IntP did [TP he dance...]]]]]]

The full extended CP below the SAP is licensed by the relationship between the Speech Act head and Force. Note that Force is not underspecified for illocutionary force, as in typical embedded clauses, because its Force feature is valued by the Speech Act head (cf. Coniglio & Zegrean 2012). This relationship between the Speech Act head and Force is important both for conceptualising the difference between an embedded speech act and a typical embedded clause and for mediating the relationship between clause type and certain discourse elements (e.g. discourse particles, as in Hill 2007).

In terms of the syntactic consequences of the SAP, the embedded CP is selected by a functional head permitting adjunction to CP of speech act adverbs and other adjuncts (cf. McCloskey 2006) and leaving C open to permit subject-auxiliary inversion, as it is not specified by the lexical verb (having not been selected by it). Secondly, long-distance wh-extraction is not possible because, if the head of SAP is taken to be the head of the Phase (via transfer of phasehood from C), then material in Spec,CP is always passed to the interfaces with the lower clause. Finally, assuming that the orientation of speech act items such as adverbs like 'frankly' is determined syntactically via control by Spec,SAP, the presence of the SAP explains why adverbs like 'frankly', which are generally considered speaker-oriented, orient to the matrix-internal argument instead when embedded in an SAP.

The blocking nature of the SAP will be then illustrated using data collected in a pilot study of 73 English speaking adults and 5 English speaking children. The data at this stage shows that extraction over embedded questions is possible in child English and, in certain circumstances, adult English also. There are also clear differences between children's comprehension of embedded polar and embedded wh-questions which will be discussed. Finally, predictions for the acquisition path in children acquiring dialects with EIIs will be presented.

Selected references: Coniglio, M. & I. Zegrean (2012). Splitting up Force. In: L. Aelbrecht et al., Main Clause Phenomena. Krifka, M. (2014). Embedding Illocutionary Acts. In: T. Roeper & M. Speas, Recursion. McCloskey, J. (2006). Questions and Questioning in a local English. In: R. Zanuttini et al., Negation, tense and clausal architecture. Configurational Properties of Point of View Roles. In: AM. Di Sciullo, Asymmetry of Grammar. Sundaresan, S. (2012). Context and (Co)reference in the syntax and its interfaces. PhD dissertation, Stuttgart/Tromsø.

No comments:

Post a Comment