Phonological Deletion in Yes/No Question Fragments



Dovnload 214.09 Kb.
Pagina1/3
Datum27.09.2016
Grootte214.09 Kb.
  1   2   3

Karen De Clercq Phonological Deletion in Yes/ No Question Fragments

CRISSP/ Catholic University of Brussels TABU Groningen, 8 June 2007




Phonological Deletion in Yes/No Question Fragments

Karen De Clercq

CRISSP/ Catholic University of Brussels



Outline

1. Introduction

2. Evidence for Ellipsis

3. Evidence for Phonological Deletion

4. Conclusion





  1. Introduction




  • Yes/No Question Fragment (YNQF)

An incomplete sentence that is nonetheless interpreted as a full yes/no-question,

i.e. a question that can be answered with yes or no.





  • Discourse Initial Context (DIC)

A situation in which the utterance of the yes/no question is initial,

i.e. where nothing has been said previously (i.e. Stainton’s DIlang).


(1) (Situation: A man sits at the bar and is slowly getting drunk.

Pointing at his empty glass, the landlady asks:)

Nog een pintje?

still a beer

‘Another beer?’




Claim 1:

The sentence in (2) is derived from an underlying syntactic structure

like the one in (3) through a process of ellipsis, as shown in (4).


(5) Wil je nog een pintje?

want you.sg. still an beer

‘Do you want another beer?’


(6) Wil je nog een pintje?

want you.sg. still a beer



‘Do you want another beer?


  • Alternative to claim 1: Direct Interpretation Approach (Stainton 1995)

(7) is a mere utterance which forces the hearer to directly interpret the missing part.

→ sufficient evidence to refute this hypothesis in my case (e.g. connectivity effects; see 2.1 below)

→ only a few kinds of utterances are candidates for a direct interpretation approach (e.g. (4))


(8) Greetings: Hello. Good bye. (Merchant 2004)


  • If there is underlying syntax in YNQF, then the next question is related to the kind of deletion that takes place: is it syntactically licensed deletion or phonological deletion?

(9) a. Wil jij een koekje?



  1. Wil jij een koekje?

  2. Wil jij een koekje?

  3. Wil jij een koekje?




Claim 2:

The deletion in (10) is a phonological phenomenon, rather than a syntactically licensed phenomenon.





  1. Evidence for Ellipsis (CLAIM 1)



Overview:

    1. Connectivity effects

      1. Case-matching

      2. Binding

      3. Tag Questions

      4. Control

    2. Questionnaire

      1. Set up

      2. Results



2.1. Connectivity effects




= Grammatical dependencies which are similar to the dependencies manifest in non-elliptical sentences.


      1. Case- matching

(11) (A man and a woman are sitting in a restaurant. They’ve had their main dish and they call the waiter to come over to their table. The waiter asks: )


a. * Noch ein Wunsch? (German)

still a.nom wish?

‘Anything else?’

b. Noch einen Wunsch?



still a.acc wish?

‘Anything else?’
c. Haben Sie noch einen/ * ein Wunsch?

have you still a.acc/ a.nom wish?

‘Would you like anything else?’



      1. Binding

 Reciprocals




  • Reciprocals require a plural antecedent:

(12) a. * Heb jij elkaar gekust?

have you.sg. each other kissed

Have you kissed each other?’
b. * Heeft Piet elkaar gekust?

has Piet each other kissed?

‘Has Piet kissed each other?’

c. *Hebben elkaar gedanst?

have each other danced

‘Have they danced with each other?’


d. Hebben ziji/ de jongensi/ [Piet en Marie] i elkaari gekust?

have they/ the boys/ Piet and Mary each other kissed

‘Have they/the boys/ Piet and Mary kissed each other?’

(13) (Situation: Carnival in Aalst: two friends meet each other on Sunday evening before going out. One has black make-up on, the other white make–up. A girlfriend enters and sees that the black guy has some white stains and the white guy some black ones. She bursts into laughter and asks:)

Elkaar gekust?

each other kissed

‘Kissed each other?’


(14) Hebben julliei elkaari gekust?

have you each other kissed?



Have you kissed each other?’



  • Anaphoric adverbs




    • Allemaal ‘all’ and allebei ‘both’ require a plural antecedent.

(15) a. Wil(len) * je/ * hij/ julliei/ zei allemaali chocomelk?



want you.sg he you.pl they all chocolate.milk?

Do *you/ *he/ you/ they all want chocolate milk?’
b. Willen julliei/ zei/ * hij/ * je allebeii chocomelk?

Want you.pl they he you. sg both chocolate milk?

Do you/ they/ *he/ *you both want chocolate milk?’
(16) a. Allemaal chocomelk?

all chocolate.milk

‘All of you chocolate milk?’
b. Allebei chocomelk?

both chocolate.milk

Both of you chocolate milk?’
(17) a. Willen julliei allemaali chocolademelk?

want you all chocolate.milk

‘Do you all want chocolate milk?
b. Willen julliei allebeii chocolademelk?

Want you both chocolate.milk



Do you both want chocolate milk?’


      1. Question Tags (Napoli 1982:90)

(18) a. It is raining, isn’t it/ *he?

b He will come, won’t he/ *they/ *she?

(19) a. Can’t sing a note, can he?

b. He can’t sing a note, can he?

→ (20a) must have a matrix subject at some point in the derivaton which is coreferential with the he in the tag.





      1. Control




  • Infinitival subjects require an antecedent:

(21) a. * Het is waarschijnlijk om PRO te vertrekken.

it is probable for to leave

*‘It is probable to leave.’
b. Mariei is klaar om PROi/*j te vertrekken.

Marie is ready for to leave



Marie is ready to leave.’

(22) (Situation: A man calls his wife who is dressing in the bathroom:)

Klaar om te vertrekken?

ready for to go

‘Ready to go?’


(23) Ben jei klaar om PROi te vertrekken?

are you ready for to go

‘Are you ready to go?’




Summary

The connectivity effects support the claim that the sentence in (24) is derived from an underlying syntactic structure like the one in (25) through a process of ellipsis, as shown in (26) .


(27) (Situation: A man sits at the bar and is slowly getting drunk.

Pointing at his empty glass, the landlady asks: )

Nog een pintje?



still a beer

‘Another beer?’


(28) Wil je nog een pintje?

want you.sing. still an beer

‘Do you want another beer?’
(29) Wil je nog een pintje?

want you.sing. still a beer



‘Do you want another beer?
All this evidence can be further supported by the results of the questionnaire I set up.



    1. questionnaire




      1. Setup




  • 24 informants

  • Scores from 1-4

1 absolutely acceptable 2 acceptable 3 hardly acceptable 4 ungrammatical

  • One ungrammatical fragment to test informants

  • Informants were asked to write down the interpretation they assigned to the stimulus fragments




      1. Results of questionnaire

 Acceptability



  • See page 6

  • column B and C were counted up = acceptable fragments

  • column E and F were counted up = unacceptable fragments

  • Fragments are ordered according to their acceptability/ grammaticality

  • Details: see page 6.

 Interpretations




  • Further testing could exclude some problems related to the interpretation of certain scores.

  • What can be deleted?

1st person pronouns

2nd person pronouns



er ‘there’

copular zijn ‘be’

auxiliaries zullen ‘shall’, moeten ‘must’, willen ‘want’, wensen ‘wish’

the motion verb gaan ‘go’




Summary

Not only semantically null auxiliaries, like have and be, can be deleted, but also semantically contentful verbs and auxiliaries can be deleted.





  1   2   3


De database wordt beschermd door het auteursrecht ©opleid.info 2019
stuur bericht

    Hoofdpagina