Echoicity and contrast in Spanish conditionals Elena Castroviejo and - - PowerPoint PPT Presentation

Echoicity and contrast in Spanish conditionals

Elena Castroviejo and Laia Mayol

Ikerbasque and UPV/EHU — UPF

Non-At-Issue Meaning and Information Structure – University of Oslo May 9th, 2017

Introduction

The phenomenon

⊚ The conditional construction in Spanish, illustrated in (1-b): (1) A: Estoy be.1SG cansado. tired ‘I’m tired.’ B: Si if t´ u you est´ as be.2SG cansado, tired yo I estoy be.1SG muerto. dead ‘(lit.) If you are tired, I am exhausted.’ → We will call them Echoic contrastive conditionals (eccs for short).

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 2 / 35

Introduction

Research questions

⊚ Are they regular (hypothetical) conditionals?

⊕ What are their properties? ⊕ What is their distribution?

⊚ What is the rhetorical relation between antecedent (p) and consequent (q)? ⊚ What is the information structure of p and q? ⊚ What formal analysis can account for the specific behavior of eccs?

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 3 / 35

Introduction

Goals & Claims

⊚ Goal: Characterize eccs and provide an analysis that is compatible with current theories of conditionals in semantics. ⊚ Claims

⊕ eccs are biscuit conditionals in the sense of e.g. Siegel (2006). ⊕ The hypothetical relation between p and q is really established between the assertion of p and the assertion of q, two propositions that stand in a relation of contrast. ⊕ Spanish conditionals with Contrastive Topic marking and no causal or epistemic dependence between p and q, yield contrast, scalarity and echoicity.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 4 / 35

Introduction

This talk

1 Introduction 2 Background

Previous work Properties

3 Proposal

Conditional Discourse structure

4 Conclusions

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 5 / 35

Background Previous work

Previous mentions

Narbona (1990) mentions eccs when discussing conditionals in Spanish. (1) A: Estoy be.1SG cansado. tired ‘I’m tired.’ B: Si if t´ u you est´ as be.2SG cansado, tired yo I estoy be.1SG muerto. dead ‘(lit.) If you are tired, I am exhausted.’ ⊚ p echoes the previous utterance in the dialogue. ⊚ The biclausal structure is used to emphasize two opposed propositions.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 6 / 35

Background Previous work

Previous mentions

Schwenter (1999): the contrastive function of the construction is evidenced by the obligatoriness of the overt subject pronouns. (1) A: Estoy be.1SG cansado. tired ‘I’m tired.’ B: Si if #(t´ u) you est´ as be.2SG cansado, tired #(yo) I estoy be.1SG muerto. dead ‘(lit.) If you are tired, I am exhausted.’ (2) A: Estoy cansado. ‘I’m tired.’ B: Si if (#t´ u) you est´ as are cansado, tired acu´ estate lie down un a rato. while ‘If you are tired, lie down for a while.’

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 7 / 35

Background Properties

Properties 1 and 2

(3) A: Ayer Yesterday me 1SG.CL beb´ ı drank una a botella bottle de

• f

vino wine entera. whole ‘Yesterday I drank a whole bottle of wine.’ B: Si if t´ u you te 2SG.CL bebiste drank una a botella, bottle yo I me 1SG.CL beb´ ı drank dos. two ‘(lit.) If you drank a bottle, I drank two.’

1 Echoicity 2 Contrastivity

→ We follow Mayol (2010) in claiming that the overt occurrence of the pronoun in a null subject language flags Contrastive Topic (CT).

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 8 / 35

Background Properties

Property 3

3 Scalarity: the focused predicates in p and q stand in a scalar

relationship, such that the one in q is higher / stronger in some contextually relevant scale. (4) [In a textbook on prehistorical animals] Si los dinosaurios daban miedo, el tiranosaurio rex era directamente aterrador. ‘If dinosaurs were scary, the tyranossaurus rex was downright terrifying.’

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 9 / 35

Background Properties

Property 3

3 Scalarity: the focused predicates in p and q stand in a scalar

relationship, such that the one in q is higher / stronger in some contextually relevant scale. (5) B: Si t´ u est´ as cansado, yo estoy muerto. ‘(lit.) If you are tired, I am exhausted.’ B’:#Si t´ u est´ as cansado, yo estoy totalmente relajado. ‘If you are tired, I am totally relaxed.’

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 10 / 35

Background Properties

Property 4

4 Lack of conditional interpretation (≈ q is true in those worlds in which

p is also true, cf. Stalnaker 1979): p doesn’t seem to be a condition for the truth of q. (6) Si if llueve, rains ir´ e will.go.I al to the cine. movies ‘If it rains, I’ll go to the movies.’ (1) A: Estoy cansado. ‘I’m tired.’ B: Si t´ u est´ as cansado, yo estoy muerto. ‘If you are tired, I am exhausted.’

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 11 / 35

Background Properties

Property 5

5 Expression of disdain: they may convey that the speaker is

challenging the truth or the relevance of her interlocutor’s utterance.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 12 / 35

Background Properties

Alert! Apparent cases of ECCS.

⊛ These are not the same kind of animal . . . (7) Si if t´ u you no NEG vas, go.2SG ellos they vuelven. come back.3PL ‘If you don’t go [to vote], they will come back.’ → There is a conditional relation between p and q, no echoicity. → The pronoun is optional in p. (8) If that’s art, then I’m the Queen of England. → Called backhanded conditionals (Mu˜ noz 2013).

⊙ Imply ¬p (hyperbole). ⊙ q is agreed not to be true in the actual world (outlandishness).

→ Not necessarily echoic, no scalarity, epistemic dependence.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 13 / 35

Proposal

Interim summary and update on our goals

⊚ p and q in eccs are CT-marked. ⊚ eccs have the form of conditional sentences, but the meaning of q is apparently not conditional on the meaning of p. ⊚ p is systematically echoic. We need an analysis that . . .

1 finds out whether a conditional relation can be preserved in eccs. 2 explains why whenever there is Contrastive Topic, scalarity, echoicity

(and potentially disdain) arise.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 14 / 35

Proposal

Our analysis in a nutshell

Main idea ⊚ A conditional relation is established between two utterances: the assertion of q is relevant given the assertion of p by a previous discourse participant. ⊚ Both p and q address the same multiple wh-question, which is flagged through Contrastive Topic. The two propositions stand in a discourse relation of contrast. ⊚ q is interpreted as winning the argument in the contrast between p and q. This coincides with the stronger scalar term.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 15 / 35

Proposal Conditional

Not a hypothetical conditional

⊚ Biscuit conditionals (Siegel 2006, Predelli 2009, Franke 2007a,b). ⊚ Also relevance or speech act conditionals. ⊚ They do not state when q is true (there is no causal or epistemical dependence between p and q). ⊚ They state when it is appropriate to utter q. (9) a. If you are hungry, (#then) there is pizza in the fridge. b. If you need anything else later, (#then) my name is James. (10) a. If you are hungry, ∃a∃p [a is an assertion of p ∧ p = ‘there’s pizza in the fridge’] b. If you are hungry, there is a relevant assertion that there’s pizza in the fridge.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 16 / 35

Proposal Conditional

Not a hypothetical conditional

⊚ In eccs there is a lack of causal or epistemic dependence between p and q (Prop. 4). q seems to be entailed. ⊚ Whenever entonces ‘then’ is included, a causal or epistemic dependence arises, (11) (no ecc reading). (11) a. Si t´ u est´ as cansado, (#entonces) yo estoy muerto. ‘If you are tired, (#then) I am exhausted.’ b. Si los dinosaurios daban miedo, (#entonces) el tiranosaurio rex era directamente aterrador. ‘If dinosaurs were scary, (#then) the tyranossaurus rex was downright terrifying.’

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 17 / 35

Proposal Conditional

Not a hypothetical conditional

A subtype of biscuit? ⊚ If the lack of apparent dependence between p and q is key to determining the kind of conditional, then eccs are biscuits. (12) In all worlds where p has been asserted, there is a relevant assertion that q. ⊚ We still need to understand why echoicity (+ taking the assertion of p as the antecedent) and contrast go hand in hand.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 18 / 35

Proposal Discourse structure

QUD

⊚ QUD-model of discourse (Roberts 1996, Ginzburg 1994): An utterance U, uttered in a context C1, addresses the most salient qud (qud-maxC1) and introduces a qud which becomes the most salient in the output context (qud-maxC2) (Vallduv´ ı 2016). (13) A: [qud-maxC1:] What are we having for dinner? B: Fish. [qud-maxC2:] Are we having fish for dinner? ⊚ ’Fish’ is a rheme, the actual update potential of an utterance, which is the locus of progression in discourse. ⊚ The theme is omitted.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 19 / 35

Proposal Discourse structure

When is the theme present?

⊚ “Theme-containing utterances prepare the input context by promoting a given qud to qud-maximality prior to being elaborated

• n” (Vallduv´

ı 2016). ⊚ When qud-maxC1 is split into several QUDs. (B¨ uring 2003) (14) How was the concert? a. The sound was awful. b. The audience was enthusiastic. c. The band was fantastic.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 20 / 35

Proposal Discourse structure

Contrastive Topic

(15) [qud-maxC1:] How is A feeling? A: Estoy cansado. ‘I’m tired.’ [qud-maxC2:] Is A tired? (16) B: Si t´ u est´ as cansado, yo estoy muerto. ‘If you are tired, I am exhausted.’ [qud-maxC3:] Who is feeling how, regarding tiredness? Who is feeling how tired? ⊚ CT indicates what QUD q is addressing: a multiple wh-question.

1 Focus-driven question: How are you feeling, regarding tiredness? 2 CT-driven question: Who is feeling how, regarding tiredness?

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 21 / 35

Proposal Discourse structure

Where does scalarity come from?

⊚ The discourse relation between the answers to the multiple wh-question can be either sequentiality, (17-b), or contrastivity, (17-c). (17) A: Who is feeling how, regarding tiredness? B1: HeCT is tired and ICT am exhausted. B2: HeCT is tired, but ICT am exhausted. ⊚ But contrastivity explains scalarity.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 22 / 35

Proposal Discourse structure

Contrast

⊚ According to a formal contrast approach (Sæbø 2003, Umbach 2005), as reported in Winterstein (2012) for but: (18) a. β = I; α = he; γ = exhausted; δ = tired b. CONTRAST conveys ¬γ(α) = he is not exhausted ⊚ On the scale <exhausted,tired>, a quantity implicature arises that he is tired and nothing more, he’s only tired.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 23 / 35

Proposal Discourse structure

Contrast

⊚ The second conjunct presents the stronger argument (Anscombre and Ducrot 1977). (19) a. The earrings are expensive, but pretty. [We’ll buy them.] b. The earrings are pretty, but expensive. [We won’t.] ⊚ As Tomioka (2016) puts it, Hurford’s Constraint does not only apply to disjunction. Contrastive Focus obeys HC. (20) a. Andy read some of the books, but Anna read all of them.

• b. ??Anna read all of the books, but Andy read some of them.

→ When two scale-mates are contrasted, the better order is the one in which the semantically stronger one follows the weaker one.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 24 / 35

Proposal Discourse structure

Contrast

⊚ If the scalar terms are reversed, a contradiction arises, (21):

⊕ If γ(β) = I am tired, then ¬γ(α) = he is not tired. ⊕ BUT “exhausted” ⊧ “tired”, so you cannot be “not tired” and “exhausted” at the same time. (21) #HeCT is exhausted, but ICT am tired.

⊚ The 2 opposed items in focus must be alternatives in a Common Integrator (here, belong to the same scale) (Sæbø 2003), (22). (22) #HeCT is tired. ICT am happy.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 25 / 35

Proposal Discourse structure

A decision problem

⊚ Communication takes places against the background of a decision problem of the hearer (What action should he carry out? What should he believe?) (Franke 2007a,b).

⊕ In biscuits, the speaker realizes that in all p worlds, the hearer has a decision problem. ⊕ p helps identify in what way q is relevant to decide for one action over the other. (23) a. There are biscuits in the cupboard. b. If you are hungry, there are biscuits in the cupboard.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 26 / 35

Proposal Discourse structure

A decision problem

⊚ In eccs, we claim, the speaker imposes her own decision problem.

⊕ She presents 2 possible answers to the CT-driven QUD (‘Who is feeling how, regarding tiredness?’). ⊕ By uttering the ecc she wants to resolve this QUD so as to settle her decision problem: Who wins the argument? (‘Who is more tired?’).

⊚ This competition does not arise when there is no contrast between p and q, even if there is CT marking, (24). (24) A: Estoy cansado. ‘I am tired.’ B: (Pues) part YoCT I estoy am muerto. dead ‘I am exhausted.’

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 27 / 35

Proposal Discourse structure

Back to biscuit conditionals

(11) In all worlds where p has been asserted, there is a relevant assertion that q. → What makes the assertion of q relevant / appropriate? ⊚ The assertion of q is justified by the previous assertion of p, because p and q are 2 different ways of answering the same QUD. ⊚ Given the previous assertion of p, the assertion of q can be interpreted as the answer that settles the decision problem, i.e. who wins the argument.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 28 / 35

Proposal Discourse structure

Echoicity

We reason how echoicity is derived, as follows: ⊚ The antecedent of the ecc is the utterance of p. ⊚ In placing p in the antecedent of a conditional, the speaker presupposes p (Isaacs and Rawlins 2008). ⊚ This implicates ¬ p, which is incompatible with BELspk(p). ∴ The author of the assertion of p must be someone else.

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 29 / 35

Proposal Discourse structure

Disdain

A function of 2 factors: a) By placing p in the antecedent of a conditional, the speaker may not be accepting p into the cg or the QUD resolved by p (as was meant by another discourse participant). b) γ is a better state of affairs than δ (γ is the stronger argument, δ falls short).

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 30 / 35

Conclusions

Conclusions

The speaker examines a previous utterance of p and imposes a decision problem: Who wins the argument? 1) The relevance of the assertion of q is motivated by her wish to settle the decision problem.

i) Lack of conditional dependence between p and q (Prop. 4)

2) She contrasts p with her own assertion that q.

i) p and q are CT marked (Prop. 2) ii) p is echoic (Prop. 1) iii) The focus in q is a stronger scalar term than the focus in p (Prop. 3)

3) She challenges the truth or relevance of p by placing it in the antecedent of a conditional.

i) Disdain (Prop. 5)

Castroviejo & Mayol (EHU & UPF) Echoicity and contrast in Spanish conditionals 05/09/2017 31 / 35

Echoicity and contrast in Spanish conditionals Thanks!

Elena Castroviejo Laia Mayol Ikerbasque and University of the Basque Country Universitat Pompeu Fabra (Ikerbasque & UPV/EHU) (UPF) elena.castroviejo@ehu.eus laia.mayol@upf.edu http://elena-castroviejo-miro.cat http://laiamayol.wordpress.com This research has been partially supported by project FFI2015-66732-P, funded by the Ministry of Economy and Competitiveness (MINECO) and the European Regional Development Fund (FEDER, UE), the IT769-13 Research Group (Basque Government), and UFI11/14 (University of the Basque Country, UPV/EHU).

References I

Anscombre, J.-C. and Ducrot, O.: 1977, Deux mais en fran¸ cais?, Lingua 43(1), 23–40. B¨ uring, D.: 2003, On D-Trees, Beans, and B-Accents, Linguistics & Philosophy 26(5), 511–545. Franke, M.: 2007a, Independence and decision-contexts for non-interference conditionals, in A. Benz, C. Ebert and R. van Rooij (eds), Proceedings of the ESSLLI 2007 Workshop on Language, Games and Evolution, pp. 9–16. Franke, M.: 2007b, The pragmatics of biscuit conditionals, in M. Aloni,

• P. Dekker and F. Roelofsen (eds), Proceedings of the 16th Amsterdam

Colloquium. Ginzburg, J.: 1994, An update semantics for dialogue, Proceedings of the first International Workshop on Computational Semantics, Tilburg, The Netherlands.

References II

Isaacs, J. and Rawlins, K.: 2008, Conditional questions, Journal of Semantics 25(3), 269–319. Mayol, L.: 2010, Contrastive pronouns in null-subject romance languages, Lingua 120(10), 2497–2514. Mu˜ noz, P.: 2013, That’s not art: backhanded sentences and distant worlds., Proceedings from the Annual Meeting of the Chicago Linguistic Society. Narbona, A.: 1990, Las subordinadas adverbiales impropias del espa˜ nol, ´ Agora, Malaga. Predelli, S.: 2009, Towards a semantics for biscuit conditionals, Philosophical Studies 142(3), 293–305. Roberts, C.: 1996, Information structure in discourse: Towards an integrated formal theory of pragmatics, OSUWPL Volume 49: Papers in Semantics pp. 91–136.

References III

Sæbø, K. J.: 2003, Presupposition and contrast: German aber as a topic particle, Proceedings of Sinn und Bedeutung, Vol. 7, pp. 257–271. Schwenter, S. A.: 1999, Pragmatics of Conditional Marking. Implicature, Scalarity, and Exclusivity, Outstanding Dissertations in Linguistics, Garland. Siegel, M. E.: 2006, Biscuit conditionals: Quantification over potential literal acts, Linguistics and Philosophy 9(2), 167–203. Stalnaker, R.: 1979, Assertion, Syntax and Semantics 9, 315–322. Tomioka, S.: 2016, Contrastiveness and scalar implicature. University College London Seminar. Umbach, C.: 2005, Contrast and information structure: A focus-based analysis of but, Linguistics 43(1), 207–232. Vallduv´ ı, E.: 2016, Information structure, Cambridge Handbook of Formal Semantics, Cambridge University Press, Cambdridge. Winterstein, G.: 2012, What but-sentences argue for: An argumentative analysis of but, Lingua 122(15), 1864–1885.