Over-arching research interest Impact of optionality in natural - PowerPoint PPT Presentation

Over-arching research interest • Impact of optionality in natural human language Coreference and focus in • Why? Ubiquitous and linguistically universal human sentence processing • Applies at two key levels: Wind Cowles Language and Cognition Lab • Multiple syntactic structures possible for same information Department of Linguistics • Multiple referential forms possible for same referent At the heart of structural optionality: Information structure Formal instantiations of information structure • Lambrecht (1999) - Independent component within sentence grammar • Information structure is formally realized via: Syntax Information Structure Semantics • Prosody • Function that mediates between the form of utterances and the current • Specialized (morpho-)syntactic markers mental states of interlocutors • Ordering and positioning of syntactic constituents • Particular grammatical constructions • (Alternative, compatible approach: Jackendo ff (2002) - Interface between components within the grammar) • Certain choices among lexical options

Major categories: Topic and Focus Major categories: Topic and Focus What did the player do? What happened? The player dropped the ball. The player dropped the ball. Topic Focus Topic Focus Aboutness condition Informative part Aboutness condition Informative part File card metaphor “pragmatically non-recoverable” File card metaphor “pragmatically non-recoverable” Major categories: Topic and Focus Formal instantiations (generalized) Topic Focus What did the player drop? reduced pitch accents prominent pitch accents Prosody The player dropped the ball. Specialized e.g. - wa in Japanese e.g. only in English (morpho-)syntactic markers Topic Focus Ordering/positioning of early late Aboutness condition Informative part syntactic constituents File card metaphor “pragmatically non-recoverable” Particular grammatical fronting clefts constructions fuller forms of Certain choices among pronouns lexical options reference

Questions for Human Coreference Processing Different types of anaphoric expressions • How is the antecedent referent determined? uniquely type in focus activated familiar identifiable referential identifiable • What are the influential factors? this/that (indef) this • Contextual, antecedent features, anaphor features it that N the N a N this N N • What is the time course of this process? • Gundel et al. (1993) Interactions with prominent/focused antecedents Interactions with prominent/focused antecedents A boxer and a trainer entered the weight room. A boxer and a trainer entered the weight room. The boxer talked to the trainer. The boxer talked to the trainer. Prominence “preference” He ... He ... Pronominal Pronominal Pronominal co-reference to prominent The athlete ... ... was preparing for the The athlete ... Categorical-Def Categorical-Def antecedents leads to faster reading upcoming competition. Categorical-Dem That athlete ... Categorical-Dem That athlete ... times - even if they are unmentioned. (e.g. Cornish, Garnham, Cowles, The boxer ... The boxer ... Repeated Repeated Fossard & Andre, 2005)

Interactions with prominent/focused antecedents Interactions with prominent/focused antecedents A boxer and a trainer entered the weight room. A boxer and a runner entered the weight room. The boxer talked to the trainer. The boxer talked to the runner. Inverse Semantic Distance E ff ects Topic maintenance/Topic shift He ... He ... Pronominal Pronominal Categorical demonstrative forms serve Categorical co-reference to atypical The athlete ... The athlete ... Categorical-Def Categorical-Def to shift attention and thus seek exemplars in prominent syntactic Categorical-Dem That athlete ... Categorical-Dem That athlete ... antecedents than pronouns (Fossard, positions leads to faster reading times Garnham & Cowles, 2012) (e.g. Almor, 1999; Cowles & Garnham, The boxer ... The boxer ... Repeated Repeated 2005) Interactions with prominent/focused antecedents Typicality Effects A boxer and a runner entered the weight room. 1. The professor and her student arranged the transportation for their field trip. The boxer talked to the runner. 2. She rented a car/boat for the second stage of the trip. 3. The vehicle was necessary for getting to the exploration site. e.g. Garrod & Sanford (1977) Repeated Name Penalty He ... Pronominal Repeated co-reference to prominent The athlete ... Categorical-Def antecedents leads to slower reading Categorical-Dem That athlete ... times (e.g. Gordon, Grosz & Gilliom, 1993; Gordon et al. 1999; Almor, 1999) The boxer ... Repeated but not always (Cowles & Dawidzuik, in press)

Inverse Typicality Effects Beyond Typicality and Clefts 1. The professor and her student arranged the transportation for their field trip. • Cowles & Garnham (2005) 2a. It was the student that rented the car/boat. 2b. What the student rented was the car/boat. • Two Experiments Tested Conceptual and Inverse Conceptual Distance E ff ects 3. The vehicle was necessary for getting to the exploration site. Almor (1999) • Exp 1: Clefts • Exp 2: No Clefts • 28 native English participants, 24 items Exp 1: Design & Sample Materials Methods - Procedure Distant Antecedent • Self-paced reading Most Specific Close Antecedent Anaphor Least Specific Focus Antecedent Setup: cobra. READY It was the mongoose that stood up to the cobra. READY The reptile hissed and got ready to strike. What the mongoose stood up to was the snake. Non-focus Antecedent Setup: cobra. It was the mongoose that stood up to the snake. Target sentence: The reptile hissed and got ready to strike.

Results - Reading time at the anaphor Results - Residual time at the predicate The reptile hissed and got ready to strike. The reptile hissed and got ready to strike. 800 400 Near Near (snake) (snake) 700 300 Distant Distant (cobra) (cobra) Reading Time for Subject NP (msec) 600 200 584 574 561 145 542 500 100 Residual Times 0 400 23 -23 -139 300 -100 -200 200 -300 100 -400 0 Non-Focus Focus Non-Focus Focus • 28 native English participants, 24 items • 28 native English participants, 24 items Experiment 2: Sample Materials Results - Reading time at the anaphor The reptile hissed and got ready to strike. Distant Antecedent Most Specific 800 Close Antecedent Near (snake) Anaphor Least Specific 700 Distant (cobra) 600 623 Reading Time for Subject NP (msec) 619 597 Focus Antecedent Setup: 576 500 The snake/cobra frightened the hunter. 400 Non-focus Antecedent Setup: 300 The hunter was frightened by the snake/cobra 200 Target sentence: 100 The reptile looked ready to strike at once if threatened. 0 Non-Focus Focus

Results - Residual time at the predicate Beyond Typicality and Clefts The reptile hissed and got ready to strike. • Inverse semantic distance e ff ects are not limited to clefts or specific to typicality 400 Near (snake) 300 Distant • What does this mean? (cobra) Reading Time for Subject NP (msec) 200 • Janus (Garnham and Cowles, 2008) 100 0 19 16 • Anaphoric processing has two aspects: -24 -6 -100 • Antecedent identification -200 -300 • Discourse structuring (cf. Marlsen-Wilson et al., 1982; Vonk, Hustinx & Simons, 1992) -400 Non-Focus Focus