Bookshelf 📖

SYNTACTIC AND NON-SYNTACTIC FACTORS IN REFLEXIVE PRONOUN RESOLUTION IN MANDARIN CHINESE (Lyn et al., 2023)

Citation

Lyn, Jyn., (2023). SYNTACTIC AND NON-SYNTACTIC FACTORS IN REFLEXIVE PRONOUN RESOLUTION IN MANDARIN CHINESE

My thoughts

My thoughts

The differences between logophoric (perspective-reporting), which might be the source, and empathic, which is perspective-taking, which might be the fear verb? - 我的想法：attitude holder算是 source / perciver吗？empathy locus可以是 empathy verb 怕吗？ - ziji can be both empathic and logophoric

Summary

The main goal of this dissertation is to use experiments to examine how syntactic factors (e.g., locality, syntactic position) and non-syntactic factors (e.g., verb semantics, perspectival information) impact reflexive resolution in Chinese and how these factors influence real-time processing, to re-examine conventional linguistic judgment, and more importantly, to provide new data contributing to the advances of linguistic theories.

The antecedent choice judgment results:

• ziji and ta-ziji are subject to syntactic, semantic, and discourse-level constraints, but the extent to which discourse topic prominence impacts the interpretation of ziji and ta-ziji differs.
• Ziji is more heavily influenced by the discourse topic status of the potential antecedent compared to ta-ziji (ziji is perspective-sensitive (empathic) while ta-ziji is not)

Self-paced reading results

• When discourse-level information is considered in the experimental design, syntax does not play a predominant role in antecedent retrieval
• more consistent with the standard retrieval model (but it can not predict the absence of the object blocking effect)

Abstract

A reflexive pronoun depends on another referent in the discourse for interpretation, and when multiple referents exist, the readers need to identify which referent the reflexive is to be anchored to.

Key Concepts

2 types of reflexive in Chinese

• morphologically simple/bare reflexive ziji (‘self’)
• morphonologically complex/compound ta-ziji (‘s/he-self’)

object blocking effect

• When a 1st person pronoun (sentence-external perspective center) in the object position cannot block long-distance binding by a discourse topic empathy locus (sentence-internal perspective center)

Introduction

• The referent that the reflexive anchors to is called the antecedent; the process of disambiguating the reflexive is called reflexive resolution
• The process of locating the antecedent is constrained not only by the human cognitive architecture but also by the linguistic properties of reflexives, so reflexive resolution is a bridge that brings cognitive architecture and linguistics theories together
• There are in-depth theorizing about the syntactic and discourse-pragmatic constraints that govern the interpretation of Mandarin reflexives (e.g., Pan, 1997, 1998, 2001; Pollard & Xue, 1998, 2001; Huang & Liu, 2001; Anand, 2006; Cole et al., 2006; Huang et al., 2009; Wang & Pan, 2014, 2015a,b; Giblin, 2016; Charnavel et al., 2017), there are relatively few studies examining these theroatical approach from a testable experimental approach (but see e.g., Schumacher et al., 2011; He & Kaiser, 2016; Han, 2020; Liu, 2020)

2 competing antecedent retrieval models debated in the sentence processing literature

• Central question: whether syntactic information is prioritized over non-syntactic information (e.g., gender, number, animacy) in early-stage processing 1. the standard cue-based retrieval model (e.g., Lewis & Vasishth, 2005; Van Dyke & McElree, 2006; Chen et al., 2012; Jäger et al., 2015; Patil et al., 2016) 2. the structure-based retrieval model (e.g., Sturt, 2003; Xiang et al., 2009; Van Dyke & McElree, 2011; Dillon et al., 2013; Cunnings & Sturt, 2014; Parker & Phillips, 2017)

Research Background

• Long-distance binding of ziji
  • One key linguistic property of ziji is that it can be exempt from Binding Principle A (Chomsky, 1981, 1986) or the locality constraint
  • Binding Principle A: A reflexive must be bound in its local domain (e.g., the smallest XP of the sentence, such as DP, TP, CP that contains a subject and a reflexive)
  • Linguistics noted early on that ziji can be free in the local domain and can be referred to a non-local antecedent (e.g., Huang et al., 1984; Battistella, 1989; Tang, 1989; Yu, 1992; Xu, 1993)
  • Exempt anaphor: Charnavel (2020) posits that plain (local) anaphors which is bound by overt antecedents, and exempt (long-distance) anaphors which covert logophoric operators bind, are simply bound by different antecedents.
  • Earlier syntactic approaches hold that LD ziji is only apparently non-local, but underlyingly locally bound, as ziji moves cyclically at the logical form (LF) to adjoin to a position that has a local relationship to its target antecedent (e.g., Battistella, 1989; Huang & Tang, 1991; Cole et al., 1990, 1993; Cole & Sung, 1994; Cole & Wang, 1996).

• Two analyses, but similar to an agreement-based account
• Head-movement analysis (presented by Cole et al., 1990): ziji moves up the tree to adjoin with the T2 head and percolates its [+3] feature to T2, because T in Chinese is arguably vacuous, after which subject-verb agreement for TP2 can proceed, ziji then moves through all the head positions to the next TP layer to feed subject-verb agreement in TP1.
• XP-adjunction analysis (Huang and Tang, 1991), which is implemented differently but close in spirit.
• 我的想法：这个的确和agreement很像，所以是自己agree with王老师和小明的3-person feature，所以可以refer到两个antecedent，但是感觉这个没有解释为什么这个可以LD bound? 只是说agreement
• 这个local relationship我没有看出来在哪里
• yaqing的解释：ziji本来是base-generate在最下面的NP那里，这时候就是local bound；但是在LF,ziji也可以继续往上move，跑到xiaoming下面的位置的时候，其实也是local bound；但是在LF上的movement我们是看不见的，所以就理解成long-distance bound （如果往上move的话感觉的确可以解释是local bound，理解了）

• Problems with the blocking effect:
• 小明，听说我喜欢自己。vs. 小明，听说你喜欢自己 （自己可以是你，但是不可以是小明, ziji can be referred to you, but not xiaoming）
  • blocking effect crashed due to subject-verb agreement = okay with subject blocking effect
  • ziji is generated with a [+1] person feature or gets its phi-feature from the local ‘I’ in TP
  • After it moves to TP1 layer and precolates its [+1] feature to T1, subject-verb agreement crashes because the matrix subject has an incompatible [+3] person feature
• 小明告诉我，王教授喜欢自己。vs. 小明告诉你，王教授喜欢自己。（自己可以是你，但是不能是小明, ziji can be referred to you, but not xiaoming）
• problem with the object blocking effect

• This approach predicts an object should not block LD binding, because an object blocker is not in a structural (e.g., subject) position to pass its phi-feature to ziji (showing that there should not be an object blocking effect, which is contrary to 小明告诉我，王教授喜欢自己。vs. 小明告诉你，王教授喜欢自己) = fail to take account of object blocking effect, ziji 不能是我，但是可以是你，这个account assume 自己有 +1 feature，所以可以bound to我，因为我也是 +1 feature

• Recent Approach
  • Giblin (2016) offers a syntactic solution to the object blocking effect
  • Discourse-Based Account: Before Giblin (2016), researchers proposed that discourse factors, such as perspective or point of view, play a critical role in LD binding of ziji and the blocking effect
    • if we assume ziji is perspective-sensitive = non-local antecedents can bind ziji when they are perspective centers or domains of point of view
      • a perspective center can be:
        • attitude holder = a person whose perspective or point of view the sentence expresses
        • empathy locus = a person whose perspective the speaker takes
    • Mandarin LD ziji is empathic = one must put themselves in the shoes of the perspective holder = the antecedent of ziji needs to be an empathy locus
    • empathy = speaker’s identification, with varying degrees, with a person who participates in the event that he describes in a sentence (Kuno & Kaburaki, 1977: 628)
  • Empathy locus and its interaction with the blocking effect
    • I or you encodes the speaker’s own perspective, then it means that LD ziji with the sentence-internal Xiaoming and sentence-external I or you includes a clash of perspectives
    • 我的想法：为什么 小明是sentence- internal，i or you 是sentence-external？
    • jun的解释：Speech Act Empathy Hierarchy (Kuno, 1987: 212)
      • The speaker cannot empathize with someone else more than with himself (speaker只能对自己更empathize比起其他人)
      • 这个approach好的地方是可以解释object blocking effect and subject blocking effect
    • overall: perspective-taking can influence the reflexive resolution of ziji

2 research questions

1. Does the discourse role (e.g., source vs. perceiver) of the perspective center impact LD binding of ziji?
2. Is there a syntactic component to the blocking effect?

1.2.1.1 Logophoricity, empathy, and ziji

• differences between empathy and logophoricity
  • logophoricity = certain pronominal forms refer to individuals whose speech, thoughts or general states of mind are reported (Clements, 1975: 141)
  • logophoricity is used in linguistics to refr to grammatical reflexes of perspectives
  • logophor = in some African languages, there is a logophoric pronoun that must uniquely refer to an attitude holder
• Logophoric hierarchy (Culy, 1994; 1062)
  • Speech > thought > knowledge > direct perception
• Logophoric hierarchy (Huang & Liu, 2001)
  • Source > Self > Pivot
    • typological observation that logophoric roles are higher on the scale
      • speech initiators or sources = more likely to be the antecedents of logophoric pronouns
      • Recipients of direct perception or perceivers are less likely to be the antecedents of logophoric pronouns
    • Each sentence or the point-of-view domain carries only one perspective center
      • each sentence can only have one referent = empathy locus = takes the perspective of it
    • perspective is not the core component of logophoricity
• In Japanese, ziben (‘self’) can have a logophoric/non-empathic use, where the external perspective center boku (I) can co-exist with the non-local antecedent whose perspective the sentence expresses
  • Oshima (2007) argued that there are three different uses of ziben
    • local zibun
    • logophoric zibun
    • empathic zibun
• One can take his/her own perspective while expressing the perspective of the non-local antecedent
• Logophoricity is related to the notion of perspective-reporting while empathy is related to the notion of perspective-taking

• Ziji is necessarily empathic and only under certain circumstances displays properties of logophoric reflexives
  • Wang and Pan 2015a,b found that
    • Ziji fits the criteria for empathic reflexives
      • the blocking effect
      • subject-orientation = only subjects can bind ziji
      • debatable = ziji must have the de se reading
  • de se reading = the antecedent must be aware of the event described by the sentence
    • If the speaker empathizes with Zhangsan, the same sentence can also be true = this suggests that ziji can be a pure empathic anaphor
    • ziji is necessarily empathic but sometimes shows logophoric properties = de se reading in the intensional context with a presence of a speech or belief verb with source subject

experimentally testing the logophoric properties of ziji

1. Liu (2020)
   • compares the acceptability of LD binding by a source referent to LD binding by a topical referent
     • 根据李四说，这件事伤害了自己 = according to lisi, this event hurt self
       • Lisi is the source
     • 说到李四，这件事伤害了自己 = speaking of lisi, this event hurt self
       • Lisi is the topic
   • using ye/no acceptability judgment task, Liu found that average proportion of acceptability for source is close to the ceiling, but the average proportion of acceptability for topic is less than 50%
   • suggest that the source is more likely to license LD binding than a topic referent
   • alternative explanation = lisi is in the object position as a topic but in the subject position as a source
   • giving the subject orientation property of ziji, this is expected
   • inanimate of the local antecedent = the event = effect of animacy? Or the logophoricity effect still remains if the local antecedent is animate
2. Xu and Runner (submitted)
   • antecedent forced-choice
   • When the local antecedent is animate, the logophoric role (believer vs. perceiver) of the matrix subject does not influence preferences for LD binding of ziji
   • They only tested the believer and the perceiver, which are not the two ends of the logophoric scale
   • This can be explained by their null result, because it does not disprove the hypothesis that ziji can show logophoric properties

1.2.1.2 Syntactic prominence and the blocking effect

• An agreement-based syntactic approach can account for the subject blocking effect, but not the object blocking
• A discourse-based approach can account for both the subject and object blocking effect
• It seems the discrouse-based account is empirically superior to the syntactic approach

Other two Approachs that can account for subject and object blocking effect

1. Contiguous-agreement account (Giblin, 2016)

• For LD binding, the C head at the left edge of the matrix clause is merged with an unvalued feature (uparticipant)
• Upon merge, this feature probes downward for a [+participant] feature in the spec-TP - If the contiguous matrix subject carries a [+participant] feature, the C head will be valued - All the T heads (matrix and local T head) will then inherit this [+participant] carried by the matrix subject (by the local T head in a T-V-reflexive chain) - one condition must be satisfied - there must be no intervening DP with a [+ participant] feature - If the contiguous DP/spec-TP is [-participant], the derivation crashes due to the violation of the contiguous Agree constraint.
• 我没有太明白这个approach是什么意思？
• 什么是contiguous Agree constraint？

1. Unified account (Cole et al., 2006; Charnavel et al., 2017)
   • Explain the object blocking effect using syntactic means
   • does not predict any difference in the size of the subject vs object blocking effects
   • blocking effect has both a syntactic componenet and a discourse componenet
   • zhangsan yiwei lisi hui ba ni dai hui ziji de jia
     - due to conflict of prespectives between empahty locus zhangsan and the speaker’s perspective - you is uttered from the perspective of the speaker

• It seems that different researchers have different tuitions about subject and object blocking effects, which suggest inter-speaker variaiton

two types of discourse-level prominence

• discrouse topic prominence
• logophoricity

“animacy blocking effect” by Pan (1998)

• if the local antecedent is both animate and loacal, then only the local binding is allowed

• animacy prominence is a type of semantic prominence

• animacy blocking effect is reminiscent of the “intervention effect” observed for “standard American English” (Pollard & Sag, 1992): when a local human c-commander is available, the reflexive must be locally bound

• The real antecedent philip is a perspective center (empathy locus) in the domain of point of view

• 43% of the time, Korean native speakers accepted LD binding of caki-casin, becuase caki-casin just like ta-ziji is considered as a typical local refleixves, but when there is a discourse prominent antecedent present in the sentence, they will also be LD bound as well.

1.2.2.2 Logophoricity and ta-ziji

• source-advantage is the logophoricity effect
• a reflexive is taken to be logophoric if the relfexive is sensitive to the logophoric hierarchy - consider the two ends of the logophoric hierarchy, which are source and perceiver
• early empirical work shown that LD reflexives include the complex reflexives which show sensitivity to logophoricity as in English himself/herself (Kaiser et al., 2009; Sloggett, 2017)
• A source matrix subject is preferred as the binder compared to a perceiver subject
• visual world eye-tracking study shown that the source preference emerges quite early, during the 200-600 ms time window
• Kaiser et al., 2009 provided crucial experimental evidence using both offline and online data that English reflexives in the picture NP constructions can be logophoric
• In contrast to Kaiser et al.’s (2009) finding that the logophoricity effect can appear across animate local antecedents, Sloggett found that with argument-position reflexives, the source advantage is only present when the local antecedent is an impossible antecedent for the reflexive, in particular, when the local antecedent mismatches with the reflexive in two phi-features (number, gender) or when the local antecedent is inanimate.
• Korean caki-casin is sensitive to the source role of the non-local topically marked antecedent.
• LD binding by source is more acceptable than LD binding by self which is in turn more acceptable than LD binding by pivot.
• Study by Xu and Runner did not see a logophoricity effect for compound reflexive, and this can due to the fail of selection of the logophric verbs, such as belief and preciver, instead of source and preceiver
• For other studies, such as Kaiser etal., 2009 and Kim & Yoon, 2009 - they found that there is an effect for logophoricity, and this can be due to that they created an exempt enviorment, which is no strictly subject to synatctic binding theory - such as Kaiser’s Picture NP - Kim and Yoon 2009’s study, the matrix subject is also the discrouse and strucutral topic

## 1.2.3 LD binding of ziji and ta-ziji: A summary

• ziji is necessarily empathic, which require the speaker/comprehender to take the pespective of its LD binder
• Empathic nature of ziji explains the blocking effect
• Not sure if ziji shows the logophricity effect with an attitude holder or source and its non-local antecedent, and the relation between the blocking effect and syntactic prominence is not determinered

1.3 Antecedent retrieval and sentence processing

• In sentence processing work on reflexive resolution, researchers are more concerned with the time course with which the various linguistic factors guide antecedent retrieval
• mismatch between people’s final offline interpretations and incremental potentially transient processing patterns - such as some linguitic factors or cues may play a role in offline judgment tasks but not necessarily in online parsing
  • could mean that vertain non-syntactic factors are accessed at the later comprehnesion stages and other such as syntactic factors are accessed much earliers

1.3.1 Standard cue-based retrieval model

• Lewis and Vasishth (2005)
• aims to explain moment-by-momemt antecedent retrieval in real-time parsing
• memory is invovled - procedural: building and attaching syntactic struccutres - declarative: non-procedural infromation: lexical information in long-term memeory, and recently processed lexical items and intermediate syntactic strcurtres in working memory
• when a new lexical item is attached to the syntactic sturcure, the parser needs to check the current linguitstic strucure in the declarative memoery and activate the relevant node for attachment of the next lexcial items
• when an anaphor is encounted, a dependence needs to be established between it and the antecedent
• antecedent-anaphor dependency is established through the assoicaiton or overlap of retrieval cues encoded on the anaphor and the potential antecedent since the earliest moments of processing
• more feature overlap means greater retrueval success and faster retrieval times
• Assume for all cues, such as semantic. discoursal, syntactic all by default have similar weights
• the standard model does not prioritize accessibility of certain linguistic cues over others

1.3.2 Structure-based retrieval model

• the main difference is that syntactic cues are priorized or weighted more heavily than semantic cues such as gender, number, animacy and other non-syntactic cues like discourse topicality
• only when syntactic constraints are satisfied, then non-syntactic constraints be considered
• syntactic accessibility/syntactic legitimacy = antecedents are syntactically more prominent (local, c-commanding)
• syntactic cues (c-command, locality, subjecthood) are more prioritized compared to non-syntactic cues
• early applicaiton of syntactic constraints does not mean that non-syntactic cues are ignored at all processing stages
• 2 views
  • only syntactic cues are accessible at the early processing stages
  • syntactic cues play a predominant role but do not entirely block the accessibility of the syntactically inaccessible antecedents
• Most studies supporting the structure-based model inspect the role of semantic cues such as gender, number, and animacy

1.3.3 A critical review of the sentence processing literature

• most studies support the strcuture-based retrieval models ((e.g., Nicole & Swinney, 1989; Sturt, 2003; Gao et al., 2005; Xiang et al., 2009; Van Dyke & McElree, 2011; Dillon et al., 2013, 2016; Qian & Wu, 2016; Cunnings & Sturt, 2014, 2018; Parker & Phillips, 2017; Wang, 2017; Jäger et al., 2020; also see the meta-analysis in Jäger et al., 2017).
• few studies found early interference effect from non-local and/or c-commanding antecedents (e.g., Runner et al., 2006; Kaiser et al., 2009; Thompson & Choy, 2009; Chen et al., 2012; Jäger et al., 2015; Patil et al., 2016; Sloggett, 2017; Chang et al., 2020)
• individual variation in working memory (Cunnings & Felser, 2013) and reading speed (Yadav et al., 2022) and activation decay of antecedents (King et al., 2012). - mixed picture come from differences in experimental designs, nature of specific linguistic properties in certain languages, variation in the test population, and statistical power

1.3.3.1 What do studies on English reflexives tell us about antecedent retrieval?

• priming effect is largest when the word is related to the local referent doctor, suggesting local antecedents are most activated arguably because participants tend to take the local referent as the antecedent.
• the locality constraint highly restricts antecedent retrieval at the earliest moments of reflexive resolution.
• Sturt (2003) found that English native speakers only considered local antecedents during the early processing stages
• the discourse prominent but syntactically non-local antecedent Jonathan was not considered by English native speakers during early stages of comprehension
• at later processing stages discourse prominence plays a role
• supporting structure-based retrieval model: syntactic constraints filter out structurally inaccessible antecedents at an early stage
• recall that Zribi-Hertz (1989) (also see Pollard & Sag, 1998, 2001) showed that literary British English allows empathic readings of himself/herself when the non-local antecedent is an empathy locus.
• Two real-time processing studies (Kaiser et al., 2009; Sloggett, 2017) suggest that logophoricity seems to modulate reflexive resolution

Kaiser et al. (2009)

• visual world eye-tracking, participants listened to audio stimuli while looking at the pictures on the computer screen
• Kaiser et al. (2009) found that in sentences like (1.29-1.30) English native speakers showed early sensitivity to the source vs. perceiver role of the matrix subject –> source advantage
• Peter {told/heard from} Andrew about the picture of himself on the wall
• Peter {told/heard from} Andrew about Greg’s picture of himself on the wall.
• higher proportions of looking at peter in the source condition compared to the perceiver condition for both matrix subject in the local and non-local antecedent position
• logophoricity modulates reflexive resolution at least with picture NP reflexvies

Sloggett (2017)

• logophoricity and (object) argument reflexive resolution
• The librarian {said/heard} that the schoolboys misrepresented herself at the meeting.
• when local antecedent schoolboys is an impossible binder (due to the mismatch of two phi-features, gender and number)
• English reflexives are “co-opted” as exempt anaphors which unsurprisingly show logophoric properties.
• intervention effect

Overall

• the syntactic cue of structural locality has higher weight than non-syntactic cues such as gender, number, discourse topic prominence, logophoricity at least in English

1.3.4 What can ziji and ta-ziji say about antecedent retrieval?

• an event-related potential (ERP) study by Li and Zhou (2010) only examined what ERP effect is triggered in long-distance binding of ziji
• This is because the local antecedent satisfies [+local] and [+animate] while the non-local antecedent only satisfies [+animate], which leads to higher probability of local retrieval due to complete cue overlap.
• In a self-paced reading study, Chen et al. (2012) manipulated the animacy of the syntactically inaccessible distractor to test if semantic animacy guides antecedent retrieval despite its syntactic inaccessibility
• reading times are slower when the syntactically inaccessible distractor is animate (e.g., ‘protesters’) compared to when it is inanimate (e.g., ‘protest’). This suggests that animacy is one of the cues that guide antecedent retrieval even though it is non-locally embedded in a temporal adjunct clause.
• However, when ziji has a local animate antecedent, Chen et al. (2012) did not find an interference effect from the syntactically inaccessible distractor
• Thus, the studies on Mandarin ziji seem to present a very different picture compared to studies on English reflexives
• the linguistic properties of ziji are the reason behind the divergent findings in English and Chinese as ziji can appear in non-local environments more freely
• partial cue match (matrix subject being [+animate, -local]) facilitates processing compared to no cue match (matrix subject being [-animate, -local])
• locality bias found in prior studies on ziji and ta-ziji can be alternatively explained as a linear recency bias, rather than a structural or syntactic locality effect ?

1.3.5 Reflexive resolution and antecedent retrieval models: A summary

• reflexive resolution is clearly influenced by syntactic information such as syntactic locality
• syntactic locality plays a predominant role in reflexive resolution, for both English and Mandarin reflexives.
• research on ziji and ta-ziji suggests that semantic cues such as gender and animacy impact early-stage processing while research on English reflexives presents a more mixed picture

Chapter 2

• “two-antecedent” design
• Mandarin offers a ideal testing ground to disentangling linear recency and syntactic locality with its prenominal relative clasue structures
• If linear recency impacts relfeixve resolution, in an incremental reading task, we will expect reading time slowdowns at and/or after the reflexive when NP1 mismatches in semantic features with the reflexive, compared to when NP1 matches the animacy of the reflexive
• If syntactic locality guides reflexive resolution, we should see reading slowdowns at and/or after NP2

2.2 Antecedent retrieval models

• the standard cue-based retrieval model and the structure-based retrieval model
• relevant retrieval cues = semantic animacy and syntactic locality
• standard cue-based retrieval model: assumes that all retrieval cues, syntactic or not, can be immediately used fro antecedent retrieval
• strcuture-based retrieval model: syntactic cues are prioritized in early-stage comprehension
• Only when syntactic constraints are satisfied is non-syntactic information accessed
• predicts that animacy of the syntactically inaccessible NPs should not impact reflexive resolution
• structure-based retrieval mechanism can make predictive use of the relative clause structure to anticipate the head noun to resolve the ambiguity.

2.4 Experiment a

• two antecedent design
  • the local NP is also the linearly closer NP
  • acceptability judgment task = final interpretation of the reflexive
  • factors: reflexive (ziji/ta-ziji)
  • distance: recent/distant
  • factorial design

Predictions

• main effect of distance is expected: sentence with recent animate antecedents should be more acceptable than those with distance animate antecedentd
• interaction of relfeixve and distance: if the recency/locality biases associated with ziji and ta-ziji are different: the aniamcy mismatch effect shown by ziji may be different compared to ta-ziji

analysis

• Mixed effects ordinal regression
• The fixed effects are Reflexive and Distance
• The random effects are slopes and intercepts grouped by participants and items
• sum coding were fitted for the factors of reflexive and distance
• Statistical models were fitted parsimoniously (Bates et al., 2015): simpler models are preferred over more complex models if the models do not differ significantly

Results

• main effect of reflexive (p - 0.001)
• main effect of distance (p < 0.001) - assumeing the higer acceptability is correalted less processing cost, then sentence containing ziji is easier to process than sentences containing ta-ziji = higher frequency of ziji
  • higher frequency = easier access
  • Chinese speakers prefer short-distance bidning over long-distance bidning
  • the interaction between reflexvie and distance is not significant

Discussion

• replicated the locality/recency bias of ziji and ta-ziji

2.4 Experiment 1b

• Experiment 1b is a self-paced reading experiment which reflects incremental comprehension processes
• pariticpants read sentences presented region by region
• at the end of each sentence, paritipants answered the two-alternative comprehension question that does not probe the interpretation of the reflexvie
• the accuarcies of paritipants’ answer choices were used as indicators of attention

2.4 Experiment 1b data analysis

• accuracies below 75% would be removed due to concern with lapse of attention during the experiment
• RTs shorter than 100 ms and longer than 3000ms were removed as outliers
• RTs above 2.5 SDs of the mean were also removed. These two steps affected 3.23% of the original data
• Statistical analyses were conducted over both raw and log-transformed RTs using mixed effects linear regression

2.4 Experiment 1b results

• at the critical region, there is a main effect of reflexive (p < 0.001)
• Main effect of distance and the reflexive x distance interaction are not significant
• Experiment 1b did not find any evidence suggesting a locality/recency bias difference between ziji and ta-ziji
• could be that by default ziji is syntactic anaphor governed by principle A, unless triggered by discourse-pragmamtic = perspective-taking factors

2.5 Experiment 2a

• acceptability judgment task
• Reflexive ziji/taziji, positon local/recent were crossed in a factorial design
• position = position of the animate antecedent
• in local condition = animate antecdents are post-reflexvie
• recent condition = animate antecedent are pre-reflexvie

Prediciton

• First, if the ‘locality bias’ is in fact linear recency, participants should prefer reflexive binding by a linearly recent, animate antecedent than by a post-reflexive, animate antecedent.
• main effect of position
• reflexive x position interaction = the acceptability difference between the recent and local conditions may be different for ziji and ta-ziji

Results

Discussion

• Chinese speakers prefer local binding instead of recent binding
• syntactic locality is a separate factor from linear recency

2.5 Experiment 2a

• self-paced reading

Prediciton

• If Chinese speakers attempt to bind the reflexvie to the linearly recent matrix subject, the inanimate matrix subjects in the local match should lead to reading slowdowns = an aniamcy misamtch effect
• Main effect of position

Results

• main effect of relfexive (p < 0.001)
• taziji is longer than ziji
• signficant reflexvie x position interaction (p < 0.005)
• the animacy of the linearly recent antecedents impacts the processing of ziji and ta-ziji in different ways

Discussion

• syntactic locality and linear recency
• providing empirical evidence that linear recency and syntactic locality both guide antecedent retrieval
• both linear recency and syntactic locality guide antecedent retrieval

2.6.1 Summary of results

• the offline acceptability judgment and online self-paced reading experiments show that syntactic locality plays a distinct role from linear recency
• syntactic locality should not be conflated with linear recency
• the experiments in this study did not show strong evidence that the locality bias strengths of ziji and ta-ziji are different
• at the early processing stages that ta-ziji is not immediately bound to the linearly recent antecedent
• ta-ziji may be more sensitive to the syntactic locality constraint than ziji is

2.6.2 Implications for sentence processing and theoretical linguistics

• both sentence processing models and lingusitic theories
• experiment 2b = predicitons with standard cue-based retrieval model which does not assign any privileged status to syntactic constraints
• animacy of the syntactically inaccessible antecedent outside of local domain has been found to modulate the processing patterns of ziji
• structure-based retrieval model is not entirely refuted either due to the early Reflexive x Position interaction at the reflexive region with ta-ziji
• However, this assumption leaves aside the following question: when is ziji used as a syntactic anaphor and when is it used as an exempt anaphor?
• In this study, only two items have perceiver verbs while the rest (22 items) have source verbs (e.g., ‘say’, ‘state’, ‘disclose’, ‘confess’, ‘claim’) and belief verbs (‘think’) which make the matrix subject attitude holders.
• According to research on exempt anaphors, attitude holders and empathy loci license exempt reading of reflexives (e.g., Charnavel & Zlogar, 2015; Charnavel, 2020, 2021). Thus, we could expect that the availability of an attitude holder can cue the participants to have an exempt reading of ziji.

Chapter 3: Discourse topic prominence and reflexive resolution

• The discourse-level factor this chapter focuses on is discourse topic prominence
• discourse topicality may impact the resolution of ziji through perspective-taking and the resolution of ta-ziji through a non-perspectiverelated prominence constraint
• examine the interaction of verb semantics and the locality constraint selfdirected verbs and other-directed verbs. Self-directed verbs (or ‘introverted’ verbs) describe actions that “one generally performs upon one’s self” and other-directed verbs (or ‘extroverted verbs’) denote actions that one “usually performs towards others”

3.2 Discourse topic prominence and Mandarin reflexives

• For instance, using Sell’s (1987) terminology, Huang and Liu (2001) showed that the long-distance use of ziji is allowed when the antecedent is a source (intentional agent of communication), self (one whose mental state or attitude the proposition describes), or pivot (one with respect to whose space-time location the proposition is evaluated)
• source, self, and pivot roles can be perspective centers (or empathy loci)
• if the speaker puts him/herself in the shoes of the non-local antecedent or assumes the perspective of the empathy locus, long-distance binding is licensed
• Topic Empathy Hierarchy (Kuno, 1987),
  • involving two participants A and B such that A is a discourse topic, it is easier to take the perspective of A than B
  • predicts that readers should prefer the perspective of Xiaoming
  • perspective-sensitive ziji should tend to be interpreted as referring to the perspective center Xiaoming (rather than Professor Wang)

3.3 Verb semantics and Mandarin reflexives

• I follow Pollard and Xue (2001) in assuming that an empathic reading can be triggered when language specific discourse/pragmatic constraints (e.g., logophoricity, discourse prominence) are satisfied
• whether the prominence constraint for ta-ziji can be extended to the discourse level and (ii) whether syntactic locality is a stronger constraint than verb semantics for ta-ziji.

3.6 Experiment 3: Processing of ziji in discourse

• Experiment 3a = forced choice judgment task
• two factors
• context (neutral/biases)
• The term ‘neutral context’ refers to a context that lacks a discourse topic character
• The term ‘biased context’ refers to a context with a discourse topic character (e.g., Xiaoming) which is the (non-local) matrix subject of the critical sentence
• verb directedness (self/other-directed)
• Experiment 3a = self-paced reading experiment

Predictions 3a

• participants should prefer local antecedents with self-directed verbs and non-local antecedents with other-directed verbs
• main effect for context = if discourse topic prominence infleucned the interpreaion of ziji
• when non-local antecedent is the discourse topic in the baised context, then non-local choice preference should be stronger relative to the neutral context
• when verb is self-directed, local referents are preferred as antecdents in the biased context
• when verb is other-directed, non-local referents are preferred

Predictions 3b

• the standard and strucure-based retrieval models will predict local bidning since there is no discrouse-level cues = netural context
• other-directed verbs will lead to reading slowdowns at the reflexive region in netural context
• baised contexts, the two processing models will make different predcitons - topic empathy hierarchy - paritipants should emphazie with or take the perspective of the discourse topic - stnadard cue-based retrieval model = predicts that the language parsers will make quick use of the [+POV] cue in addition to [+Subj] for antecedent retrieval - self-directed verb should cause reading slowdowns
  • main effect of context is expected
  • should prefer the prespective of the discourse topic after reading the context sentence = in reading the ensuing critical sentence with the discourse topic as the subject, we expect paritiapnts to maintain the perspective

Results 3a

• in the neutral context, the proportions of local choices under self-directed verbs are higher
• proportions of non-local choices under other directed verbs are similar
• both verbs show a preference for the referents favroed by the verb’s semantic directionlity bias
• discourse topicality influences paritiapnts antecedent choices as the discourse topic status of the non-local subject lowers the proportions of local choices

Results 3b

• First, words in the biased context are processed faster compared to the neutral context.
• Second, the verb directedness effects already appear at the verb, prior to the reflexive ziji.
• At the reflexive itself, RTs are faster in biased contexts than in neutral contexts, but there appears to be no effect of verb directedness.

Discussion of Experiments 3a and 3b

• invesitage the locality constraint, verb semantics and discourse topic prominenve guide the offline and online interpretation of ziji
• judgement data show the interpreatation of ziji is strongly infleucned by verb directedness and the discourse topicality also increase the proporitons of non-local choices
• standard cue-based retrieval model predicts that paritipants will immediate locate the discourse topic - notated with [+POV] due to its tendency to be empathized with
• as the antecedent at the earliest possible moments
• strcutre-based retrieval modle predicts that only syntactic cues such as locality information will be accessed by the paser and only local antecedents should be considered at this early stages