West Coast Conference on Formal Linguistics (WCCFL) (38th : 2020)

(Bound) Pronouns in Competition : Ambiguity Avoidance in Romanian Anaphora Production Ivan, Rodica; Dillon, Brian 2020-03-07

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Notice for Google Chrome users:
If you are having trouble viewing or searching the PDF with Google Chrome, please download it here instead.

Item Metadata

Download

Media
73804-Ivan_R_et_al_Bound_pronouns_WCCFL38_2020.pdf [ 24.67MB ]
Metadata
JSON: 73804-1.0389861.json
JSON-LD: 73804-1.0389861-ld.json
RDF/XML (Pretty): 73804-1.0389861-rdf.xml
RDF/JSON: 73804-1.0389861-rdf.json
Turtle: 73804-1.0389861-turtle.txt
N-Triples: 73804-1.0389861-rdf-ntriples.txt
Original Record: 73804-1.0389861-source.json
Full Text
73804-1.0389861-fulltext.txt
Citation
73804-1.0389861.ris

Full Text

Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General Discussion(Bound) Pronouns in Competition:Ambiguity Avoidance in RomanianAnaphora ProductionRodica Ivan & Brian DillonUMass Amherst{rivan, brian}@linguist.umass.eduWCCFL 38 • UBC •March 7, 2020Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:https://tinyurl.com/boundWCCFLBackground Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionToday’s TalkDespite the fact that Romanian seems to be exempt from classicCondition B, speakers are sensitive to a generic pragmatic constraintwhich regulates pronominal reference.We provide psycholinguistic evidence from production whichsupports the hypothesis that ambiguity avoidance strategies steerpronominal usage in cases of local coreference, and, surprisingly tothe BT literature, also for bound variable dependencies.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionToday’s TalkDespite the fact that Romanian seems to be exempt from classicCondition B, speakers are sensitive to a generic pragmatic constraintwhich regulates pronominal reference.We provide psycholinguistic evidence from production whichsupports the hypothesis that ambiguity avoidance strategies steerpronominal usage in cases of local coreference, and, surprisingly tothe BT literature, also for bound variable dependencies.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat are the grammatical constraints on pronominalinterpretation?The traditional answer pertains to the Binding Theory.(1) CONDITION BA pronoun must be free in its binding domain.(Chomsky, 1981, 1986; Bu¨ring, 2005)However, Condition B is not a crosslinguistic fact: it does not hold fora number of different languages (Khanty - Volkova & Reuland, 2014;, Jambi -Cole et al., 2017; Chamorro - Wagers et al., 2018; a.o.).Romanian is one such language.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat are the grammatical constraints on pronominalinterpretation?The traditional answer pertains to the Binding Theory.(1) CONDITION BA pronoun must be free in its binding domain.(Chomsky, 1981, 1986; Bu¨ring, 2005)However, Condition B is not a crosslinguistic fact: it does not hold fora number of different languages (Khanty - Volkova & Reuland, 2014;, Jambi -Cole et al., 2017; Chamorro - Wagers et al., 2018; a.o.).Romanian is one such language.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat are the grammatical constraints on pronominalinterpretation?The traditional answer pertains to the Binding Theory.(1) CONDITION BA pronoun must be free in its binding domain.(Chomsky, 1981, 1986; Bu¨ring, 2005)However, Condition B is not a crosslinguistic fact: it does not hold fora number of different languages (Khanty - Volkova & Reuland, 2014;, Jambi -Cole et al., 2017; Chamorro - Wagers et al., 2018; a.o.).Romanian is one such language.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat are the grammatical constraints on pronominalinterpretation?The traditional answer pertains to the Binding Theory.(1) CONDITION BA pronoun must be free in its binding domain.(Chomsky, 1981, 1986; Bu¨ring, 2005)However, Condition B is not a crosslinguistic fact: it does not hold fora number of different languages (Khanty - Volkova & Reuland, 2014;, Jambi -Cole et al., 2017; Chamorro - Wagers et al., 2018; a.o.).Romanian is one such language.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRomanian pronouns are not subject to Condition B!In Romanian, clitics obey Condition B, in the case of (di)transitivepredicates. Furthermore, these clitics constrain the reference of overtpronouns in direct and indirect object positions.(2) a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.The pronoun el is compatible with either interpretation.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRomanian pronouns are not subject to Condition B!In Romanian, clitics obey Condition B, in the case of (di)transitivepredicates.Furthermore, these clitics constrain the reference of overtpronouns in direct and indirect object positions.(2) a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.The pronoun el is compatible with either interpretation.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRomanian pronouns are not subject to Condition B!In Romanian, clitics obey Condition B, in the case of (di)transitivepredicates. Furthermore, these clitics constrain the reference of overtpronouns in direct and indirect object positions.(2) a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.The pronoun el is compatible with either interpretation.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRomanian pronouns are not subject to Condition B!In Romanian, clitics obey Condition B, in the case of (di)transitivepredicates. Furthermore, these clitics constrain the reference of overtpronouns in direct and indirect object positions.(2) a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.The pronoun el is compatible with either interpretation.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRomanian pronouns are not subject to Condition B!In Romanian, clitics obey Condition B, in the case of (di)transitivepredicates. Furthermore, these clitics constrain the reference of overtpronouns in direct and indirect object positions.(2) a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.The pronoun el is compatible with either interpretation.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRomanian pronouns are not subject to Condition B!In Romanian, clitics obey Condition B, in the case of (di)transitivepredicates. Furthermore, these clitics constrain the reference of overtpronouns in direct and indirect object positions.(2) a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.The pronoun el is compatible with either interpretation.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsIn the case of non-transitive predicates, however, cliticization is not anoption. There are no oblique clitics in Romanian!Consequently, regular pronouns are ambiguous between reflexiveand non-reflexive interpretations.(3) Lockhart1Lockhartahasvorbittalkeddespreaboutel1/2.himLockhart talked about himself / about someone else.blaRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsIn the case of non-transitive predicates, however, cliticization is not anoption. There are no oblique clitics in Romanian!Consequently, regular pronouns are ambiguous between reflexiveand non-reflexive interpretations.(3) Lockhart1Lockhartahasvorbittalkeddespreaboutel1/2.himLockhart talked about himself / about someone else.blaRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsIn the case of non-transitive predicates, however, cliticization is not anoption. There are no oblique clitics in Romanian!Consequently, regular pronouns are ambiguous between reflexiveand non-reflexive interpretations.(3) Lockhart1Lockhartahasvorbittalkeddespreaboutel1/2.himLockhart talked about himself / about someone else.blaRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsIn the case of non-transitive predicates, however, cliticization is not anoption. There are no oblique clitics in Romanian!Consequently, regular pronouns are ambiguous between reflexiveand non-reflexive interpretations.(3) Lockhart1Lockhartahasvorbittalkeddespreaboutel1/2.himLockhart talked about himself / about someone else.blaRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsIn the case of non-transitive predicates, however, cliticization is not anoption. There are no oblique clitics in Romanian!Consequently, regular pronouns are ambiguous between reflexiveand non-reflexive interpretations.(3) Lockhart1Lockhartahasvorbittalkeddespreaboutel1/2.himLockhart talked about himself / about someone else.blaRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsThis ambiguity is especially surprising given the existence of reflexiveanaphora in Romanian.Languages with dedicated reflexives arepredicted to obey Condition B (Rooryck & vanden Wyngaerd, 2011).(4) a. LockhartLockhartahasvorbittalkeddespreaboutel.himPRONOUNLockhart talked about him / himself.b. LockhartLockhartahasvorbittalkeddespreaboutsine.selfREFLEXIVELockhart talked about himself.c. LockhartLockhartahasvorbittalkeddespreaboutelhimıˆnsus¸i.himselfEMPHATICLockhart talked about himself.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsThis ambiguity is especially surprising given the existence of reflexiveanaphora in Romanian. Languages with dedicated reflexives arepredicted to obey Condition B (Rooryck & vanden Wyngaerd, 2011).(4) a. LockhartLockhartahasvorbittalkeddespreaboutel.himPRONOUNLockhart talked about him / himself.b. LockhartLockhartahasvorbittalkeddespreaboutsine.selfREFLEXIVELockhart talked about himself.c. LockhartLockhartahasvorbittalkeddespreaboutelhimıˆnsus¸i.himselfEMPHATICLockhart talked about himself.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsThis ambiguity is especially surprising given the existence of reflexiveanaphora in Romanian. Languages with dedicated reflexives arepredicted to obey Condition B (Rooryck & vanden Wyngaerd, 2011).(4) a. LockhartLockhartahasvorbittalkeddespreaboutel.himPRONOUNLockhart talked about him / himself.b. LockhartLockhartahasvorbittalkeddespreaboutsine.selfREFLEXIVELockhart talked about himself.c. LockhartLockhartahasvorbittalkeddespreaboutelhimıˆnsus¸i.himselfEMPHATICLockhart talked about himself.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsThis ambiguity is especially surprising given the existence of reflexiveanaphora in Romanian. Languages with dedicated reflexives arepredicted to obey Condition B (Rooryck & vanden Wyngaerd, 2011).(4) a. LockhartLockhartahasvorbittalkeddespreaboutel.himPRONOUNLockhart talked about him / himself.b. LockhartLockhartahasvorbittalkeddespreaboutsine.selfREFLEXIVELockhart talked about himself.c. LockhartLockhartahasvorbittalkeddespreaboutelhimıˆnsus¸i.himselfEMPHATICLockhart talked about himself.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThe ambiguity of pronominal PP objectsThis ambiguity is especially surprising given the existence of reflexiveanaphora in Romanian. Languages with dedicated reflexives arepredicted to obey Condition B (Rooryck & vanden Wyngaerd, 2011).(4) a. LockhartLockhartahasvorbittalkeddespreaboutel.himPRONOUNLockhart talked about him / himself.b. LockhartLockhartahasvorbittalkeddespreaboutsine.selfREFLEXIVELockhart talked about himself.c. LockhartLockhartahasvorbittalkeddespreaboutelhimıˆnsus¸i.himselfEMPHATICLockhart talked about himself.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionBackgroundRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionBinding TheoriesSentences like (5) traditionally fall under the purview of the classicBinding constraints.(5) * Lockhart admires him. RULED OUT BY CONDITION BHowever, some theoretical accounts of reference suggest thatambiguity avoidancemodulates the competition between reflexiveand non-reflexive pronouns in these sentences.(Bolinger 1979; Dowty, 1980; Reinhart, 1983; Levinson, 1987, 2000; a.o.)(6) a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionBinding TheoriesSentences like (5) traditionally fall under the purview of the classicBinding constraints.(5) * Lockhart admires him. RULED OUT BY CONDITION BHowever, some theoretical accounts of reference suggest thatambiguity avoidancemodulates the competition between reflexiveand non-reflexive pronouns in these sentences.(Bolinger 1979; Dowty, 1980; Reinhart, 1983; Levinson, 1987, 2000; a.o.)(6) a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionBinding TheoriesSentences like (5) traditionally fall under the purview of the classicBinding constraints.(5) * Lockhart admires him. RULED OUT BY CONDITION BHowever, some theoretical accounts of reference suggest thatambiguity avoidancemodulates the competition between reflexiveand non-reflexive pronouns in these sentences.(Bolinger 1979; Dowty, 1980; Reinhart, 1983; Levinson, 1987, 2000; a.o.)(6) a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionBinding TheoriesSentences like (5) traditionally fall under the purview of the classicBinding constraints.(5) * Lockhart admires him. RULED OUT BY CONDITION BHowever, some theoretical accounts of reference suggest thatambiguity avoidancemodulates the competition between reflexiveand non-reflexive pronouns in these sentences.(Bolinger 1979; Dowty, 1980; Reinhart, 1983; Levinson, 1987, 2000; a.o.)(6) a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionBinding TheoriesSentences like (5) traditionally fall under the purview of the classicBinding constraints.(5) * Lockhart admires him. RULED OUT BY CONDITION BHowever, some theoretical accounts of reference suggest thatambiguity avoidancemodulates the competition between reflexiveand non-reflexive pronouns in these sentences.(Bolinger 1979; Dowty, 1980; Reinhart, 1983; Levinson, 1987, 2000; a.o.)(6) a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?Inspired by Dowty (1980)’s avoid ambiguity principle and GRICEANreasoning, a number of approaches to the Binding Theory argue infavor of pragmatic constraints regulating the competition betweenhimself and him in ambiguous contexts.→ Rule I (Reinhart, 1984, 2006; Grodzinsky & Reinhart 1993)→ Coreference Rule (Heim, 1993; Roelofsen, 2010)→ Minimize Restrictors! (Schlenker, 2005; Johnson 2013)This reasoning also led to a number of accounts which constrain thecompetition between pronouns and reflexives in terms of ECONOMYconsiderations. (Safir 2004, 2014; Rooryck & vanden Wyngaerd, 2011; a.o.)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?Inspired by Dowty (1980)’s avoid ambiguity principle and GRICEANreasoning, a number of approaches to the Binding Theory argue infavor of pragmatic constraints regulating the competition betweenhimself and him in ambiguous contexts.→ Rule I (Reinhart, 1984, 2006; Grodzinsky & Reinhart 1993)→ Coreference Rule (Heim, 1993; Roelofsen, 2010)→ Minimize Restrictors! (Schlenker, 2005; Johnson 2013)This reasoning also led to a number of accounts which constrain thecompetition between pronouns and reflexives in terms of ECONOMYconsiderations. (Safir 2004, 2014; Rooryck & vanden Wyngaerd, 2011; a.o.)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?Inspired by Dowty (1980)’s avoid ambiguity principle and GRICEANreasoning, a number of approaches to the Binding Theory argue infavor of pragmatic constraints regulating the competition betweenhimself and him in ambiguous contexts.→ Rule I (Reinhart, 1984, 2006; Grodzinsky & Reinhart 1993)→ Coreference Rule (Heim, 1993; Roelofsen, 2010)→ Minimize Restrictors! (Schlenker, 2005; Johnson 2013)This reasoning also led to a number of accounts which constrain thecompetition between pronouns and reflexives in terms of ECONOMYconsiderations. (Safir 2004, 2014; Rooryck & vanden Wyngaerd, 2011; a.o.)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?Inspired by Dowty (1980)’s avoid ambiguity principle and GRICEANreasoning, a number of approaches to the Binding Theory argue infavor of pragmatic constraints regulating the competition betweenhimself and him in ambiguous contexts.→ Rule I (Reinhart, 1984, 2006; Grodzinsky & Reinhart 1993)→ Coreference Rule (Heim, 1993; Roelofsen, 2010)→ Minimize Restrictors! (Schlenker, 2005; Johnson 2013)This reasoning also led to a number of accounts which constrain thecompetition between pronouns and reflexives in terms of ECONOMYconsiderations. (Safir 2004, 2014; Rooryck & vanden Wyngaerd, 2011; a.o.)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?Inspired by Dowty (1980)’s avoid ambiguity principle and GRICEANreasoning, a number of approaches to the Binding Theory argue infavor of pragmatic constraints regulating the competition betweenhimself and him in ambiguous contexts.→ Rule I (Reinhart, 1984, 2006; Grodzinsky & Reinhart 1993)→ Coreference Rule (Heim, 1993; Roelofsen, 2010)→ Minimize Restrictors! (Schlenker, 2005; Johnson 2013)This reasoning also led to a number of accounts which constrain thecompetition between pronouns and reflexives in terms of ECONOMYconsiderations. (Safir 2004, 2014; Rooryck & vanden Wyngaerd, 2011; a.o.)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?Inspired by Dowty (1980)’s avoid ambiguity principle and GRICEANreasoning, a number of approaches to the Binding Theory argue infavor of pragmatic constraints regulating the competition betweenhimself and him in ambiguous contexts.→ Rule I (Reinhart, 1984, 2006; Grodzinsky & Reinhart 1993)→ Coreference Rule (Heim, 1993; Roelofsen, 2010)→ Minimize Restrictors! (Schlenker, 2005; Johnson 2013)This reasoning also led to a number of accounts which constrain thecompetition between pronouns and reflexives in terms of ECONOMYconsiderations. (Safir 2004, 2014; Rooryck & vanden Wyngaerd, 2011; a.o.)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?As the inspiration for many modern binding theories, it istheoretically important to ascertain whether ambiguity avoidanceactually does apply to within-sentence pronouns.However, there is little psycholinguistic evidence that ambiguityavoidance strategies affect the choice of referring expressions inintrasentential contexts.And, in fact, the competition between pronouns and reflexives iscouched in an important larger question ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?As the inspiration for many modern binding theories, it istheoretically important to ascertain whether ambiguity avoidanceactually does apply to within-sentence pronouns.However, there is little psycholinguistic evidence that ambiguityavoidance strategies affect the choice of referring expressions inintrasentential contexts.And, in fact, the competition between pronouns and reflexives iscouched in an important larger question ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionIntrasentential Ambiguity Avoidance?As the inspiration for many modern binding theories, it istheoretically important to ascertain whether ambiguity avoidanceactually does apply to within-sentence pronouns.However, there is little psycholinguistic evidence that ambiguityavoidance strategies affect the choice of referring expressions inintrasentential contexts.And, in fact, the competition between pronouns and reflexives iscouched in an important larger question ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionHow do speakers choose between differentreferring expressions?There is psycholinguistic evidence that the choice of pronouns acrosssentences is steered by ambiguity avoidance.(Arnold et al., 2000; Fukumura et al., 2010; Fukumura et al., 2011; a.o. )Arnold & Griffin (2007)Continue the following story based on the two-panel cartoon.Mickey went for a walk with Daisyin the hills the other day.Mickey went for a walk with Donaldin the hills the other day.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionHow do speakers choose between differentreferring expressions?There is psycholinguistic evidence that the choice of pronouns acrosssentences is steered by ambiguity avoidance.(Arnold et al., 2000; Fukumura et al., 2010; Fukumura et al., 2011; a.o. )Arnold & Griffin (2007)Continue the following story based on the two-panel cartoon.Mickey went for a walk with Daisyin the hills the other day.Mickey went for a walk with Donaldin the hills the other day.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionHow do speakers choose between differentreferring expressions?There is psycholinguistic evidence that the choice of pronouns acrosssentences is steered by ambiguity avoidance.(Arnold et al., 2000; Fukumura et al., 2010; Fukumura et al., 2011; a.o. )Arnold & Griffin (2007)Continue the following story based on the two-panel cartoon.Mickey went for a walk with Daisyin the hills the other day.Mickey went for a walk with Donaldin the hills the other day.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionHow do speakers choose between differentreferring expressions?There is psycholinguistic evidence that the choice of pronouns acrosssentences is steered by ambiguity avoidance.(Arnold et al., 2000; Fukumura et al., 2010; Fukumura et al., 2011; a.o. )Arnold & Griffin (2007)Continue the following story based on the two-panel cartoon.Mickey went for a walk with Daisyin the hills the other day.Mickey went for a walk with Donaldin the hills the other day.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionHow do speakers choose between differentreferring expressions?There is psycholinguistic evidence that the choice of pronouns acrosssentences is steered by ambiguity avoidance.(Arnold et al., 2000; Fukumura et al., 2010; Fukumura et al., 2011; a.o. )Arnold & Griffin (2007)Continue the following story based on the two-panel cartoon.Mickey went for a walk with Daisyin the hills the other day.Mickey went for a walk with Donaldin the hills the other day.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDIFFERENT GENDERMickey went for a walk with Daisyin the hills the other day.SAME GENDERMickey went for a walk with Donaldin the hills the other day.Sample possible continuations:He got tired. /Mickey got tired.Speakers use less pronouns inambiguous contextscross-sententially.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDIFFERENT GENDERMickey went for a walk with Daisyin the hills the other day.SAME GENDERMickey went for a walk with Donaldin the hills the other day.Sample possible continuations:He got tired. /Mickey got tired.Speakers use less pronouns inambiguous contextscross-sententially.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDIFFERENT GENDERMickey went for a walk with Daisyin the hills the other day.SAME GENDERMickey went for a walk with Donaldin the hills the other day.Sample possible continuations:He got tired. /Mickey got tired.Speakers use less pronouns inambiguous contextscross-sententially.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDIFFERENT GENDERMickey went for a walk with Daisyin the hills the other day.SAME GENDERMickey went for a walk with Donaldin the hills the other day.Sample possible continuations:He got tired. /Mickey got tired.Speakers use less pronouns inambiguous contextscross-sententially.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWill this effect extend to within sentence pronouns?Not necessarily.The production literature assumes a division of labor betweenintrasentential and cross-sentential reference. The choice of referenceacross sentences reflects preferences, while reference to a clausemateantecedent is governed by syntactic constraints. (Almor & Nair, 2007)Furthermore, experimental evidence shows that speakers do notalways avoid syntactic ambiguity, like garden path sentences (Arnold etal., 2004; Ferreira & Hudson, 2011; Ferreira & Schotter, 2013; Jaeger, 2010, 2011)While there is evidence in favor of ambiguity avoidancecross-sententially, it is not obvious that the same pressures shouldhold for intrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWill this effect extend to within sentence pronouns?Not necessarily.The production literature assumes a division of labor betweenintrasentential and cross-sentential reference. The choice of referenceacross sentences reflects preferences, while reference to a clausemateantecedent is governed by syntactic constraints. (Almor & Nair, 2007)Furthermore, experimental evidence shows that speakers do notalways avoid syntactic ambiguity, like garden path sentences (Arnold etal., 2004; Ferreira & Hudson, 2011; Ferreira & Schotter, 2013; Jaeger, 2010, 2011)While there is evidence in favor of ambiguity avoidancecross-sententially, it is not obvious that the same pressures shouldhold for intrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWill this effect extend to within sentence pronouns?Not necessarily.The production literature assumes a division of labor betweenintrasentential and cross-sentential reference.The choice of referenceacross sentences reflects preferences, while reference to a clausemateantecedent is governed by syntactic constraints. (Almor & Nair, 2007)Furthermore, experimental evidence shows that speakers do notalways avoid syntactic ambiguity, like garden path sentences (Arnold etal., 2004; Ferreira & Hudson, 2011; Ferreira & Schotter, 2013; Jaeger, 2010, 2011)While there is evidence in favor of ambiguity avoidancecross-sententially, it is not obvious that the same pressures shouldhold for intrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWill this effect extend to within sentence pronouns?Not necessarily.The production literature assumes a division of labor betweenintrasentential and cross-sentential reference. The choice of referenceacross sentences reflects preferences, while reference to a clausemateantecedent is governed by syntactic constraints. (Almor & Nair, 2007)Furthermore, experimental evidence shows that speakers do notalways avoid syntactic ambiguity, like garden path sentences (Arnold etal., 2004; Ferreira & Hudson, 2011; Ferreira & Schotter, 2013; Jaeger, 2010, 2011)While there is evidence in favor of ambiguity avoidancecross-sententially, it is not obvious that the same pressures shouldhold for intrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWill this effect extend to within sentence pronouns?Not necessarily.The production literature assumes a division of labor betweenintrasentential and cross-sentential reference. The choice of referenceacross sentences reflects preferences, while reference to a clausemateantecedent is governed by syntactic constraints. (Almor & Nair, 2007)Furthermore, experimental evidence shows that speakers do notalways avoid syntactic ambiguity, like garden path sentences (Arnold etal., 2004; Ferreira & Hudson, 2011; Ferreira & Schotter, 2013; Jaeger, 2010, 2011)While there is evidence in favor of ambiguity avoidancecross-sententially, it is not obvious that the same pressures shouldhold for intrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWill this effect extend to within sentence pronouns?Not necessarily.The production literature assumes a division of labor betweenintrasentential and cross-sentential reference. The choice of referenceacross sentences reflects preferences, while reference to a clausemateantecedent is governed by syntactic constraints. (Almor & Nair, 2007)Furthermore, experimental evidence shows that speakers do notalways avoid syntactic ambiguity, like garden path sentences (Arnold etal., 2004; Ferreira & Hudson, 2011; Ferreira & Schotter, 2013; Jaeger, 2010, 2011)While there is evidence in favor of ambiguity avoidancecross-sententially, it is not obvious that the same pressures shouldhold for intrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionHypothesisRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningQuestion: Is there evidence that GRICEAN reasoning gives rise todisjoint reference effects for regular pronouns in Condition Benvironments?Inspired by Dowty (1980), let’s assume that the following constraint isemployed as a general ambiguity avoidance strategy:(7) BE CLEAR!When choosing between two alternative sentences, S and S’,speak S’ iff:i. S and S’ have indistinguishable interpretations in a context C,andii. the set of possible interpretations for S’ is a proper subsetof the set of possible interpretations for S.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningQuestion: Is there evidence that GRICEAN reasoning gives rise todisjoint reference effects for regular pronouns in Condition Benvironments?Inspired by Dowty (1980), let’s assume that the following constraint isemployed as a general ambiguity avoidance strategy:(7) BE CLEAR!When choosing between two alternative sentences, S and S’,speak S’ iff:i. S and S’ have indistinguishable interpretations in a context C,andii. the set of possible interpretations for S’ is a proper subsetof the set of possible interpretations for S.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningQuestion: Is there evidence that GRICEAN reasoning gives rise todisjoint reference effects for regular pronouns in Condition Benvironments?Inspired by Dowty (1980), let’s assume that the following constraint isemployed as a general ambiguity avoidance strategy:(7) BE CLEAR!When choosing between two alternative sentences, S and S’,speak S’ iff:i. S and S’ have indistinguishable interpretations in a context C,andii. the set of possible interpretations for S’ is a proper subsetof the set of possible interpretations for S.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningQuestion: Is there evidence that GRICEAN reasoning gives rise todisjoint reference effects for regular pronouns in Condition Benvironments?Inspired by Dowty (1980), let’s assume that the following constraint isemployed as a general ambiguity avoidance strategy:(7) BE CLEAR!When choosing between two alternative sentences, S and S’,speak S’ iff:i. S and S’ have indistinguishable interpretations in a context C,andii. the set of possible interpretations for S’ is a proper subsetof the set of possible interpretations for S.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningQuestion: Is there evidence that GRICEAN reasoning gives rise todisjoint reference effects for regular pronouns in Condition Benvironments?Inspired by Dowty (1980), let’s assume that the following constraint isemployed as a general ambiguity avoidance strategy:(7) BE CLEAR!When choosing between two alternative sentences, S and S’,speak S’ iff:i. S and S’ have indistinguishable interpretations in a context C,andii. the set of possible interpretations for S’ is a proper subsetof the set of possible interpretations for S.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningWe’ve seen BE CLEAR! at work in English Condition B environments.(8) CONTEXT: Lockhart admires Lockhart.a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSIt works similarly in the case of Romanian clitics.(9) CONTEXT: Lockhart admires Lockhart.a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. #Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningWe’ve seen BE CLEAR! at work in English Condition B environments.(8) CONTEXT: Lockhart admires Lockhart.a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSIt works similarly in the case of Romanian clitics.(9) CONTEXT: Lockhart admires Lockhart.a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. #Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGricean ReasoningWe’ve seen BE CLEAR! at work in English Condition B environments.(8) CONTEXT: Lockhart admires Lockhart.a. Lockhart admires himself. UNAMBIGUOUSb. # Lockhart admires him. AMBIGUOUSIt works similarly in the case of Romanian clitics.(9) CONTEXT: Lockhart admires Lockhart.a. Lockhart1Lockhartse1/∗2REFL.CLadmirăadmires(pe(ACCel1/∗2).him)Lockhart admires himself.b. #Lockhart1Lockhartıˆl2/∗1MASC.SG.CLadmirăadmires(pe(ACCel2/∗1).him)Lockhart admires someone else.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsAn important insight (Heim, 1982; Reinhart, 1983): binding must bedistinguished from coreference. Rule I, the Coreference Rule andsimilar constraints were designed to account for disjoint referenceeffectswith a referential antecedent (one that does not undergo QR).(10) a. * Lockhart1 λ1 [t1 talked about him1]. CONDITION Bb. Lockhart talked about him. RULED OUT BY RULE IThese constraints stipulate preference for bound variables overcoreference. A similar consequence is obtained by syntactic-basedcompetition accounts (Safir 2004, 2014; Rooryck & vanden Wyngaerd 2011; a.o.).Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsAn important insight (Heim, 1982; Reinhart, 1983): binding must bedistinguished from coreference.Rule I, the Coreference Rule andsimilar constraints were designed to account for disjoint referenceeffectswith a referential antecedent (one that does not undergo QR).(10) a. * Lockhart1 λ1 [t1 talked about him1]. CONDITION Bb. Lockhart talked about him. RULED OUT BY RULE IThese constraints stipulate preference for bound variables overcoreference. A similar consequence is obtained by syntactic-basedcompetition accounts (Safir 2004, 2014; Rooryck & vanden Wyngaerd 2011; a.o.).Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsAn important insight (Heim, 1982; Reinhart, 1983): binding must bedistinguished from coreference. Rule I, the Coreference Rule andsimilar constraints were designed to account for disjoint referenceeffectswith a referential antecedent (one that does not undergo QR).(10) a. * Lockhart1 λ1 [t1 talked about him1]. CONDITION Bb. Lockhart talked about him. RULED OUT BY RULE IThese constraints stipulate preference for bound variables overcoreference. A similar consequence is obtained by syntactic-basedcompetition accounts (Safir 2004, 2014; Rooryck & vanden Wyngaerd 2011; a.o.).Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsAn important insight (Heim, 1982; Reinhart, 1983): binding must bedistinguished from coreference. Rule I, the Coreference Rule andsimilar constraints were designed to account for disjoint referenceeffectswith a referential antecedent (one that does not undergo QR).(10) a. * Lockhart1 λ1 [t1 talked about him1]. CONDITION Bb. Lockhart talked about him. RULED OUT BY RULE IThese constraints stipulate preference for bound variables overcoreference. A similar consequence is obtained by syntactic-basedcompetition accounts (Safir 2004, 2014; Rooryck & vanden Wyngaerd 2011; a.o.).Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsAn important insight (Heim, 1982; Reinhart, 1983): binding must bedistinguished from coreference. Rule I, the Coreference Rule andsimilar constraints were designed to account for disjoint referenceeffectswith a referential antecedent (one that does not undergo QR).(10) a. * Lockhart1 λ1 [t1 talked about him1]. CONDITION Bb. Lockhart talked about him. RULED OUT BY RULE IThese constraints stipulate preference for bound variables overcoreference.A similar consequence is obtained by syntactic-basedcompetition accounts (Safir 2004, 2014; Rooryck & vanden Wyngaerd 2011; a.o.).Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsAn important insight (Heim, 1982; Reinhart, 1983): binding must bedistinguished from coreference. Rule I, the Coreference Rule andsimilar constraints were designed to account for disjoint referenceeffectswith a referential antecedent (one that does not undergo QR).(10) a. * Lockhart1 λ1 [t1 talked about him1]. CONDITION Bb. Lockhart talked about him. RULED OUT BY RULE IThese constraints stipulate preference for bound variables overcoreference. A similar consequence is obtained by syntactic-basedcompetition accounts (Safir 2004, 2014; Rooryck & vanden Wyngaerd 2011; a.o.).Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsBy design, constraints like Rule I and the Coreference Rule do NOTpredict that the form of a bound variable should be affected bycontextual ambiguity.According to BE CLEAR!, regular pronouns should be disfavored inambiguous contexts, irrespective of the type of antecedent: referentialor quantificational.The experiments target this difference.Experimental Question:Is the rate of production of pronouns affected by context ambiguity indisjoint reference, local coreference and bound variable contexts?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsBy design, constraints like Rule I and the Coreference Rule do NOTpredict that the form of a bound variable should be affected bycontextual ambiguity.According to BE CLEAR!, regular pronouns should be disfavored inambiguous contexts, irrespective of the type of antecedent: referentialor quantificational.The experiments target this difference.Experimental Question:Is the rate of production of pronouns affected by context ambiguity indisjoint reference, local coreference and bound variable contexts?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsBy design, constraints like Rule I and the Coreference Rule do NOTpredict that the form of a bound variable should be affected bycontextual ambiguity.According to BE CLEAR!, regular pronouns should be disfavored inambiguous contexts, irrespective of the type of antecedent: referentialor quantificational.The experiments target this difference.Experimental Question:Is the rate of production of pronouns affected by context ambiguity indisjoint reference, local coreference and bound variable contexts?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsBy design, constraints like Rule I and the Coreference Rule do NOTpredict that the form of a bound variable should be affected bycontextual ambiguity.According to BE CLEAR!, regular pronouns should be disfavored inambiguous contexts, irrespective of the type of antecedent: referentialor quantificational.The experiments target this difference.Experimental Question:Is the rate of production of pronouns affected by context ambiguity indisjoint reference, local coreference and bound variable contexts?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsBy design, constraints like Rule I and the Coreference Rule do NOTpredict that the form of a bound variable should be affected bycontextual ambiguity.According to BE CLEAR!, regular pronouns should be disfavored inambiguous contexts, irrespective of the type of antecedent: referentialor quantificational.The experiments target this difference.Experimental Question:Is the rate of production of pronouns affected by context ambiguity indisjoint reference, local coreference and bound variable contexts?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential vs. Quantified SubjectsBy design, constraints like Rule I and the Coreference Rule do NOTpredict that the form of a bound variable should be affected bycontextual ambiguity.According to BE CLEAR!, regular pronouns should be disfavored inambiguous contexts, irrespective of the type of antecedent: referentialor quantificational.The experiments target this difference.Experimental Question:Is the rate of production of pronouns affected by context ambiguity indisjoint reference, local coreference and bound variable contexts?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionReferential Subjects & Quantified SubjectsMain Questions:1 Is the rate of pronoun production affected by ambiguity in thecase of coreference / variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Hypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns inambiguous contexts.• more names in ambiguousdisjoint contexts.• more reflexive pronouns inambiguous reflexive contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionExperiment 1: Referential SubjectsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN• picture description task• 2 x 2 design: PICTURE TYPE x AMBIGUITYLocal Coreferent/Local Disjoint x Character Gender Match/Mismatch• 16 items (distributed in 4 Latin Squared lists)• only 2 referents in the context per item• 20 fillersPARTICIPANTS• 68 participants (62 female), University of Bucharest students• The age range was between 18 and 30, with an average age of 20.4• reimbursed 30 RON (≈ 8 USD) for participationRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN• picture description task• 2 x 2 design: PICTURE TYPE x AMBIGUITYLocal Coreferent/Local Disjoint x Character Gender Match/Mismatch• 16 items (distributed in 4 Latin Squared lists)• only 2 referents in the context per item• 20 fillersPARTICIPANTS• 68 participants (62 female), University of Bucharest students• The age range was between 18 and 30, with an average age of 20.4• reimbursed 30 RON (≈ 8 USD) for participationRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN• picture description task• 2 x 2 design: PICTURE TYPE x AMBIGUITYLocal Coreferent/Local Disjoint x Character Gender Match/Mismatch• 16 items (distributed in 4 Latin Squared lists)• only 2 referents in the context per item• 20 fillersPARTICIPANTS• 68 participants (62 female), University of Bucharest students• The age range was between 18 and 30, with an average age of 20.4• reimbursed 30 RON (≈ 8 USD) for participationRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic,Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniellaughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext (introducing the referents in random order).• Each target sentence consists of:→ a topic PP which names one of the referents→ an overt subject which refers to the other referent by name→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Monica’s picnic, Daniel laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSAMPLE ITEMThis is Andrei. This is Mihai.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSAMPLE ITEMThis is Andrei. This is Mihai.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionLOCAL COREFERENT, CHARACTER GENDER MATCHAt Mihai’s house, Andrei talked about ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSample Item Pictures & Target Sentences by ConditionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSample Item Pictures & Target Sentences by ConditionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionParticipant Response TypesOn Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Mihai, Andrei, Irina• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: heremotional states, his friend, hisglasses, etc.• random NPs: love, magicalpowers, girls• full sentences: how he feels,what happened last night, howthey met at a restaurant 10years ago, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionData Exclusion• excluded non-target responses• excluded data from 2 participants due to a low rate of targetresponses (<30%)• lost 15 responses due to a PsychoPy error• in total, 10.81% of responses were removed• analysis ran on 957 target responses out of a total of 1073Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Coreferent/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionData Exclusion• excluded non-target responses• excluded data from 2 participants due to a low rate of targetresponses (<30%)• lost 15 responses due to a PsychoPy error• in total, 10.81% of responses were removed• analysis ran on 957 target responses out of a total of 1073Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Coreferent/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionData Exclusion• excluded non-target responses• excluded data from 2 participants due to a low rate of targetresponses (<30%)• lost 15 responses due to a PsychoPy error• in total, 10.81% of responses were removed• analysis ran on 957 target responses out of a total of 1073Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Coreferent/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionData Exclusion• excluded non-target responses• excluded data from 2 participants due to a low rate of targetresponses (<30%)• lost 15 responses due to a PsychoPy error• in total, 10.81% of responses were removed• analysis ran on 957 target responses out of a total of 1073Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Coreferent/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionData Exclusion• excluded non-target responses• excluded data from 2 participants due to a low rate of targetresponses (<30%)• lost 15 responses due to a PsychoPy error• in total, 10.81% of responses were removed• analysis ran on 957 target responses out of a total of 1073Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Coreferent/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionData Exclusion• excluded non-target responses• excluded data from 2 participants due to a low rate of targetresponses (<30%)• lost 15 responses due to a PsychoPy error• in total, 10.81% of responses were removed• analysis ran on 957 target responses out of a total of 1073Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Coreferent/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRaw ResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRaw ResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRaw ResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRaw ResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRaw ResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects both coreference and disjoint reference contexts.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 26.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous coreference contexts.X a 16.9% increase from LOCAL COREFERENT MISMATCH to LOCALCOREFERENT MATCH (p<0.01)Does BE CLEAR! impact the choice of pronominal form for locallybound variables as well?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionExperiment 2: Quantified SubjectsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN, MATERIALS, ANALYSISThe hypothesis is that BE CLEAR! is a generic ambiguity avoidanceconstraint: the same predictions are made for referential expressionstargeting a local referential subject or a local quantified subject.Experiment 2 is a replication of Experiment 1: the same design,procedure, data annotation and analysis were used.The relevant differences:• item subjects are quantified expressions - like every boy• there are 4 referents per context per itemRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN, MATERIALS, ANALYSISThe hypothesis is that BE CLEAR! is a generic ambiguity avoidanceconstraint:the same predictions are made for referential expressionstargeting a local referential subject or a local quantified subject.Experiment 2 is a replication of Experiment 1: the same design,procedure, data annotation and analysis were used.The relevant differences:• item subjects are quantified expressions - like every boy• there are 4 referents per context per itemRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN, MATERIALS, ANALYSISThe hypothesis is that BE CLEAR! is a generic ambiguity avoidanceconstraint: the same predictions are made for referential expressionstargeting a local referential subject or a local quantified subject.Experiment 2 is a replication of Experiment 1: the same design,procedure, data annotation and analysis were used.The relevant differences:• item subjects are quantified expressions - like every boy• there are 4 referents per context per itemRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN, MATERIALS, ANALYSISThe hypothesis is that BE CLEAR! is a generic ambiguity avoidanceconstraint: the same predictions are made for referential expressionstargeting a local referential subject or a local quantified subject.Experiment 2 is a replication of Experiment 1: the same design,procedure, data annotation and analysis were used.The relevant differences:• item subjects are quantified expressions - like every boy• there are 4 referents per context per itemRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDESIGN, MATERIALS, ANALYSISThe hypothesis is that BE CLEAR! is a generic ambiguity avoidanceconstraint: the same predictions are made for referential expressionstargeting a local referential subject or a local quantified subject.Experiment 2 is a replication of Experiment 1: the same design,procedure, data annotation and analysis were used.The relevant differences:• item subjects are quantified expressions - like every boy• there are 4 referents per context per itemRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSAMPLE ITEM(11) Context Screen:Grandma Laura was recently visited by her family.Monica, Elena and Irina were there too.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSAMPLE ITEM(11) Context Screen:Grandma Laura was recently visited by her family.Monica, Elena and Irina were there too.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionLOCAL BOUND, CHARACTER GENDER MATCHAt Grandma Laura’s house, every girl talked about ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSample Item Pictures & Target Sentences by ConditionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionSample Item Pictures & Target Sentences by ConditionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionRaw ResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionResultsRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionDiscussionHypothesis:BE CLEAR! affects coreferent, disjoint and locally bound variables.Predictions:• less regular pronouns in ambiguous contexts.X overall effect of AMBIGUITY (p<0.001)• more names in ambiguous disjoint contexts.X a 25.9% increase from LOCAL DISJOINT MISMATCH to LOCALDISJOINT MATCH (p<0.001)• more reflexive pronouns in ambiguous bound contexts.X a 18.6% increase from LOCAL BOUND MISMATCH to LOCALBOUND MATCH (p<0.01)An ambiguity avoidance (BE CLEAR!) constraint impacts the choiceof pronominal form in all three contexts: local disjoint reference, localcoreference and locally bound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionGeneral DiscussionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about Romanian pronouns?• Pronouns do not obeyCondition B, but their rate ofproduction is affected bycontext ambiguity.• In unambiguous coreferent andbound contexts, the preferredpronominal form is the regularpronoun (around 50%), and notreflexive expressions.• In ambiguous coreferent andbound variable contexts, theemphatic reflexive is thepreferred pronominal form.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about Romanian pronouns?• Pronouns do not obeyCondition B, but their rate ofproduction is affected bycontext ambiguity.• In unambiguous coreferent andbound contexts, the preferredpronominal form is the regularpronoun (around 50%), and notreflexive expressions.• In ambiguous coreferent andbound variable contexts, theemphatic reflexive is thepreferred pronominal form.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about Romanian pronouns?• Pronouns do not obeyCondition B, but their rate ofproduction is affected bycontext ambiguity.• In unambiguous coreferent andbound contexts, the preferredpronominal form is the regularpronoun (around 50%), and notreflexive expressions.• In ambiguous coreferent andbound variable contexts, theemphatic reflexive is thepreferred pronominal form.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about Romanian pronouns?• Pronouns do not obeyCondition B, but their rate ofproduction is affected bycontext ambiguity.• In unambiguous coreferent andbound contexts, the preferredpronominal form is the regularpronoun (around 50%), and notreflexive expressions.• In ambiguous coreferent andbound variable contexts, theemphatic reflexive is thepreferred pronominal form.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about Romanian pronouns?• Pronouns do not obeyCondition B, but their rate ofproduction is affected bycontext ambiguity.• In unambiguous coreferent andbound contexts, the preferredpronominal form is the regularpronoun (around 50%), and notreflexive expressions.• In ambiguous coreferent andbound variable contexts, theemphatic reflexive is thepreferred pronominal form.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about ambiguity avoidance?• Ever since Dowty (1980), there has been a long-debated intuitionthat ambiguity avoidancemodulates the competition betweenreflexive and non-reflexive pronouns.• Despite a large body of work in the psycholinguistic literaturethat ambiguity avoidance affects the choice of referringexpressions cross-sententially, there was a lack of evidence thatthe same strategy applies in intrasentential contexts.• The results from Experiment 1 and Experiment 2 show thatambiguity avoidance strategies affect the production rate ofregular pronouns for both coreference and bound variables inintrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about ambiguity avoidance?• Ever since Dowty (1980), there has been a long-debated intuitionthat ambiguity avoidancemodulates the competition betweenreflexive and non-reflexive pronouns.• Despite a large body of work in the psycholinguistic literaturethat ambiguity avoidance affects the choice of referringexpressions cross-sententially, there was a lack of evidence thatthe same strategy applies in intrasentential contexts.• The results from Experiment 1 and Experiment 2 show thatambiguity avoidance strategies affect the production rate ofregular pronouns for both coreference and bound variables inintrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about ambiguity avoidance?• Ever since Dowty (1980), there has been a long-debated intuitionthat ambiguity avoidancemodulates the competition betweenreflexive and non-reflexive pronouns.• Despite a large body of work in the psycholinguistic literaturethat ambiguity avoidance affects the choice of referringexpressions cross-sententially, there was a lack of evidence thatthe same strategy applies in intrasentential contexts.• The results from Experiment 1 and Experiment 2 show thatambiguity avoidance strategies affect the production rate ofregular pronouns for both coreference and bound variables inintrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about ambiguity avoidance?• Ever since Dowty (1980), there has been a long-debated intuitionthat ambiguity avoidancemodulates the competition betweenreflexive and non-reflexive pronouns.• Despite a large body of work in the psycholinguistic literaturethat ambiguity avoidance affects the choice of referringexpressions cross-sententially, there was a lack of evidence thatthe same strategy applies in intrasentential contexts.• The results from Experiment 1 and Experiment 2 show thatambiguity avoidance strategies affect the production rate ofregular pronouns for both coreference and bound variables inintrasentential contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionWhat have we learned about BE CLEAR! ?• BE CLEAR! is substantiated.• The experiments provided clear evidence of a genericintrasentential ambiguity avoidance principle.• This is a violable, preferential constraint in Romanian.• The effect was qualitatively the same as the ambiguity avoidanceeffect found in the psycholinguistic literature for cross-sententialpronouns.• The effect was of the same magnitude for local coreference(Experiment 1) and bound variable contexts (Experiment 2).Question: Why is the regular pronoun el/ea the preferred form inunambiguous contexts?Ask me about it!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionFurther Questions• Is there pyscholinguistic evidence in favor of competitionbetween forms? (Yes!!)• Should we expect BE CLEAR! to affect pronoun production inother languages? (Of course!)• Is the data from comprehension consistent with the data fromthe production experiments? (Yes!)• so many other questions...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Background Hypothesis Experiment 1: Referential Subjects Experiment 2: Quantified Subjects General DiscussionThank You!Acknowledgements: We are grateful for all of the feedback, support andencouragement of Kyle Johnson, Seth Cable, Isabelle Charnavel, AlexandraCornilescu, Marcel den Dikken, Lyn Frazier, Ken Safir, Dominique Sportiche,Adrian Staub, the audiences ofWCCFL 36 @ UCLA (April 2018), Pronouns inCompetition @ Santa Cruz (April 2018), and many others we’re bound toforget. We are also very thankful to the NSF, the University of Bucharest, andthe UMass Amherst Predissertation Grant for their support.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionRule IRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic competitionLike its predecessors, BE CLEAR! is a pragmatic constraint whichcompares sentences that have the same meaning in a given context.(12) RULE I: INTRASENTENTIAL COREFERENCENP A cannot corefer with NP B if replacing A with C, avariable A-bound by B, yields an indistinguishable interpretation.x (Grodzinsky & Reinhart, 1993, p.79, ex. (20))(13) COREFERENCE RULEA speaker will never use a logical form LF in a context C if theLF is semantically indistinguishable from one of its bindingalternatives.x (Roelofsen, 2010, p.119)Unlike its predecessors, BE CLEAR! is a general ambiguity avoidancestrategy: it does not favor bound variables over unbound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic competitionLike its predecessors, BE CLEAR! is a pragmatic constraint whichcompares sentences that have the same meaning in a given context.(12) RULE I: INTRASENTENTIAL COREFERENCENP A cannot corefer with NP B if replacing A with C, avariable A-bound by B, yields an indistinguishable interpretation.x (Grodzinsky & Reinhart, 1993, p.79, ex. (20))(13) COREFERENCE RULEA speaker will never use a logical form LF in a context C if theLF is semantically indistinguishable from one of its bindingalternatives.x (Roelofsen, 2010, p.119)Unlike its predecessors, BE CLEAR! is a general ambiguity avoidancestrategy: it does not favor bound variables over unbound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic competitionLike its predecessors, BE CLEAR! is a pragmatic constraint whichcompares sentences that have the same meaning in a given context.(12) RULE I: INTRASENTENTIAL COREFERENCENP A cannot corefer with NP B if replacing A with C, avariable A-bound by B, yields an indistinguishable interpretation.x (Grodzinsky & Reinhart, 1993, p.79, ex. (20))(13) COREFERENCE RULEA speaker will never use a logical form LF in a context C if theLF is semantically indistinguishable from one of its bindingalternatives.x (Roelofsen, 2010, p.119)Unlike its predecessors, BE CLEAR! is a general ambiguity avoidancestrategy: it does not favor bound variables over unbound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic competitionLike its predecessors, BE CLEAR! is a pragmatic constraint whichcompares sentences that have the same meaning in a given context.(12) RULE I: INTRASENTENTIAL COREFERENCENP A cannot corefer with NP B if replacing A with C, avariable A-bound by B, yields an indistinguishable interpretation.x (Grodzinsky & Reinhart, 1993, p.79, ex. (20))(13) COREFERENCE RULEA speaker will never use a logical form LF in a context C if theLF is semantically indistinguishable from one of its bindingalternatives.x (Roelofsen, 2010, p.119)Unlike its predecessors, BE CLEAR! is a general ambiguity avoidancestrategy: it does not favor bound variables over unbound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic competitionLike its predecessors, BE CLEAR! is a pragmatic constraint whichcompares sentences that have the same meaning in a given context.(12) RULE I: INTRASENTENTIAL COREFERENCENP A cannot corefer with NP B if replacing A with C, avariable A-bound by B, yields an indistinguishable interpretation.x (Grodzinsky & Reinhart, 1993, p.79, ex. (20))(13) COREFERENCE RULEA speaker will never use a logical form LF in a context C if theLF is semantically indistinguishable from one of its bindingalternatives.x (Roelofsen, 2010, p.119)Unlike its predecessors, BE CLEAR! is a general ambiguity avoidancestrategy: it does not favor bound variables over unbound variables.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionReferential vs. Quantified SubjectsRule I and the Coreference Rule assume Condition B:pronouns cannot be bound by an antecedent in the same clause.(14) a. *Every boy1 λ1 [t1 talked about him1].b. *Lockhart1 λ1 [t1 talked about him1].c. RULED OUT BY CONDITION BTPTPvPvPtalked about him1VPvt1Tλ1every boy1 / Lockhart1DPRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionReferential vs. Quantified SubjectsRule I and the Coreference Rule assume Condition B:pronouns cannot be bound by an antecedent in the same clause.(14) a. *Every boy1 λ1 [t1 talked about him1].b. *Lockhart1 λ1 [t1 talked about him1].c. RULED OUT BY CONDITION BTPTPvPvPtalked about him1VPvt1Tλ1every boy1 / Lockhart1DPRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionReferential vs. Quantified SubjectsRule I and the Coreference Rule assume Condition B:pronouns cannot be bound by an antecedent in the same clause.(14) a. *Every boy1 λ1 [t1 talked about him1].b. *Lockhart1 λ1 [t1 talked about him1].c. RULED OUT BY CONDITION BTPTPvPvPtalked about him1VPvt1Tλ1every boy1 / Lockhart1DPRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionReferential vs. Quantified SubjectsRule I and the Coreference Rule assume Condition B:pronouns cannot be bound by an antecedent in the same clause.(14) a. *Every boy1 λ1 [t1 talked about him1].b. *Lockhart1 λ1 [t1 talked about him1].c. RULED OUT BY CONDITION BTPTPvPvPtalked about him1VPvt1Tλ1every boy1 / Lockhart1DPRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionReferential vs. Quantified SubjectsRule I and the Coreference Rule assume Condition B:pronouns cannot be bound by an antecedent in the same clause.(14) a. *Every boy1 λ1 [t1 talked about him1].b. *Lockhart1 λ1 [t1 talked about him1].c. RULED OUT BY CONDITION BTPTPvPvPtalked about him1VPvt1Tλ1every boy1 / Lockhart1DPRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionReferential vs. Quantified SubjectsRule I and the Coreference Rule assume Condition B:pronouns cannot be bound by an antecedent in the same clause.(14) a. *Every boy1 λ1 [t1 talked about him1].b. *Lockhart1 λ1 [t1 talked about him1].c. RULED OUT BY CONDITION BTPTPvPvPtalked about him1VPvt1Tλ1every boy1 / Lockhart1DPRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBe Small!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic and Syntactic ConsiderationsPragmatic competition based accounts of the Binding Theory (Reinhart,1983, 2006; Roelofsen, 2010) argue that a pragmatic constraint restricts thedistribution of pronouns in Condition B environments.Syntactic competition based accounts (Safir, 2004, 2014; Reuland, 2011;Rooryck & vanden Wyngaerd, 2011) propose that the choice of reflexivepronouns over non-reflexive pronouns in Condition B environmentshas to do with economy considerations: smaller forms are preferred.We suggest a generic economy constraint (inspired byMinimizeRestrictors, Schlenker, 2005) BE SMALL and the generic pragmaticconstraint BE CLEAR!! jointly determine the distribution of pronounsin Romanian. The experiments corroborate this assumption.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic and Syntactic ConsiderationsPragmatic competition based accounts of the Binding Theory (Reinhart,1983, 2006; Roelofsen, 2010) argue that a pragmatic constraint restricts thedistribution of pronouns in Condition B environments.Syntactic competition based accounts (Safir, 2004, 2014; Reuland, 2011;Rooryck & vanden Wyngaerd, 2011) propose that the choice of reflexivepronouns over non-reflexive pronouns in Condition B environmentshas to do with economy considerations: smaller forms are preferred.We suggest a generic economy constraint (inspired byMinimizeRestrictors, Schlenker, 2005) BE SMALL and the generic pragmaticconstraint BE CLEAR!! jointly determine the distribution of pronounsin Romanian. The experiments corroborate this assumption.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic and Syntactic ConsiderationsPragmatic competition based accounts of the Binding Theory (Reinhart,1983, 2006; Roelofsen, 2010) argue that a pragmatic constraint restricts thedistribution of pronouns in Condition B environments.Syntactic competition based accounts (Safir, 2004, 2014; Reuland, 2011;Rooryck & vanden Wyngaerd, 2011) propose that the choice of reflexivepronouns over non-reflexive pronouns in Condition B environmentshas to do with economy considerations: smaller forms are preferred.We suggest a generic economy constraint (inspired byMinimizeRestrictors, Schlenker, 2005) BE SMALL and the generic pragmaticconstraint BE CLEAR!! jointly determine the distribution of pronounsin Romanian. The experiments corroborate this assumption.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic and Syntactic ConsiderationsPragmatic competition based accounts of the Binding Theory (Reinhart,1983, 2006; Roelofsen, 2010) argue that a pragmatic constraint restricts thedistribution of pronouns in Condition B environments.Syntactic competition based accounts (Safir, 2004, 2014; Reuland, 2011;Rooryck & vanden Wyngaerd, 2011) propose that the choice of reflexivepronouns over non-reflexive pronouns in Condition B environmentshas to do with economy considerations: smaller forms are preferred.We suggest a generic economy constraint (inspired byMinimizeRestrictors, Schlenker, 2005) BE SMALL and the generic pragmaticconstraint BE CLEAR!! jointly determine the distribution of pronounsin Romanian.The experiments corroborate this assumption.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPragmatic and Syntactic ConsiderationsPragmatic competition based accounts of the Binding Theory (Reinhart,1983, 2006; Roelofsen, 2010) argue that a pragmatic constraint restricts thedistribution of pronouns in Condition B environments.Syntactic competition based accounts (Safir, 2004, 2014; Reuland, 2011;Rooryck & vanden Wyngaerd, 2011) propose that the choice of reflexivepronouns over non-reflexive pronouns in Condition B environmentshas to do with economy considerations: smaller forms are preferred.We suggest a generic economy constraint (inspired byMinimizeRestrictors, Schlenker, 2005) BE SMALL and the generic pragmaticconstraint BE CLEAR!! jointly determine the distribution of pronounsin Romanian. The experiments corroborate this assumption.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionEconomy and Grice AgainThe preference for smaller forms is also predicted by Gricean Maxims(the inspiration forMinimize Restrictors - Schlenker, 2005).(15) (TWO) GRICEAN MAXIMS (1975)a. Maxim of MannerAvoid ambiguity. BE CLEAR!b. Maxim of QuantityMake your contribution as informative as is required (forthe current purposes of the exchange).Do not make your contribution more informative than isrequired. BE SMALL!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionEconomy and Grice AgainThe preference for smaller forms is also predicted by Gricean Maxims(the inspiration forMinimize Restrictors - Schlenker, 2005).(15) (TWO) GRICEAN MAXIMS (1975)a. Maxim of MannerAvoid ambiguity. BE CLEAR!b. Maxim of QuantityMake your contribution as informative as is required (forthe current purposes of the exchange).Do not make your contribution more informative than isrequired. BE SMALL!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionEconomy and Grice AgainThe preference for smaller forms is also predicted by Gricean Maxims(the inspiration forMinimize Restrictors - Schlenker, 2005).(15) (TWO) GRICEAN MAXIMS (1975)a. Maxim of MannerAvoid ambiguity. BE CLEAR!b. Maxim of QuantityMake your contribution as informative as is required (forthe current purposes of the exchange).Do not make your contribution more informative than isrequired. BE SMALL!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionEconomy and Grice AgainThe preference for smaller forms is also predicted by Gricean Maxims(the inspiration forMinimize Restrictors - Schlenker, 2005).(15) (TWO) GRICEAN MAXIMS (1975)a. Maxim of MannerAvoid ambiguity. BE CLEAR!b. Maxim of QuantityMake your contribution as informative as is required (forthe current purposes of the exchange).Do not make your contribution more informative than isrequired. BE SMALL!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionEconomy and Grice AgainThe preference for smaller forms is also predicted by Gricean Maxims(the inspiration forMinimize Restrictors - Schlenker, 2005).(15) (TWO) GRICEAN MAXIMS (1975)a. Maxim of MannerAvoid ambiguity. BE CLEAR!b. Maxim of QuantityMake your contribution as informative as is required (forthe current purposes of the exchange).Do not make your contribution more informative than isrequired.BE SMALL!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionEconomy and Grice AgainThe preference for smaller forms is also predicted by Gricean Maxims(the inspiration forMinimize Restrictors - Schlenker, 2005).(15) (TWO) GRICEAN MAXIMS (1975)a. Maxim of MannerAvoid ambiguity. BE CLEAR!b. Maxim of QuantityMake your contribution as informative as is required (forthe current purposes of the exchange).Do not make your contribution more informative than isrequired. BE SMALL!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBe Small!(16) BE SMALL!When choosing between two alternative phrases, XP and XP’,speak XP’ iff:i. XP and XP’ have indistinguishable interpretations in a contextC, andii. the set of morphosyntactic features of XP’ is a proper sub-set of the set of morphosyntactic features of XP.Prediction: regular pronouns like el are preferred to emphaticreflexives like el ıˆnsus¸iwhen they achieve the same meaning.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBe Small!(16) BE SMALL!When choosing between two alternative phrases, XP and XP’,speak XP’ iff:i. XP and XP’ have indistinguishable interpretations in a contextC, andii. the set of morphosyntactic features of XP’ is a proper sub-set of the set of morphosyntactic features of XP.Prediction: regular pronouns like el are preferred to emphaticreflexives like el ıˆnsus¸iwhen they achieve the same meaning.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBe Small!(16) BE SMALL!When choosing between two alternative phrases, XP and XP’,speak XP’ iff:i. XP and XP’ have indistinguishable interpretations in a contextC, andii. the set of morphosyntactic features of XP’ is a proper sub-set of the set of morphosyntactic features of XP.Prediction: regular pronouns like el are preferred to emphaticreflexives like el ıˆnsus¸iwhen they achieve the same meaning.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBe Small!(16) BE SMALL!When choosing between two alternative phrases, XP and XP’,speak XP’ iff:i. XP and XP’ have indistinguishable interpretations in a contextC, andii. the set of morphosyntactic features of XP’ is a proper sub-set of the set of morphosyntactic features of XP.Prediction: regular pronouns like el are preferred to emphaticreflexives like el ıˆnsus¸iwhen they achieve the same meaning.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBe Small!(16) BE SMALL!When choosing between two alternative phrases, XP and XP’,speak XP’ iff:i. XP and XP’ have indistinguishable interpretations in a contextC, andii. the set of morphosyntactic features of XP’ is a proper sub-set of the set of morphosyntactic features of XP.Prediction: regular pronouns like el are preferred to emphaticreflexives like el ıˆnsus¸iwhen they achieve the same meaning.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionExperiment 1Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionProcedure Experiment 1 & 2• The experiment took place at the University of Bucharest.• Participants were recruited through flyers, class anouncementsand via online platforms.• The experiment was coded and ran in PsychoPy.• Participants were walked through the Instructions.• Participants were instructed to choose a continuation beforeuttering the entire sentence.• Participants’ responses were recorded, transcribed andannotated.• The entire process, including debriefing, lasted between 45-60minutes per participant.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionRaw Results Experiment 1The rate of production for each on-target response type is as follows:The rate of regular pronoun production (el, ea) can be used tomeasure the effect of AMBIGUITY in both contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 1Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of coreferencewith the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z =2.88, S.E = 0.29, p < 0.01) in the Local Coreferent conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z =5.34, S.E = 0.31, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -2.68, p < 0.01)Participants used more regular pronouns in the Local Coreferentconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionExperiment 2Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDESIGN EXPERIMENT 2• picture description task• 2 x 2 design: PICTURE TYPE x AMBIGUITYLocal Bound/Local Disjoint x Character Gender Match/Mismatch• 16 items (distributed in 4 Latin Squared lists)• 20 fillers• same procedure as in Experiment 1• 4 referents in the context per itemPARTICIPANTS• 68 participants (60 female)• University of Bucharest students• The age range was between 18 and 33, with an average age of 21.3• reimbursed 30 RON (≈ 8 USD) for participationRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionData Exclusion Experiment 2• excluded non-target responses• no participants were excluded from the data analysis• in total, 1.83% of the collected data was removed• analysis ran on 1068 target responses out of a total of 1088• improved rate of target responses in comparison to Experiment 1due to having tweaked the instructions to emphasize that theexperiment was not a test of creativity.Data AnalysisFor all of the on-target responses, logistic mixed effects regressionwas used to model:• the effect of AMBIGUITY (Character Gender Mismatch/Match)• the effect of PICTURE TYPE (Local Bound/Local Disjoint)A second nested model was fitted to estimate size of AMBIGUITYwithin each picture type.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic, every girl laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic, every girl laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic, every girl laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic,every girl laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic, every girllaughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic, every girl laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionMATERIALS• Each item involved a target sentence and a target picture.• Participants continued the target sentence fragment so that itmatched the visually-provided scenario.• Each target picture and sentence set was preceded by a shortcontext introducing all 4 characters.• 3 of the characters matched in age and gender: 3 boys or 3 girls• the 4th character was always an older relative (an aunt/uncle orgrandma/grandpa)• Each target sentence consists of:→ a topic PP which names the older character→ a quantificational subject which targets the 3 other characters→ a predicate which takes a PP object (laugh at, cook for) and is equallyplausible with a reflexive and non-reflexive continuation→ No transitive verbs were used to avoid clitic doubling.At Grandpa Vlad’s picnic, every girl laughed at ...Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionCHARACTERS IN EXPERIMENT 2Given the larger number of characters in Experiment 2 and due totheir familial relationships, the entire list of characters was presentedto the participants during the instructions.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionParticipant Response Types Experiment 2On Target Responses• pronoun: el, ea ‘him, her’• emphatic reflexive: el ıˆnsus¸i,ea ıˆnsăs¸i ‘himself, herself’• reflexive: sine ‘self’• other reflexives: propriapersoana ‘own person’,persoana lui/ei ‘his/herperson’, etc.• names: Grandma Laura• demonstrative: acesta, aceasta‘this one’Non Target Responses• pronouns/names targetingwrong referent• possessive constructions: hishair, his grandpa, her success• random NPs: dissatisfactions,etc.• full sentences: what he did,how she feels, etc.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionRaw Results Experiment 2The rate of production for each on-target response type is as follows:The rate of regular pronoun production (el, ea) can be used tomeasure the effect of AMBIGUITY in both contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDiscussion Experiment 2Main Questions:1 Is the rate of pronoun production affected by context ambiguityin the case of variables bound by the local subject?2 Is the rate of pronoun production affected by context ambiguityin the case of disjoint referencewith the local subject?Answers:1 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.2,S.E = 0.3, p < 0.001) in the Local Bound conditions.2 Yes! The nested model found a clear effect of AMBIGUITY (z = 5.1,S.E = 0.32, p < 0.001) in the Local Disjoint conditions.We also found an overall effect of PICTURE TYPE (z = -3.1, p < 0.01)Participants used more regular pronouns in the Local Boundconditions in both ambiguous and unambiguous contexts.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionProduction ComparisonRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionComprehensionRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionWe also ran 2 comprehension experiments in the same vein.• REFERENTIAL SUBJECTS COMPREHENSIONthe comprehension equivalent of EXPERIMENT 1 (2 referents)• QUANTIFIED SUBJECTS COMPREHENSIONthe comprehension equivalent of EXPERIMENT 2 (4 referents)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionWe also ran 2 comprehension experiments in the same vein.• REFERENTIAL SUBJECTS COMPREHENSIONthe comprehension equivalent of EXPERIMENT 1 (2 referents)• QUANTIFIED SUBJECTS COMPREHENSIONthe comprehension equivalent of EXPERIMENT 2 (4 referents)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionWe also ran 2 comprehension experiments in the same vein.• REFERENTIAL SUBJECTS COMPREHENSIONthe comprehension equivalent of EXPERIMENT 1 (2 referents)• QUANTIFIED SUBJECTS COMPREHENSIONthe comprehension equivalent of EXPERIMENT 2 (4 referents)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionThe production experiments clearly target Ambiguity Avoidance,which is competition at the level of meaning.We also wanted to directly manipulate how competition plays out atthe level of form.In both comprehension experiments, we split participants into twogroups:• half of the participants only heard sentences with him/her• half of the participants heard sentences with regular pronouns,emphatic pronouns and demonstrativesRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionThe production experiments clearly target Ambiguity Avoidance,which is competition at the level of meaning.We also wanted to directly manipulate how competition plays out atthe level of form.In both comprehension experiments, we split participants into twogroups:• half of the participants only heard sentences with him/her• half of the participants heard sentences with regular pronouns,emphatic pronouns and demonstrativesRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionThe production experiments clearly target Ambiguity Avoidance,which is competition at the level of meaning.We also wanted to directly manipulate how competition plays out atthe level of form.In both comprehension experiments, we split participants into twogroups:• half of the participants only heard sentences with him/her• half of the participants heard sentences with regular pronouns,emphatic pronouns and demonstrativesRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDESIGN• 2 subgroups in each experiment: Gender & Form• 3 conditions: Ambiguous / Reflexive / Disjoint• 15 items• 20 fillers• 68 participants (University of Bucharest, 20ish years old)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDESIGN• 2 subgroups in each experiment: Gender & Form• 3 conditions: Ambiguous / Reflexive / Disjoint• 15 items• 20 fillers• 68 participants (University of Bucharest, 20ish years old)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDESIGN• 2 subgroups in each experiment: Gender & Form• 3 conditions: Ambiguous / Reflexive / Disjoint• 15 items• 20 fillers• 68 participants (University of Bucharest, 20ish years old)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionDESIGN• 2 subgroups in each experiment: Gender & Form• 3 conditions: Ambiguous / Reflexive / Disjoint• 15 items• 20 fillers• 68 participants (University of Bucharest, 20ish years old)Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionGENDER GROUP, REFLEXIVED KAuditorily: At Irina’s house, Andrei talked about him.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionGENDER GROUP, DISJOINTD KAuditorily: At Irina’s house, Andrei talked about her.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionAMBIGUOUSD KAuditorily: At Mihai’s house, Andrei talked about him.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionFORM GROUP, REFLEXIVED KAuditorily: At Mihai’s house, Andrei talked about him himself.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionFORM GROUP, DISJOINTD KAuditorily: At Mihai’s house, Andrei talked about this one.Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionREFERENTIAL SUBJECTSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionQUANTIFIED SUBJECTSRodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBAYES’ RULE!p(referent|pronoun) = p(referent) ∗ p(pronoun|referent)p(pronoun)We know p(pronoun|referent) and p(pronoun) by looking at theproportions in the production experiments.We know p(referent) due to the experimental design.We can calculate p(referent|pronoun) for the comprehensionexperiments!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBAYES’ RULE!p(referent|pronoun) = p(referent) ∗ p(pronoun|referent)p(pronoun)We know p(pronoun|referent) and p(pronoun) by looking at theproportions in the production experiments.We know p(referent) due to the experimental design.We can calculate p(referent|pronoun) for the comprehensionexperiments!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBAYES’ RULE!p(referent|pronoun) = p(referent) ∗ p(pronoun|referent)p(pronoun)We know p(pronoun|referent) and p(pronoun) by looking at theproportions in the production experiments.We know p(referent) due to the experimental design.We can calculate p(referent|pronoun) for the comprehensionexperiments!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBAYES’ RULE!p(referent|pronoun) = p(referent) ∗ p(pronoun|referent)p(pronoun)We know p(pronoun|referent) and p(pronoun) by looking at theproportions in the production experiments.We know p(referent) due to the experimental design.We can calculate p(referent|pronoun) for the comprehensionexperiments!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionBAYES’ RULE!p(referent|pronoun) = p(referent) ∗ p(pronoun|referent)p(pronoun)We know p(pronoun|referent) and p(pronoun) by looking at theproportions in the production experiments.We know p(referent) due to the experimental design.We can calculate p(referent|pronoun) for the comprehensionexperiments!Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:Rule I Be Small! Experiment 1 Experiment 2 Production Comparison ComprehensionPlugging in the data from Experiment 1we get the following:p(reflexive|el) = 0.56p(disjoint|el) = 0.45ACTUAL RESULTS FROM COMPREHENSION:Rate of Reflexive Interpretation of el: 57%Rate of Disjoint Interpretation of el: 43%Magic!BIG IMPORTANT QUESTIONHow do our models of production and comprehension differ?Rodica Ivan & Brian Dillon UMass Amherst(Bound) Pronouns in Competition:

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            data-media="{[{embed.selectedMedia}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
https://iiif.library.ubc.ca/presentation/dsp.73804.1-0389861/manifest

Comment

Related Items