A proposta deste trabalho é implementar um conjunto de elementos do diálogo a decorrer entre dois agentes humanos. As anáforas pronominais e certos pronomes déiticos (eu, você, sua, seu, meu, minha), que eventualmente surgirem durante o diálogo, são resolvidas. Basicamente, este trabalho está dividido em quatro partes: 1. Estudo introdutório da Discourse Representation Theory (DRT) [KAM88, KAM90]. A DRT é um formalismo para a representação do discurso que utiliza modelos na avaliação semântica das estruturas representacionais. Neste estudo são considerados somente os aspectos representacionais, dando enfoque à representação de sentenças simples. 2. Um estudo baseado em [HIR81, CAR87] sobre: tipos de ambigüidades, o que são anáforas, tipos de anáforas etc, visa fornecer ao leitor um conhecimento mínimo sobre o aspecto lingüístico do tratamento das anáforas. Dentro do estudo realizado sobre anáforas destacam-se os seguintes tipos: coespecificação pessoal e colocação em coesão léxica, que são os tipos previstos na implementação. Estes dois tipos de anáforas são enquadrados nos seguintes grupos: coespecificação pessoal no grupo das anáforas pronominais e colocação em coesão léxica no grupo das anáforas nominais. 3. Considerando que a DRT somente representa o discurso, sem contudo resolver as anafóras que surgem no discurso, incorporou-se a Teoria do Foco [SID79, COR92] como ferramenta para a resolução das anáforas pronominais. A Teoria do Foco trabalha com as informações temáticas das sentenças, de maneira a reduzir o universo dos possíveis antecedentes para uma anáfora e prover um conjunto de regras que permita um caminhamento inteligente, dentro deste universo. O algoritmo de focalização aqui utilizado e o proposto por Sophie Cormack [COR92], que foi por sua vez baseado no algoritmo original de Candace Sidner [SID79]. 4. E por último a implementação, que foi realizada em C-Prolog [PER87], onde as principais funções são: (a) Um gerador de DRSs. (b) Algoritmo de focalização. (c) Integração do algoritmo de focalização e do gerador de DRSs. Descrevendo de maneira geral o funcionamento da implementação: as falas (conjunto de sentenças) de cada interlocutor são lidas através do teclado, as sentenças de cada fala são analisadas individualmente pelo analisador sintático, que gera uma árvore de derivação sintática. A árvore gerada é então repassada ao gerador de DRSs, que irá reduzi-la a referentes e condições. Eventualmente, se surgirem anáforas pronominais, é chamado o algoritmo de focalização. Caso surjam pronomes deíticos a resolução é realizada pelo gerador de DRSs. / The proposal of this work is to implement a set of dialog elements expressed by two human agents. The pronominal anaphora and some deixis pronoums (in portuguese: I, you, your, my) that eventually appear during the dialog are resolved. Basically, this work is divided in four parts: 1. An introdutory study of the Discourse Representation Theory (DRT) [KAM88, KAM90]. The DRT is a formalism for discourse representation that uses models for semantic evaluation of the representation structures. This study considers only the representational aspects, looking for single sentences. 2. A study based on [HIR81, CAR87] about: some kinds of ambiguity, what anaphora are, kinds of anaphora etc. This study intends to give the reader the minimal knowledge about the linguistic aspects of anaphora. In this study, we point out two types of anaphora: personal coespecification and lexical placement, this two were the ones that we used in the system. Those two types are in the following groups: personal coespecification in the pronominal anaphora and lexical placement in the nominal anaphora. 3. Considering that DRT only represent the discourse without resolving the anaphora, we used the Focus Theory [SID79, COR92] as a tool for pronominal anaphora resolution. The Focus Theory works on the thematic informations of the sentences. It reduces the universe of the possible antecedents and give some rules to walk throught this universe. We used the focalization algorithm presented by Sophie Cormack [COR92] which is based on the original version of Candace Sidner [SID79]. 4. Finally, the system was implemented in C-Prolog [PER87], and its main functions are: (a) a DRS generator, (b) a focalization algorithm, (c) the integration of the focalization algorithm and the DRS generator. Basically, what the system does is: the discourse of the agent is read in the keyboard, and each sentence of the discourse is analised by the sintatic analyser, generating a parsing tree. Then the DRS generator reduces this tree into referents and DRS-conditions. Eventually, the focalization algorithm will be called when the sentence contain some pronominal anaphora. The deixis resolution is made by the DRS-generator.

This thesis investigates coreference and its resolution within the domain of health education. Coreference is the relationship between two linguistic expressions that refer to the same real-world entity, and resolution involves identifying this relationship among sets of referring expressions. The coreference resolution task is considered among the most difficult of problems in Artificial Intelligence; in some cases, resolution is impossible even for humans. For example, "she" in the sentence "Lynn called Jennifer while she was on vacation" is genuinely ambiguous: the vacationer could be either Lynn or Jennifer. <br/><br/> There are three primary motivations for this thesis. The first is that health education has never before been studied in this context. So far, the vast majority of coreference research has focused on news. Secondly, achieving domain-independent resolution is unlikely without understanding the extent to which coreference varies across different genres. Finally, coreference pervades language and is an essential part of coherent discourse. Its effective use is a key component of easy-to-understand health education materials, where readability is paramount. <br/><br/> No suitable corpus of health education materials existed, so our first step was to create one. The comprehensive analysis of this corpus, which required manual annotation of coreference, confirmed our hypothesis that the coreference used in health education differs substantially from that in previously studied domains. This analysis was then used to shape the design of a knowledge-lean algorithm for resolving coreference. This algorithm performed surprisingly well on this corpus, e.g., successfully resolving over 85% of all pronouns when evaluated on unseen data. <br/><br/> Despite the importance of coreferentially annotated corpora, only a handful are known to exist, likely because of the difficulty and cost of reliably annotating coreference. The paucity of genres represented in these existing annotated corpora creates an implicit bias in domain-independent coreference resolution. In an effort to address these issues, we plan to make our health education corpus available to the wider research community, hopefully encouraging a broader focus in the future.

This thesis investigates coreference and its resolution within the domain of health education. Coreference is the relationship between two linguistic expressions that refer to the same real-world entity, and resolution involves identifying this relationship among sets of referring expressions. The coreference resolution task is considered among the most difficult of problems in Artificial Intelligence; in some cases, resolution is impossible even for humans. For example, "she" in the sentence "Lynn called Jennifer while she was on vacation" is genuinely ambiguous: the vacationer could be either Lynn or Jennifer. <br/><br/> There are three primary motivations for this thesis. The first is that health education has never before been studied in this context. So far, the vast majority of coreference research has focused on news. Secondly, achieving domain-independent resolution is unlikely without understanding the extent to which coreference varies across different genres. Finally, coreference pervades language and is an essential part of coherent discourse. Its effective use is a key component of easy-to-understand health education materials, where readability is paramount. <br/><br/> No suitable corpus of health education materials existed, so our first step was to create one. The comprehensive analysis of this corpus, which required manual annotation of coreference, confirmed our hypothesis that the coreference used in health education differs substantially from that in previously studied domains. This analysis was then used to shape the design of a knowledge-lean algorithm for resolving coreference. This algorithm performed surprisingly well on this corpus, e.g., successfully resolving over 85% of all pronouns when evaluated on unseen data. <br/><br/> Despite the importance of coreferentially annotated corpora, only a handful are known to exist, likely because of the difficulty and cost of reliably annotating coreference. The paucity of genres represented in these existing annotated corpora creates an implicit bias in domain-independent coreference resolution. In an effort to address these issues, we plan to make our health education corpus available to the wider research community, hopefully encouraging a broader focus in the future.

This dissertation examines genericity from a textual perspective. The material consists of popular science texts about species of animals. The investigation concerns both the distribution of different forms of generic noun phrases and the boundary between generic and non-generic noun phrases. The analytical tools are taken from Accessibility Theory and Blending Theory. Two separate studies have been undertaken. The results of the first study indicate that generic reference on the whole follows the same principles of accessibility as non-generic reference, although there are some differences that can be attributed to the distinction between generic and non-generic reference. Some results suggest that our mental representations of generic referents are generally less accessible than those of non-generic referents. Factors other than accessibility influencing the choice of generic noun phrases are also identified. While genericity is generally treated as an all-or-nothing phenomenon, an important experience of this first study concerns the difficulties facing anyone who tries to distinguish between generic and non-generic noun phrases in authentic texts. These difficulties are the centre of attention in the second study, which shows that genericity is an extremely context-dependent phenomenon. The sentence context may clearly indicate a particular, non-generic reference, while the wider context of the text reveals that the noun phrase in question is in fact generic. Not infrequently, chains of reference involve a great deal of shifting and slithering between a generic and a non-generic meaning, although the references are seemingly coreferential. It is sometimes difficult to decide on the real referents intended. At times there are also clear cases where the noun phrase must be analysed as referring to both generic and non-generic entities at the same time. This implies that it is unlikely that we actually decide for every reference if it is generic or non-generic.

This dissertation examines the processing of sentences involving long-distance linguistic dependencies, or sentences containing elements that must be linked across intervening words and phrases. Specifically, both behavioral (self-paced reading and eye tracking) and neurophysiological (electroencephalography) methods were used (a) to evaluate the relative importance of backward- and forward-looking dependency satisfaction processes in the comprehension of sentences involving wh-dependencies and (b) to determine the extent to which common neurocognitive mechanisms are involved the processing of wh- and anaphoric dependencies. With respect to the first issue, both behavioral andneurophysiological results indicated a core role for forward-looking, expectancy-based processes in the comprehension of wh-dependency sentences. Regarding the latter issue, despite considerable overlap in the reading patterns associated with wh-dependencies and (at least some types of) anaphora, the neurophysiological responses related to these dependency types indicated that their processing draws on distinct neurocognitive mechanisms.

The thesis deals with the relationship between anaforization of name syntagma and determination , which can be seen as an exponent of the anaforization process. The purpose of this work is to evaluate what role had the individual determinants in old French in comparison to their role in contemporary French. The thesis is based on the specialized literature and analysis of anaforical name syntagmas in the text Aucassin et Nicolette. The work is divided into two main parts. The first theoretical part is focused on defining the basic concepts using specialized literature on which the analysis is based and the second part is mainly the analysis itself of nominal syntagmas in the text Aucassin et Nicolette.

A proposta deste trabalho é implementar um conjunto de elementos do diálogo a decorrer entre dois agentes humanos. As anáforas pronominais e certos pronomes déiticos (eu, você, sua, seu, meu, minha), que eventualmente surgirem durante o diálogo, são resolvidas. Basicamente, este trabalho está dividido em quatro partes: 1. Estudo introdutório da Discourse Representation Theory (DRT) [KAM88, KAM90]. A DRT é um formalismo para a representação do discurso que utiliza modelos na avaliação semântica das estruturas representacionais. Neste estudo são considerados somente os aspectos representacionais, dando enfoque à representação de sentenças simples. 2. Um estudo baseado em [HIR81, CAR87] sobre: tipos de ambigüidades, o que são anáforas, tipos de anáforas etc, visa fornecer ao leitor um conhecimento mínimo sobre o aspecto lingüístico do tratamento das anáforas. Dentro do estudo realizado sobre anáforas destacam-se os seguintes tipos: coespecificação pessoal e colocação em coesão léxica, que são os tipos previstos na implementação. Estes dois tipos de anáforas são enquadrados nos seguintes grupos: coespecificação pessoal no grupo das anáforas pronominais e colocação em coesão léxica no grupo das anáforas nominais. 3. Considerando que a DRT somente representa o discurso, sem contudo resolver as anafóras que surgem no discurso, incorporou-se a Teoria do Foco [SID79, COR92] como ferramenta para a resolução das anáforas pronominais. A Teoria do Foco trabalha com as informações temáticas das sentenças, de maneira a reduzir o universo dos possíveis antecedentes para uma anáfora e prover um conjunto de regras que permita um caminhamento inteligente, dentro deste universo. O algoritmo de focalização aqui utilizado e o proposto por Sophie Cormack [COR92], que foi por sua vez baseado no algoritmo original de Candace Sidner [SID79]. 4. E por último a implementação, que foi realizada em C-Prolog [PER87], onde as principais funções são: (a) Um gerador de DRSs. (b) Algoritmo de focalização. (c) Integração do algoritmo de focalização e do gerador de DRSs. Descrevendo de maneira geral o funcionamento da implementação: as falas (conjunto de sentenças) de cada interlocutor são lidas através do teclado, as sentenças de cada fala são analisadas individualmente pelo analisador sintático, que gera uma árvore de derivação sintática. A árvore gerada é então repassada ao gerador de DRSs, que irá reduzi-la a referentes e condições. Eventualmente, se surgirem anáforas pronominais, é chamado o algoritmo de focalização. Caso surjam pronomes deíticos a resolução é realizada pelo gerador de DRSs. / The proposal of this work is to implement a set of dialog elements expressed by two human agents. The pronominal anaphora and some deixis pronoums (in portuguese: I, you, your, my) that eventually appear during the dialog are resolved. Basically, this work is divided in four parts: 1. An introdutory study of the Discourse Representation Theory (DRT) [KAM88, KAM90]. The DRT is a formalism for discourse representation that uses models for semantic evaluation of the representation structures. This study considers only the representational aspects, looking for single sentences. 2. A study based on [HIR81, CAR87] about: some kinds of ambiguity, what anaphora are, kinds of anaphora etc. This study intends to give the reader the minimal knowledge about the linguistic aspects of anaphora. In this study, we point out two types of anaphora: personal coespecification and lexical placement, this two were the ones that we used in the system. Those two types are in the following groups: personal coespecification in the pronominal anaphora and lexical placement in the nominal anaphora. 3. Considering that DRT only represent the discourse without resolving the anaphora, we used the Focus Theory [SID79, COR92] as a tool for pronominal anaphora resolution. The Focus Theory works on the thematic informations of the sentences. It reduces the universe of the possible antecedents and give some rules to walk throught this universe. We used the focalization algorithm presented by Sophie Cormack [COR92] which is based on the original version of Candace Sidner [SID79]. 4. Finally, the system was implemented in C-Prolog [PER87], and its main functions are: (a) a DRS generator, (b) a focalization algorithm, (c) the integration of the focalization algorithm and the DRS generator. Basically, what the system does is: the discourse of the agent is read in the keyboard, and each sentence of the discourse is analised by the sintatic analyser, generating a parsing tree. Then the DRS generator reduces this tree into referents and DRS-conditions. Eventually, the focalization algorithm will be called when the sentence contain some pronominal anaphora. The deixis resolution is made by the DRS-generator.

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A presente dissertação intitulada Anáforas encapsuladoras na voz do narrador de Menino de engenho tem como objetivo primeiramente identificar os encapsulamentos anafóricos que ocorrem em alguns fragmentos da voz do narrador da obra literária de José Lins do Rego. Sobre as anáforas, observamos a sua importância no contexto em que estão sendo utilizadas, facilitando a compreensão do sentido dos objetos-de-discurso previamente colocados no texto. Levando-se em consideração os estudos sobre a Linguística textual, especificamente a teoria da Referenciação, baseamo-nos no referencial teórico dos trabalhos de Koch (2011b, 2014); Koch & Elias (2012); Mondada & Dubois (2003); Cavalcante (2013); Cortez (2012) para respaldar a investigação sobre as anáforas. Além disso, temos como objetivo discutir as anáforas encapsuladoras na voz do narrador, que, concedeu-nos, à luz da polifonia de Oswald Ducrot (1987), distinguir o narrador/locutor do narrador/enunciador. Ademais, há um jogo de vozes, haja vista que surge a indagação: ora é o narrador/locutor que surge ou o narrador/enunciador? A partir disso, evidenciamos a relevância do contexto para identificação da voz que se enuncia. Analisamos, a partir dos estudos realizados sobre as anáforas encapsuladoras e a teoria polifônica de Ducrot, alguns fragmentos da obra que comprovam os encapsulamentos e sua relação com a voz do narrador. As expressões anafóricas fortalecem o dito pelo narrador e revelam os enunciados pelas vozes que surgem no contexto. A escolha do referido cânone literário deve-se à analogia entre a narrativa memorialista e a própria anáfora pelo fato de que em ambas há um movimento de retorno, que faz retomar algo que foi vivido (no caso da narrativa) ou que foi dito (no caso da anáfora). A narrativa memorialista trabalhada não no viés da literariedade em si, mas na análise da profusão de vozes que surgem nas expressões anafóricas evidencia a relevância dos estudos linguísticos bem como seus usos na obra literária em pauta que, por sua vez, reflete a pluralidade da ação e diz algo sobre o mundo real. / This Anaphoras dissertation entitled encapsulating the voice of the Menino de engenho narrator aims to first identify the anaphoric tunnels that occur in some fragments of the voice of the narrator of the novel by José Lins do Rego. About the anaphora, observe its importance in the context in which they are being used, facilitating the understanding of the meaning of the objects of-speech previously placed in the text. Taking into account the studies on the textual linguistics, specifically the theory of referencing, we rely on the theoretical framework of Koch's work (2011b, 2014); Koch & Elias (2012); Mondada & Dubois (2003); Cavalcante (2013); Cortez (2012) to support research on anaphora. In addition, we aim to discuss the encapsulation anaphora in the narrator's voice, which gave us the light of polyphony Oswald Ducrot (1987), distinguish the narrator / announcer of the narrator / enunciator. In addition, there is a "mess" of voices, given that arises the question: now is o narrador / announcer who comes or the narrator / enunciator? From this, we saw the relevance of voice identification context that sets out. Analyzed, from studies on the encapsulation anaphora and polyphonic theory Ducrot, some fragments of the work that prove the tunnels and their relationship to the voice of the narrator. Anaphoric expressions strengthen said by the narrator and reveal the statements by the voices that arise in the context. The choice of such literary canon is due to the analogy between narrative and memoir itself anaphora by the fact that in both there is a return movement, which is getting back something that was lived (in the case of the narrative) or what was said (in case of anaphora). The narrative memoir worked not in the bias of literariness itself, but the analysis of the multitude of voices that arise in anaphoric expressions highlights the relevance of linguistic studies and their uses in the literary work in question which, in turn, reflects the plurality of action and says something about the real world.

A proposta deste trabalho é implementar um conjunto de elementos do diálogo a decorrer entre dois agentes humanos. As anáforas pronominais e certos pronomes déiticos (eu, você, sua, seu, meu, minha), que eventualmente surgirem durante o diálogo, são resolvidas. Basicamente, este trabalho está dividido em quatro partes: 1. Estudo introdutório da Discourse Representation Theory (DRT) [KAM88, KAM90]. A DRT é um formalismo para a representação do discurso que utiliza modelos na avaliação semântica das estruturas representacionais. Neste estudo são considerados somente os aspectos representacionais, dando enfoque à representação de sentenças simples. 2. Um estudo baseado em [HIR81, CAR87] sobre: tipos de ambigüidades, o que são anáforas, tipos de anáforas etc, visa fornecer ao leitor um conhecimento mínimo sobre o aspecto lingüístico do tratamento das anáforas. Dentro do estudo realizado sobre anáforas destacam-se os seguintes tipos: coespecificação pessoal e colocação em coesão léxica, que são os tipos previstos na implementação. Estes dois tipos de anáforas são enquadrados nos seguintes grupos: coespecificação pessoal no grupo das anáforas pronominais e colocação em coesão léxica no grupo das anáforas nominais. 3. Considerando que a DRT somente representa o discurso, sem contudo resolver as anafóras que surgem no discurso, incorporou-se a Teoria do Foco [SID79, COR92] como ferramenta para a resolução das anáforas pronominais. A Teoria do Foco trabalha com as informações temáticas das sentenças, de maneira a reduzir o universo dos possíveis antecedentes para uma anáfora e prover um conjunto de regras que permita um caminhamento inteligente, dentro deste universo. O algoritmo de focalização aqui utilizado e o proposto por Sophie Cormack [COR92], que foi por sua vez baseado no algoritmo original de Candace Sidner [SID79]. 4. E por último a implementação, que foi realizada em C-Prolog [PER87], onde as principais funções são: (a) Um gerador de DRSs. (b) Algoritmo de focalização. (c) Integração do algoritmo de focalização e do gerador de DRSs. Descrevendo de maneira geral o funcionamento da implementação: as falas (conjunto de sentenças) de cada interlocutor são lidas através do teclado, as sentenças de cada fala são analisadas individualmente pelo analisador sintático, que gera uma árvore de derivação sintática. A árvore gerada é então repassada ao gerador de DRSs, que irá reduzi-la a referentes e condições. Eventualmente, se surgirem anáforas pronominais, é chamado o algoritmo de focalização. Caso surjam pronomes deíticos a resolução é realizada pelo gerador de DRSs. / The proposal of this work is to implement a set of dialog elements expressed by two human agents. The pronominal anaphora and some deixis pronoums (in portuguese: I, you, your, my) that eventually appear during the dialog are resolved. Basically, this work is divided in four parts: 1. An introdutory study of the Discourse Representation Theory (DRT) [KAM88, KAM90]. The DRT is a formalism for discourse representation that uses models for semantic evaluation of the representation structures. This study considers only the representational aspects, looking for single sentences. 2. A study based on [HIR81, CAR87] about: some kinds of ambiguity, what anaphora are, kinds of anaphora etc. This study intends to give the reader the minimal knowledge about the linguistic aspects of anaphora. In this study, we point out two types of anaphora: personal coespecification and lexical placement, this two were the ones that we used in the system. Those two types are in the following groups: personal coespecification in the pronominal anaphora and lexical placement in the nominal anaphora. 3. Considering that DRT only represent the discourse without resolving the anaphora, we used the Focus Theory [SID79, COR92] as a tool for pronominal anaphora resolution. The Focus Theory works on the thematic informations of the sentences. It reduces the universe of the possible antecedents and give some rules to walk throught this universe. We used the focalization algorithm presented by Sophie Cormack [COR92] which is based on the original version of Candace Sidner [SID79]. 4. Finally, the system was implemented in C-Prolog [PER87], and its main functions are: (a) a DRS generator, (b) a focalization algorithm, (c) the integration of the focalization algorithm and the DRS generator. Basically, what the system does is: the discourse of the agent is read in the keyboard, and each sentence of the discourse is analised by the sintatic analyser, generating a parsing tree. Then the DRS generator reduces this tree into referents and DRS-conditions. Eventually, the focalization algorithm will be called when the sentence contain some pronominal anaphora. The deixis resolution is made by the DRS-generator.

Submitted by Morgana Silva (morgana_linhares@yahoo.com.br) on 2016-10-11T17:46:11Z No. of bitstreams: 1 arquivototal.pdf: 1246064 bytes, checksum: 48d673716d069f78f8353ea6c93230d5 (MD5) / Made available in DSpace on 2016-10-11T17:46:11Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 1246064 bytes, checksum: 48d673716d069f78f8353ea6c93230d5 (MD5) Previous issue date: 2016-02-19 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This study aims to investigate whether logophoric structures are acceptable or not for the speakers of Brazilian Portuguese. Much of the issue literature on the subject (Reinhart and Reuland (1993), Menuzzi (1997)) points out that logophoric anaphora exists separately from syntactic anaphora in Universal Grammar. The syntactic anaphoras are guided by syntactic factors such as location and c-command. But logophor may or may not notice these syntactic conditions. These studies also demonstrate that where there is logophoricity, the pronoun and anaphora are interchangeable, so, either one or the other can be used in the same position in the sentence. The conclusion we reached after the literature review is that anaphora resolution in logophorics contexts seems to involve not only the influence of structural information but also and above all, the influence of pragmatic and discursive information, as seen in pronouns. Based on these theoretical assumptions, we examine the acceptability of logophorics sentences in PB in order to know if they are present in the grammar of the speakers of that language. Two acceptability judgment experiments were conducted in order to know if the structures with the logophoric anaphor “ele mesmo” and the logophoric pronoun “ele” are acceptable in BP, based on the English structures presented by Reinhart e Reulnad (1993). Our hypothesis was that both structures are acceptable to speakers of BP and that in contexts where there is a logophoric anaphor, there can also be a logophoric pronoun, as shown by Reinhart e Reuland (1993). The results showed an acceptance of both structures in BP, supporting the formulated hypothesis. With these results, we obtained evidence that the structures with the logophoric anaphora “ele mesmo” and the logophoric pronoun “ele” observed in this study are acceptable in Brazilian Portuguese, as in English. / Este trabalho busca investigar se as estruturas logofóricas são aceitáveis ou não para os falantes do português brasileiro (PB). Grande parte da literatura sobre o assunto (Reinhart e Reuland (1993), Menuzzi (1997), dentre outros) aponta que a anáfora logofórica existe separadamente da anáfora sintática na Gramática Universal. As anáforas sintáticas são guiadas por fatores sintáticos como localidade e c-comando. Já a logofórica pode ou não observar essas condições sintáticas. Esses estudos também demonstram que nos casos em que há logoforicidade, a anáfora e o pronome são intercambiáveis, ou seja, tanto um quanto o outro podem ser usados na mesma posição na frase. A conclusão a que chegamos após a revisão de literatura é que a resolução de anáforas em contextos logofóricos parece envolver não apenas a influência de informações estruturais, mas também, e sobretudo, a influência de informações pragmáticas e discursivas, tal como se observa nos pronomes. Apontamos que a Teoria da Reflexividade parece ser mais coerente para a explicação desse fenômeno. Com base nesses pressupostos teóricos, examinamos a aceitabilidade de construções logofóricas em PB com a finalidade de saber se elas estão presentes na gramática dos falantes dessa língua. Foram realizados dois experimentos de julgamento de aceitabilidade no intuito de saber se as estruturas com a anáfora logofórica ‘ele mesmo’ e com o pronome logofórico ‘ele’ são aceitáveis em PB, tomando como base as estruturas em inglês apresentadas por Reinhart e Reuland (1993). A nossa hipótese foi de que ambas as estruturas são aceitáveis para os falantes do PB e que nos contextos em que há uma anáfora logofórica também pode haver um pronome logofórico, conforme aponta Reinhart e Reuland (1993). Os resultados encontrados evidenciaram uma aceitação de ambas as estruturas em PB, corroborando a hipótese formulada. Com estes resultados, obtivemos evidências de que as estruturas com a anáfora logofórica ‘ele mesmo’ e com o pronome logofórico ‘ele’ observadas nesse estudo são aceitas em PB, como em inglês.