This dissertation is an extended argument for the syntactic structure proposed in (1), referred to as the 'VICTR Hierarchy' after the initials of the five functional heads it comprises: (1) [VolitionP Volition [InitiationP Initiation [ConsecutionP Consecution [TransitionP Transition [ResultP Result [VP ] ] ] ] ] ] The VICTR Hierarchy is a hierarchy of functional heads corresponding to the part of the clause generally known in the minimalist literature as `vP' or the `thematic domain'. Nominal arguments are merged in the specifiers of one or more of these heads and receive their thematic interpretations on the basis of their merged positions. Evidence for a model of thematic roles and syntactic argument structure based in the VICTR Hierarchy is presented for a range of domains, with a focus on 'split intransitivity'. Split intransitivity is explored initially in regard to English, with close consideration of a range of split intransitive diagnostics (e.g. 'out'-prefixation, the resultative construction); a VICTR account of these patterns is presented. A VICTR account of auxiliary selection patterns in Western European languages is also given. This is followed by analysis of split intransitive case and agreement systems. A formal account of the case and agreement patterns in these languages based in the VICTR hierarchy is presented, derived in part from the inherent case theory of ergativity (Legate 2002, Aldridge 2004 and others) and drawing on a detailed typology. The dissertation then proceeds to detailed analysis of the semantic basis of split intransitive alignment in two languages, Basque and Georgian. Other split intransitive behaviours in these languages are also considered in VICTR terms. Throughout, the VICTR approach is compared to other approaches to split intransitivity following Perlmutter's (1978) Unaccusative Hypothesis. The VICTR Hierarchy is also compared to the similar proposal of Ramchand (2008). It is argued that the VICTR Hierarchy accounts more readily than these other approaches for the particular classes of verbs identified by split intransitivity diagnostics in the languages considered, and also for cross-linguistic variation in split intransitive behaviours. Much support, with some caveats, is also found in the data considered for the applicability of Sorace's (2000) Auxiliary Selection Hierarchy (ASH) to a range of split intransitive phenomena cross-linguistically. Together with acquisitional considerations, the VICTR features are argued to allow for a formalisation of the patterns described by the ASH.

As construções passivas no Português Brasileiro e no Espanhol apresentam tanto tendências comuns, quanto assimetrias. Este trabalho tem o propósito de investigar os efeitos desse fenômeno na aquisição/aprendizagem de Espanhol por estudantes brasileiros. Inicialmente, propõe-se uma descrição das construções passivas nas duas línguas. Em seguida, faz-se uma análise contrastiva dessas construções no Português Brasileiro e no Espanhol. Finalmente, investiga-se o comportamento das construções passivas (sobretudo as perifrásticas e as lexicais) na produção escrita de brasileiros aprendizes de Espanhol língua estrangeira. Os resultados obtidos na análise dos corpora sugerem algumas hipóteses sobre os fatores que possivelmente influenciam a preferência dos aprendizes por determinadas construções em detrimento de outras / Passive constructions in Brazilian Portuguese and Spanish present both common tendencies and assimetries. The aim of this project is to investigate the effects of such phenomenon on Spanish acquisition/learning process by brazilian students. Initially, a description of passive constructions is proposed for both languages. After that, a contrastive analysis is made of those constructions in Brazilian Portuguese and Spanish. Finally, the behavior of passive constructions is analyzed in the written production of brazilian students who learn Spanish as a foreign language. Results from corpora analysis suggest some hypotheses on the factors which possibly influence learners\' choices for certain constructions to the detriment of others

As construções passivas no Português Brasileiro e no Espanhol apresentam tanto tendências comuns, quanto assimetrias. Este trabalho tem o propósito de investigar os efeitos desse fenômeno na aquisição/aprendizagem de Espanhol por estudantes brasileiros. Inicialmente, propõe-se uma descrição das construções passivas nas duas línguas. Em seguida, faz-se uma análise contrastiva dessas construções no Português Brasileiro e no Espanhol. Finalmente, investiga-se o comportamento das construções passivas (sobretudo as perifrásticas e as lexicais) na produção escrita de brasileiros aprendizes de Espanhol língua estrangeira. Os resultados obtidos na análise dos corpora sugerem algumas hipóteses sobre os fatores que possivelmente influenciam a preferência dos aprendizes por determinadas construções em detrimento de outras / Passive constructions in Brazilian Portuguese and Spanish present both common tendencies and assimetries. The aim of this project is to investigate the effects of such phenomenon on Spanish acquisition/learning process by brazilian students. Initially, a description of passive constructions is proposed for both languages. After that, a contrastive analysis is made of those constructions in Brazilian Portuguese and Spanish. Finally, the behavior of passive constructions is analyzed in the written production of brazilian students who learn Spanish as a foreign language. Results from corpora analysis suggest some hypotheses on the factors which possibly influence learners\' choices for certain constructions to the detriment of others

本論文藉由動補語後方名詞，依理論架構重新分類趨向動補語Ｖ上，並探討其前後的搭配詞在分類底下與「上」之間的互動情形。首先，根據Talmy (2000) 所提出的動態事件理論區辨後方名詞，其乃是具有傳達出動態語意的名詞特徵。其次，並以V上的論旨角色作為分類，因而總共得出三種型態的V上：[ V上Ground]，[ V上Patient]，[ V上Extent]，藉此看到前方動詞和後方名詞與V上之間的緊密關係。除外並以認知語意學角度呈現其三種類型之間具有語意關連，表現出多義詞的演變型態，且符合語法化的過程。 / When the morpheme shang4 ‘up’ 上 is attached to a verb, it composes a V-shang constituent, which is a member of the group of ‘verb-complement’ marking directional motion of the action (Liu, 1983/1998). V-shang signifies the movement of an object through an action from a low position to a higher position within an upward orientation; for instance, a person moves from the ground to the top of a mountain. The previous studies have listed several extended meanings, besides the basic meaning of motion (Liu, 1998, Jiang, 2003; Pan, 2005; Yu, 2010). Most studies categorize V-shang by mixed criteria of semantic meanings and syntactic functions. Noting the lack of a theoretical basis in the classification of V-shang, this thesis depends on motion-event theory, in which an event is composed of Figure, Motion and Ground, to examine the motion event conducted by V-shang (Talmy, 2000) and Lexical Functional Grammar (Kaplan and Bresnan, 1982; Dowty, 1988; Bresnan and Zaenen, 1990; Bresnan, 2001; Her, 2009) to analyze the syntactic structures of the first verbs and shang4 in V-shang. By analyzing the noun phrases following V-shang, the study divides [V-shang NP] into three categories: [V-shang Ground], [V-shang Patient], [V-shang Extent] and observes the verb types in each pattern. There are the changes of semantic meanings from directional motion into resultative state and of orientation from vertical path into horizontal path. Within Lexical Functional Grammar, the verb types of the first verbs in V-shang change from the physical motion verbs which require thematic roles of Theme and Location to the transitive verbs which require thematic roles of Agent and Patient. [V-shang Extent] is the most abstract category of its semantic meaning, and the verb types of the first verbs are less restricted than the verb types in other two categories. The cognitive approach also explains the relationship of the extended meanings in V-shang and shang4 (Lu, 2011). Shang4 presents the features of polysemy.

Inspired by an ongoing debate in the clinical sciences concerning the value of evil as a label for human behavior (Mowrer 1960, Staub 1999, Wellman 2000, Williams 2004 etc.), this thesis examines the semantic role of AGENT in the lexical items sick and evil. Williams makes the argument that the label evil removes responsibility from the doctor, whereas, the label sick empowers the doctor in bringing about a cure. While this view is not universally accepted in the field, it does bring to light an interesting question in applied linguistic semantics as to the assignment of agency with respect to sick and evil. Based on the close association of the meanings of sick and evil that stems from historical, psychological, and legal perspectives, this thesis assumes that the semantic feature [+/-RESPONSIBILITY] is assigned to either sick or evil at some point along a continuum. This continuum establishes EVIL at one pole and receives [+RESPONSIBILITY] while SICK is at the opposite pole and receives [-RESPONSIBILITY]. Using a variety of prompts to survey 106 respondents, the continuum model is shown to be only partially true. There is a correlation between NON-RESPONSIBILITY and SICK. Also, a continuum exists that allows the assignment of PARTIAL RESPONSIBILITY to both terms. However, there is no definitive significant correlation between RESPONSIBILITY and EVIL. Further conclusions include the indication of adherence to a legal model of guilt, innocence, and insanity in the general conceptions of SICK and EVIL. Also, demographic variation shows little predictive potential in how people perceive SICK and EVIL. This thesis concludes with a proposal for an alternative model using a Greimas Square to represent the conceptions of SICK and EVIL that more appropriately fits the trends found in the survey data.

A criação de recursos linguístico-computacionais de base, como é o caso dos léxicos computacionais, é um dos focos da área de Processamento de Línguas Naturais (PLN). Porém, a maioria dos recursos léxicos computacionais existentes é específica da língua inglesa. Dentre os recursos já desenvolvidos para a língua inglesa, tem-se a VerbNet, que é um léxico com informações semânticas e sintáticas dos verbos do inglês, independente de domínio, construído com base nas classes verbais de Levin, além de possuir mapeamentos para a WordNet de Princeton (WordNet). Considerando que há poucos estudos computacionais sobre as classes de Levin, que é a base da VerbNet, para línguas diferentes do inglês, e dada a carência de um léxico para o português nos moldes da VerbNet do inglês, este trabalho teve como objetivo a criação de um recurso léxico para o português do Brasil (chamado VerbNet.Br), semelhante à VerbNet. A construção manual destes recursos geralmente é inviável devido ao tempo gasto e aos erros inseridos pelo autor humano. Portanto, há um grande esforço na área para a criação destes recursos apoiada por técnicas computacionais. Uma técnica reconhecida e bastante usada é o uso de aprendizado de máquina em córpus para extrair informação linguística. A outra é o uso de recursos já existentes para outras línguas, em geral o inglês, visando à construção de um novo recurso alinhado, aproveitando-se de atributos multilíngues/cross-linguísticos (cross-linguistic) (como é o caso da classificação verbal de Levin). O método proposto neste mestrado para a construção da VerbNet.Br é genérico, porque pode ser utilizado para a construção de recursos semelhantes para outras línguas, além do português do Brasil. Além disso, futuramente, será possível estender este recurso via criação de subclasses de conceitos. O método para criação da VerbNet.Br é fundamentado em quatro etapas: três automáticas e uma manual. Porém, também foram realizados experimentos sem o uso da etapa manual, constatando-se, com isso, que ela pode ser descartada sem afetar a precisão e abrangência dos resultados. A avaliação do recurso criado foi realizada de forma intrínseca qualitativa e quantitativa. A avaliação qualitativa consistiu: (a) da análise manual de algumas classes da VerbNet, criando um gold standard para o português do Brasil; (b) da comparação do gold standard criado com os resultados da VerbNet.Br, obtendo resultados promissores, por volta de 60% de f-measure; e (c) da comparação dos resultados da VerbNet.Br com resultados de agrupamento de verbos, concluindo que ambos os métodos apresentam resultados similares. A avaliação quantitativa considerou a taxa de aceitação dos membros das classes da VerbNet.Br, apresentando resultados na faixa de 90% de aceitação dos membros em cada classe. Uma das contribuições deste mestrado é a primeira versão da VerbNet.Br, que precisa de validação linguística, mas que já contém informação para ser utilizada em tarefas de PLN, com precisão e abrangência de 44% e 92,89%, respectivamente / Building computational-linguistic base resources, like computational lexical resources (CLR), is one of the goals of Natural Language Processing (NLP). However, most computational lexicons are specific to English. One of the resources already developed for English is the VerbNet, a lexicon with domain-independent semantic and syntactic information of English verbs. It is based on Levin\'s verb classification, with mappings to Princeton\'s WordNet (WordNet). Since only a few computational studies for languages other than English have been made about Levin\'s classification, and given the lack of a Portuguese CLR similar to VerbNet, the goal of this research was to create a CLR for Brazilian Portuguese (called VerbNet.Br). The manual building of these resources is usually unfeasible because it is time consuming and it can include many human-made errors. Therefore, great efforts have been made to build such resources with the aid of computational techniques. One of these techniques is machine learning, a widely known and used method for extracting linguistic information from corpora. Another one is the use of pre-existing resources for other languages, most commonly English, to support the building of new aligned resources, taking advantage of some multilingual/cross-linguistic features (like the ones in Levin\'s verb classification). The method proposed here for the creation of VerbNet.Br is generic, therefore it may be used to build similar resources for languages other than Brazilian Portuguese. Moreover, the proposed method also allows for a future extension of the resource via subclasses of concepts. The VerbNet.Br has a four-step method: three automatic and one manual. However, experiments were also carried out without the manual step, which can be discarded without affecting precision and recall. The evaluation of the resource was intrinsic, both qualitative and quantitative. The qualitative evaluation consisted in: (a) manual analysis of some VerbNet classes, resulting in a Brazilian Portuguese gold standard; (b) comparison of this gold standard with the VerbNet.Br results, presenting promising results (almost 60% of f-measure); and (c), comparison of the VerbNet.Br results to verb clustering results, showing that both methods achieved similar results. The quantitative evaluation considered the acceptance rate of candidate members of VerbNet.Br, showing results around 90% of acceptance. One of the contributions of this research is to present the first version of VerbNet.Br. Although it still requires linguistic validation, it already provides information to be used in NLP tasks, with precision and recall of 44% and 92.89%, respectively

O presente trabalho possui como foco de estudo a análise da representação de papéis temáticos desempenhados por atores sociais, em redações escritas nos moldes da produção textual exigida pelo Exame Nacional do Ensino Médio (Enem). Nesse sentido, o participante deve produzir um texto em prosa, do tipo dissertativo-argumentativo, na modalidade escrita formal da Língua Portuguesa, sobre o tema indicado de modo que apresente uma ou mais propostas de intervenção social para o(s) problema(s) discutido(s) no texto elaborado e que respeite os direitos humanos. Nosso corpus foi composto por 1.405 redações com conteúdo equivalente a 367.565 tokens (total de palavras) e 22.707 types (total de palavras distintas), produzidas entre 2009 e 2014, com as características exigidas pelo Enem e que estão disponíveis no Banco de redações (http://educacao.uol.com.br/bancoderedacoes/) do site UOL Educação (http://educacao.uol.com.br/), o qual possui uma seção especializada em produção textual que disponibiliza propostas de redação para os internautas assim como realiza a correção de vários textos enviados pelos interessados. Diante disso, nosso principal objetivo foi compreender os usos das representações para atender aos requisitos da prova pelo viés de teorias, metodologias e trabalhos da Linguística Cognitiva, especificamente quanto à Gramática de Papéis (CANÇADO, 2000; 2008), da Análise Crítica do Discurso, especialmente sobre a representação de atores sociais (van LEEUWEN, 1996; 2008), da Linguística Sistêmico-Funcional (HALLIDAY e MATTHIESSEN, 2004) e da Linguística de Corpus (NOVODVORSKI, 2008; 2013 e BERBER SARDINHA, 2004; 2009), tanto em pesquisa baseada em corpus quanto guiada pelo corpus com o auxílio de ferramentas do programa WordSmith Tools®, versão 6.0 (SCOTT, 2012). Assim, conheceu-se as características da proposição para a produção textual do Enem, a principal prova do Brasil que possibilita o ingresso no ensino superior, e os aspectos relevantes de textos elaborados nesse contexto. Assim, realizamos uma análise que nos permitiu identificar e descrever quais os principais papéis temáticos desempenhados pelos atores sociais por nós selecionados de nosso corpus de estudo, essa análise perpassou pela identificação de campos semânticos e investigação das relações semânticas constituídas entre os verbos (processos) e seus sujeitos e complementos (argumentos, atores sociais ou participantes). / This work has as focus of study the analysis of thematic roles representation performed by social actors in texts written according to the patterns demanded by the Exame Nacional do Ensino Médio (Enem). Such examination demands the participant to produce an argumentative dissertation, in prose, using the formal register of Portuguese language, and considering a suggested theme. Also, it is necessary for the participant to present one or more intervention proposals to the problem(s) discussed in the text, having in mind the respect for human rights. Thus, in order to develop the present work, it was used a corpus of study composed by 1.405 texts. Their content was equivalent to 367.565 tokens (running words) and 22.707 types (distinct words), produced between 2009 and 2014, and followed the characteristics demanded by Enem, which are available in Banco de redações (http://educacao.uol.com.br/bancoderedacoes/) from UOL Educação website (http://educacao.uol.com.br/). This website has a section specialized in text production where thematic proposals are provided, and which correct many texts sent by web surfers. Accordingly, the goal of this work was to comprehend the use of representations in order to meet the Enem examination requirements. For this purpose, this research was guided by the bias of theory, methodology and works from Cognitive Linguistics, specifically in which concerns to Grammar of roles (CANÇADO, 2000; 2008). Also, it was based on works from Critical Discourse Analyses, specially about the representation of social agents (van LEEUWEN, 1996; 2008), from Functional-systemic Linguistics (HALLIDAY e MATTHIESSEN, 2004) and from the Corpus Linguistics (NOVODVORSKI, 2008; 2013 e BERBER SARDINHA, 2004; 2009). It was considered both researches based on and guided by corpus, with the aid of the tools provided by the software WordSmith Tools® 6,0 version (SCOTT, 2012). In light of this, it was possible to recognize the characteristics of the proposition to the text production of Enem – considered the most important examination in Brazil that enable the entry into higher education – as well as relevant aspects of texts approached in this context. Therefore, it was accomplished an analysis that allowed to identify and to describe the main thematic roles performed by the social actors selected from this work's corpus of study. This analysis went through the identification of semantic fields and the investigation of the semantic relation constituted between verbs (processes) and its subjects and complements (arguments, social actors, and participants). / Dissertação (Mestrado)

A criação de recursos linguístico-computacionais de base, como é o caso dos léxicos computacionais, é um dos focos da área de Processamento de Línguas Naturais (PLN). Porém, a maioria dos recursos léxicos computacionais existentes é específica da língua inglesa. Dentre os recursos já desenvolvidos para a língua inglesa, tem-se a VerbNet, que é um léxico com informações semânticas e sintáticas dos verbos do inglês, independente de domínio, construído com base nas classes verbais de Levin, além de possuir mapeamentos para a WordNet de Princeton (WordNet). Considerando que há poucos estudos computacionais sobre as classes de Levin, que é a base da VerbNet, para línguas diferentes do inglês, e dada a carência de um léxico para o português nos moldes da VerbNet do inglês, este trabalho teve como objetivo a criação de um recurso léxico para o português do Brasil (chamado VerbNet.Br), semelhante à VerbNet. A construção manual destes recursos geralmente é inviável devido ao tempo gasto e aos erros inseridos pelo autor humano. Portanto, há um grande esforço na área para a criação destes recursos apoiada por técnicas computacionais. Uma técnica reconhecida e bastante usada é o uso de aprendizado de máquina em córpus para extrair informação linguística. A outra é o uso de recursos já existentes para outras línguas, em geral o inglês, visando à construção de um novo recurso alinhado, aproveitando-se de atributos multilíngues/cross-linguísticos (cross-linguistic) (como é o caso da classificação verbal de Levin). O método proposto neste mestrado para a construção da VerbNet.Br é genérico, porque pode ser utilizado para a construção de recursos semelhantes para outras línguas, além do português do Brasil. Além disso, futuramente, será possível estender este recurso via criação de subclasses de conceitos. O método para criação da VerbNet.Br é fundamentado em quatro etapas: três automáticas e uma manual. Porém, também foram realizados experimentos sem o uso da etapa manual, constatando-se, com isso, que ela pode ser descartada sem afetar a precisão e abrangência dos resultados. A avaliação do recurso criado foi realizada de forma intrínseca qualitativa e quantitativa. A avaliação qualitativa consistiu: (a) da análise manual de algumas classes da VerbNet, criando um gold standard para o português do Brasil; (b) da comparação do gold standard criado com os resultados da VerbNet.Br, obtendo resultados promissores, por volta de 60% de f-measure; e (c) da comparação dos resultados da VerbNet.Br com resultados de agrupamento de verbos, concluindo que ambos os métodos apresentam resultados similares. A avaliação quantitativa considerou a taxa de aceitação dos membros das classes da VerbNet.Br, apresentando resultados na faixa de 90% de aceitação dos membros em cada classe. Uma das contribuições deste mestrado é a primeira versão da VerbNet.Br, que precisa de validação linguística, mas que já contém informação para ser utilizada em tarefas de PLN, com precisão e abrangência de 44% e 92,89%, respectivamente / Building computational-linguistic base resources, like computational lexical resources (CLR), is one of the goals of Natural Language Processing (NLP). However, most computational lexicons are specific to English. One of the resources already developed for English is the VerbNet, a lexicon with domain-independent semantic and syntactic information of English verbs. It is based on Levin\'s verb classification, with mappings to Princeton\'s WordNet (WordNet). Since only a few computational studies for languages other than English have been made about Levin\'s classification, and given the lack of a Portuguese CLR similar to VerbNet, the goal of this research was to create a CLR for Brazilian Portuguese (called VerbNet.Br). The manual building of these resources is usually unfeasible because it is time consuming and it can include many human-made errors. Therefore, great efforts have been made to build such resources with the aid of computational techniques. One of these techniques is machine learning, a widely known and used method for extracting linguistic information from corpora. Another one is the use of pre-existing resources for other languages, most commonly English, to support the building of new aligned resources, taking advantage of some multilingual/cross-linguistic features (like the ones in Levin\'s verb classification). The method proposed here for the creation of VerbNet.Br is generic, therefore it may be used to build similar resources for languages other than Brazilian Portuguese. Moreover, the proposed method also allows for a future extension of the resource via subclasses of concepts. The VerbNet.Br has a four-step method: three automatic and one manual. However, experiments were also carried out without the manual step, which can be discarded without affecting precision and recall. The evaluation of the resource was intrinsic, both qualitative and quantitative. The qualitative evaluation consisted in: (a) manual analysis of some VerbNet classes, resulting in a Brazilian Portuguese gold standard; (b) comparison of this gold standard with the VerbNet.Br results, presenting promising results (almost 60% of f-measure); and (c), comparison of the VerbNet.Br results to verb clustering results, showing that both methods achieved similar results. The quantitative evaluation considered the acceptance rate of candidate members of VerbNet.Br, showing results around 90% of acceptance. One of the contributions of this research is to present the first version of VerbNet.Br. Although it still requires linguistic validation, it already provides information to be used in NLP tasks, with precision and recall of 44% and 92.89%, respectively

TITLE: Comprehension of syntactic structures in young school-age children ABSTRACT: The thesis deals with the comprehension of syntactic structures in children attending the second grade of elementary school which was part of a larger research project exploring assumptions regarding reading and writing. The comprehension of syntactic structures was studied using an experimental version of a test which monitors the processing of different types of simple phrases and sentences. Data show the level of comprehension of children at this age, how the children understand the sentences, and what the most common mistakes are. We were able to identify the different strategies that are used in the processing of sentences and explain some of the typical ways to erroneous interpretations of sentences using the concepts described in literature. We also investigated relationships with other skills. In addition to analysis of the role of memory in sentence comprehension, the work deals mainly with the relationship between reading efficiency and reading comprehension. The results show the relationship between these variables in the Slovak language, and compare the findings with data from studies in other languages. KEY WORDS: Comprehension of syntactic structures, process of comprehension, thematic roles, comprehension...

With the recent rise in interest in critical readings of our history, scholars have begun noticing that historical documents such as the Anglo-Saxon Chronicle may have functioned as propaganda (e.g., Yorke, 2006; Konshuh, 2020). The present study examines how Britons (i.e., Brythonic-Celtic-speaking people) are discursively represented in the Anglo-Saxon Chronicle. The aim of the study is twofold. First, drawing upon theories from Role and Reference Grammar (RRG, e.g., Van Valin, 1993), this study applies the notion of thematic roles to investigate what roles Britons may take as arguments in the Chronicle. Secondly, the study takes a socio-cognitive approach to discourse (e.g., van Dijk, 2017), conceptualizing arguments and the roles they take as corresponding to the Anglo-Saxons’ shared mental representations of the referents. The results show Britons are construed mainly as passive, often taking the roles of Theme or Patient as they are fought, slain and put to flight. When taking an Agent role, Britons are mostly performing actions related to fear caused by the Anglo-Saxons.  Based on this, it was concluded that the Britons are construed as a cowardly and rather anonymous group, whom the Anglo-Saxons have all the right to eradicate. The portrayal of Britons in the Chronicle furthermore resembles portrayals of other colonized peoples, and it appears that strategies observed in orientalist discourses may have parallels in medieval English discourses.  Moreover, the results provide linguistic empirical evidence for what previous research in history has indicated: that the Anglo-Saxons used opposing groups as part of their narrative in which they were the rightful rulers of Britain (Yorke, 2006; Konshuh, 2020).