Descriptive perception verbs have failed to receive a uniform analysis in previous verb classifications (cf. Chomsky 1965, Rogers 1974, Hengeveld 1992, Levin 1993, Van Eynde et al. 2014). This thesis argues that the descriptive perception verbs in Dutch (i.e. eruitzien 'look', klinken 'sound', voelen 'feel', ruiken 'smell', and smaken 'taste') should be classified as copular verbs, much like lijken 'seem' and schijnen 'seem'. This classification is supported by both the synchronic and diachronic behaviour of these verbs in Dutch. Synchronically, proposing that Germanic copular verbs (as opposed to copulas) are defined by their syntax rather than their (empty) semantics, I discuss that the Dutch descriptive perception verbs behave like stereotypical copular verbs: they require a predicative complement, usually in the form of an adjective. Semantically, the Dutch descriptive perception verbs are much like the copular verbs blijken 'turn out', lijken 'seem' and schijnen 'seem' in terms of epistemicity and evidentiality. Diachronically, I hypothesize that the Dutch descriptive perception verbs have evolved from one of the following two origins: either from intransitive verbs (as is the case for klinken and ruiken), much like English remain, through grammaticalization processes of semantic bleaching and reanalysis; or from cognitive perception verbs (as is the case of eruitzien and voelen), as found in Latin, Japanese and Zulu, through the process of argument reordering. The origin of smaken is not clear, and is left for future research. I show that other Germanic evidential copular verbs (i.e. lijken, schijnen 'seem', scheinen 'seem', seem) have developed diachronically in a uniform fashion, suggesting the following grammaticalization path: from a lexical verb to a copular verb, to taking a that-complement, an infinitival complement or a like-complement, and eventually being used in parenthetical constructions. The results of this thesis indicate that the Dutch descriptive perception verbs are only at the beginning of this grammaticalization path, but are on their way to becoming grammaticalized evidential copular verbs.

Orientador: Marize M. Dall'Aglio-Hattnher / Banca: Flávia Bezerra de Menezes Hirata-Vale / Banca: Anna Flora Brunelli / Resumo: O objetivo do presente trabalho é analisar o funcionamento dos evidenciais e modalizadores no discurso da autoajuda, comparando o uso dessas categorias em dois tipos temáticos desse discurso. O primeiro tipo, o qual denominamos autoajuda genérica, refere a obras em que não há uma especificação dos tipos de objetivos que se pretende alcançar com a aplicação dos métodos propostos pela autoajuda. O segundo tipo, denominado autoajuda da saúde, se refere a obras em que o objetivo é a cura e prevenção de doenças. A hipótese de que partimos é a de que a escolha temática irá influenciar na manifestação da modalidade e da evidencialidade, uma vez que essa escolha implica em relações interdiscursivas específicas. Para a análise dessas categorias, utilizamos a abordagem da Gramática Discursivo-Funcional (HENGEVELD; MACKENZIE, 2008), por se tratar de uma abordagem estratificada do enunciado, possibilitando, dessa forma, melhor compreensão dos diversos efeitos de sentido gerados por cada tipo de modalizador e evidencial. Como resultado de nossa análise, vimos que a relação interdiscursiva da autoajuda da saúde com o discurso médico influenciou a escolha dos modalizadores epistêmicos, com a preferência do enunciador da autoajuda da saúde por modalizadores objetivos, manifestando baixo comprometimento com relação à avaliação realizada. A relação com o discurso médico, influenciou também a escolha dos evidenciais, havendo a preferência do enunciador da autoajuda da saúde por evidenciais reportativos, buscando, assim, aparentar maior cientificidade do que o enunciador da autoajuda genérica. Outra diferença no comportamento dos modalizadores se deu nos domínios deôntico e dinâmico, para os quais constatamos uma maior imposição por parte do enunciador da autoajuda genérica / Abstract: The objective of the present work is to analyze the functioning of evidentials and modalizers in the discourse of Self-help, comparing how these categories are used in two different theme-specified types of the afore mentioned discourse. The first type, denominated Generic Self-help, refers to books in which the type of goals to be attained through self-help is not specified, ranging from financial success to happiness, love, or any other type of goal. The second type, denominated Health Self-help, refers to books in which the sole objective is to cure and prevent diseases. The working hypothesis is that by choosing a certain theme, the interdiscoursive relationships will also be different, resulting in different choices in the use of evidentials and modalizers. In order to analyze these categories, we have adopted the approach of Discoursive Funcional Grammar (HENGEVELD; MACKENZIE, 2008), for its stratified view of the utterance allows to capture effectively the various effect senses produced by the different types of modalizers and evidentials. As a result of the analysis, we observed that the relationship with the medical discourse in the Health Self-Help discourse influenced in the use of epistemic modality, with the preference for objective modalizers. Moreover, the medical discourse influenced in the use of evidentials, noticed by the preference in the Health Self-help discourse for reportative evidentials. Also it was noticed the difference in the use of deontic and dynamic modalities, with the preference for more imposing forms in the Generic Self-help. Since the two theme-specified types are from the same discourse, we found also similarities in the use of evidentials and modalizers, the most important one being the compromise with certainty evaluations in opposition to the lack of commitment to uncertainty evaluations / Mestre

In this dissertation, a normative prescriptive paradigm, namely a Bayesian theory of content-dependent argument strength, was employed in order to investigate argumentation, specifically the classic fallacy of the &ldquo / argument from ignorance&rdquo / or &ldquo / argumentum ad ignorantiam&rdquo / . The study was carried out in Turkish with Turkish participants. In the Bayesian framework, argument strength is determined by the interactions between three major factors: prior belief, polarity, and evidence reliability. In addition, topic effects are considered. Three experiments were conducted. The first experiment replicated Hahn et al.&rsquo / s (2005) study in Turkish to investigate whether similar results would be obtained in a different linguistic and cultural community. We found significant main effects of three of the manipulated factors in Oaksford and Hahn (2004) and Hahn et al. (2005): prior belief, reliability and topic. With respect to the Bayesian analysis, the overall fit between the data and the model was very good. The second experiment tested the hypothesis that argument acceptance would not vary across different intelligence levels. There was no significant main effect of prior belief, polarity, topic, and intelligence. We found a main effect of reliability only. However, further analyses on significant interactions showed that more intelligent subjects were less inclined to accept negative polarity items. Finally, the third experiment investigated the hypothesis that argument acceptance would vary depending on the presence of and the kind of evidentiality markers prevalent in Turkish, indicating the certainty with which events in the past have happened, marked with overt morpho-syntactic markers (&ndash / DI or &ndash / mIs). The experiment found a significant main effect of evidentiality as well as replicating the significant main effects of the two of the manipulated factors (prior belief and reliability) in Oaksford and Hahn (2004), Hahn et al. (2005) and in our first experiment. Furthermore, reliability and evidentiality interacted, indicating separate as well as combined effects of the two. With respect to the Bayesian analysis, the overall fit between the data and the model was lower than the one in the first experiment, but still acceptable. Overall, this study supported the normative Bayesian approach to studying argumentation in an interdisciplinary perspective, combining computation, psychology, linguistics, and philosophy.

In this study, the development of a second-order false belief task is investigated by considering the impact of the acquisition of Turkish evidential markers, namely &ndash / DI (direct evidence) and &ndash / mIs (inference or hearsay). A neutral version of the tasks served as a control form. 21 kindergarten children (aged 4-5 years), 47 primary school children (aged 6- 12 years) and 10 adults participated in the study. Our results revealed that there is no effect of acquisition of evidentials on false belief understanding. Together with the other studies, there is a facilitative effect of &ndash / DI (direct evidence) in understanding of stories/narratives in general rather than false belief understanding for the children at the age of 4 to 6/7. In addition to the second-order false belief tasks (FBT_2), a simple working memory task (WST), a complex working memory task (LST), a perspective taking task (PTT) and a double- embedded relative clause task (REL_2) were used in order to investigate the developmental trend of these tasks and their possible relationship with second-order false belief understanding. Also, to the best of our knowledge this is the first time that a REL_2 task has been devised in a Turkish study. The general developmental trend was found for all tasks. Even if some significant correlations were found for FBT_2 score predicted from other tasks, analyses showed that only the contribution of age was significant. Since all of these domains are not related to second-order false belief reasoning but develop at the same time, it is not incompatible with the serial bottleneck hypothesis. In sum, the findings are matching with the modularity view that ToM is a faculty of the human mind at their own pace that does not share intrinsic content with other faculties such as language and working memory (Leslie et al., 2004). However, it develops together with those other faculties and they may constrain the expression of child

Disertacijoje analizuojami ne tik vientisinių būtųjų lietuvių kalbos laikų reikšmės ir jų atitikmenys anglų kalboje, bet ir praeities raiškos formos, įtraukiant ir vadinamąsias sudurtines arba sudėtines laikų formas, dalyvius, vartojamus vietoj tiesioginės nuosakos atpasakojamojoje kalboje, ir predikatines nelinksniuojamas neveikiamosios rūšies būtojo laiko dalyvių formas. Lietuvių kalboje yra du vientisiniai laikai: būtasis kartinis ir būtasis dažninis. Abiems šiems laikams bendra tai, kad jie reiškia veiksmą, vykusį prieš kalbamąjį momentą, o skiria juos pagrindinė santykio su kalbamuoju momentu reikšmė: būtasis kartinis laikas gali reikšti labai artimą veiksmą atskaitos momentui: ar tai būtų dabartis, ar kitas praeities veiksmas. Taigi būtasis kartinis laikas turi ir perfekto, ir pliuskvamperfekto reikšmes. Šių reikšmių neturi būtasis dažninis laikas. Jis reiškia tolesnės praeities kartojamą veiksmą, kurio atžvilgiu jis sudaro opoziciją būtajam kartiniam laikui. Darbe taip pat nagrinėjamos analitinės formos, nustatomas jų santykis su vientisinėmis: ar jos sudaro vientisinėms opozicijas, ar yra sinoniminės, ar visos santykiauja vienodai. Taip pat tiriama vadinamoji netiesioginė nuosaka (modus relativus) ir neveikiamosios rūšies formą turinčių predikatinių dalyvių reikšmė. Analitinis anglų kalbos laikų pobūdis geriau padeda suprasti ir įvertinti lietuvių kalbos laikų reikšmes ir funkcijas. Ieškojimas lietuvių kalbos laikų formoms atitikmenų anglų kalboje padeda geriau... [toliau žr. visą tekstą] / The meanings of Lithuanian past simple tenses and their equivalents in English is the object of the thesis. Other forms expressing past, i.e. compound forms, participles used instead of indicative mood in the narrative speech as well as predicative indeclinable forms of past passive participles are the subject matter in the thesis. There are two simple past tenses in Lithuanian: the past simple and the past frequentative tense. Both tenses are used to express an action which happened before the moment of speech; however, the relation to the moment of speech is different: the past simple tense is used to express a very close action to the reference time (whether it is present or other past action). Thus, the past simple tense has the perfective and plusquamperfective meaning. Whereas, the past frequentative tense does not possess those meanings; it is used to express a repetitive action of a more distant past. Moreover, the relationship of analytic forms to the simple ones is determined in the study. The problem of modus relativus and the meaning of predicative participles of passive voice are covered in the thesis as well. Analytical nature of English tenses enables to perceive and evaluate the meanings and functions of Lithuanian tenses more accurately. The search for the equivalents of Lithuanian tenses in English allows determining more precise meanings of Lithuanian past tenses, particularly the grammaticalization of analytic constructions, the relationships of... [to full text]

The aim of the research is to analyse the expression and contents of non-morphological evidentiality in Lithuanian realized by neuter (non-agreeing) adjectives and adverbs. The thesis focuses on the non-agreeing adjectives and the adverbs that derive from the semantic domains of perception, comparison and knowledge and thus have the potential to acquire evidential meanings. The study explores the morphosyntactic properties (Complement-Taking-Predicates and adverbials) of the markers under consideration, their quantitative parameters and semantic functional distribution in fiction and academic discourse. The study is corpus-driven and the data have been obtained from the Corpus of the Contemporary Lithuanian Language, namely from the subcorpus of fiction, and from the Corpus of Academic Lithuanian. The analysis of the morphosyntactic properties of the markers and their semantic functional distribution suggests that in Lithuanian the non-agreeing adjectives convey evidential meanings more frequently than the adverbs because the adverbs mainly function as predicate modifiers. The main evidential value is inference based on perceptual or conceptual sources of evidence. The non-agreeing adjectives and the adverbs under consideration can also express the meanings of epistemic modality, expectation or function as pragmatic markers. / Pagrindinis darbo tikslas yra išanalizuoti nemorfologinio evidencialumo raišką ir turinį, realizuojamus bevardės giminės (nederinamaisiais) būdvardžiais ir prieveiksmiais grožinės literatūros tekstuose ir lietuvių mokslo kalboje. Tyrimo objektą sudaro nederinamieji būdvardžiai ir prieveiksmiai, kurių leksinės reikšmės atspindi semantinius laukus, turinčius potencialą žymėti žinių šaltinį. Darbe analizuojami šių vienetų morfosintaksiniai bruožai (komplementiniai predikatai ir adverbialai), tiriami jų kiekybiniai rodikliai ir atskleidžiama semantinė funkcinė distribucija. Šioje disertacijoje atliekamas tyrimas naudojant tekstynų metodologiją. Tyrimo medžiaga yra surinkta iš grožinės literatūros patekstynio, esančio Dabartinės lietuvių kalbos tekstyne ir Lietuvių mokslo kalbos tekstyno. Nederinamųjų būdvardžių ir prieveiksmių morfosintaksinių bruožų ir semantinių funkcinių ypatybių analizė suponuoja, kad lietuvių kalboje nederinamieji būdvardžiai dažniau perteikia žinių šaltinio reikšmes negu prieveiksmiai, nes šie vartojami kaip predikato modifikatoriai. Pagrindinė evidencinė reikšmė, būdinga nagrinėjamiems žymikliams, yra numanymas, pagrįstas percepciniu arba konceptualiuoju žinių šaltiniu. Nagrinėjami vienetai vartojami ir kaip episteminiai, vertinimo arba pragmatiniai žymikliai.

Pagrindinis darbo tikslas yra išanalizuoti nemorfologinio evidencialumo raišką ir turinį, realizuojamus bevardės giminės (nederinamaisiais) būdvardžiais ir prieveiksmiais grožinės literatūros tekstuose ir lietuvių mokslo kalboje. Tyrimo objektą sudaro nederinamieji būdvardžiai ir prieveiksmiai, kurių leksinės reikšmės atspindi semantinius laukus, turinčius potencialą žymėti žinių šaltinį. Darbe analizuojami šių vienetų morfosintaksiniai bruožai (komplementiniai predikatai ir adverbialai), tiriami jų kiekybiniai rodikliai ir atskleidžiama semantinė funkcinė distribucija. Šioje disertacijoje atliekamas tyrimas naudojant tekstynų metodologiją. Tyrimo medžiaga yra surinkta iš grožinės literatūros patekstynio, esančio Dabartinės lietuvių kalbos tekstyne ir Lietuvių mokslo kalbos tekstyno. Nederinamųjų būdvardžių ir prieveiksmių morfosintaksinių bruožų ir semantinių funkcinių ypatybių analizė suponuoja, kad lietuvių kalboje nederinamieji būdvardžiai dažniau perteikia žinių šaltinio reikšmes negu prieveiksmiai, nes šie vartojami kaip predikato modifikatoriai. Pagrindinė evidencinė reikšmė, būdinga nagrinėjamiems žymikliams, yra numanymas, pagrįstas percepciniu arba konceptualiuoju žinių šaltiniu. Nagrinėjami vienetai gali būti vartojami ir kaip episteminiai, vertinimo arba pragmatiniai žymikliai. / The aim of the research is to analyse the expression and contents of non-morphological evidentiality in Lithuanian realized by neuter (non-agreeing) adjectives and adverbs. The thesis focuses on the non-agreeing adjectives and the adverbs that derive from the semantic domains of perception, comparison and knowledge and thus have the potential to acquire evidential meanings. The study explores the morphosyntactic properties (Complement-Taking-Predicates and adverbials) of the markers under consideration, their quantitative parameters and semantic functional distribution in fiction and academic discourse. The study is corpus-driven and the data have been obtained from the Corpus of the Contemporary Lithuanian Language, namely from the subcorpus of fiction, and from the Corpus of Academic Lithuanian. The analysis of the morphosyntactic properties of the markers and their semantic functional distribution suggests that in Lithuanian the non-agreeing adjectives convey evidential meanings more frequently than the adverbs because the adverbs mainly function as predicate modifiers. The main evidential value is inference based on perceptual or conceptual sources of evidence. The non-agreeing adjectives and the adverbs under consideration can also express the meanings of epistemic modality, expectation or function as pragmatic markers.

Made available in DSpace on 2014-06-11T19:30:28Z (GMT). No. of bitstreams: 0 Previous issue date: 2010-08-13Bitstream added on 2014-06-13T19:40:09Z : No. of bitstreams: 1 vendrame_v_dr_sjrp.pdf: 1109440 bytes, checksum: 425cc6fee8d27dd9cb5223e5c30b8dbe (MD5) / A evidencialidade corresponde à explicitação da fonte da informação contida em um enunciado. Em muitas línguas, ela é codificada por meios gramaticais que indicam se a informação transmitida foi obtida de maneira direta, por meio de percepção sensorial, ou de maneira indireta, por meio de um relato ou uma inferência. Apesar de existir um grande número de trabalhos que tratam da evidencialidade gramatical, poucos são os estudos que consideram a evidencialidade expressa lexicalmente. A presente pesquisa surge como uma proposta de descrição da evidencialidade lexical expressa por meio de verbos de percepção em língua portuguesa. Mais especificamente, busca-se investigar quais são os tipos de contextos sintático-semânticos em que os verbos de percepção ver, ouvir e sentir têm valor evidencial e quais sentidos evidenciais estão relacionados com cada contexto. A teoria que embasa este estudo é a da Gramática Discursivo-Funcional (GDF), um modelo de gramática que “procura entender como unidades linguísticas são estruturadas em termos da realidade que elas descrevem e das intenções comunicativas com as quais elas são produzidas, e molda isso em uma implementação dinâmica da gramática” (HENGEVELD; MACKENZIE, 2008, p.2). Dentro da GDF, a evidencialidade é descrita em termos de níveis ou camadas de acordo com as características semântico-pragmáticas e morfossintáticas que a estrutura evidencial apresenta. O material de análise desta pesquisa é composto por amostras do português brasileiro falado e escrito. A partir da análise dos dados, foi possível identificar quatro tipos evidenciais codificados pelos verbos analisados: evidencialidade reportativa, expressa pelos verbos ver e ouvir, evidencialidade inferida, deduzida e direta, esses três tipos codificados pelos verbos ver, ouvir e sentir. Uma vez que um mesmo verbo pode codificar diferentes... / Evidentiality is understood as the marking of the source of evidence a speaker has for his/her statement. In many languages, it is expressed by grammatical means that indicate whether the information was obtained in a direct way, via sensorial perception, or in an indirect way, via report or inference. Although there are many studies on grammatical evidentiality, only a few take lexical evidentiality into account. This research aims at describing lexical evidentiality expressed by perception verbs in Portuguese by investigating the verbs ver (to see), ouvir (to hear) and sentir (to feel) in order to find out the evidential meanings related to them and the syntactic and semantic contexts in which they occur. The theoretical background of accounting perception verbs here rests on Functional Discourse-Grammar (FDG), a model of grammar which “aims to understand how linguistic units are structured in terms of the world they describe and the communicative intentions with which they are produced, and models this in a dynamic implementation” (HENGEVELD; MACKENZIE, 2008, p.2). Within FDG, evidentiality is described in terms of levels and layers according to the semantic, pragmatic and morphosyntactic features the evidential structure presents. The data analyzed here comprise samples of spoken and written Brazilian Portuguese. The results show that the three verbs are used to express four types of evidentiality: reportative evidentiality, expressed by ver and ouvir, inferential, deductive and direct evidentiality, all these three kinds expressed by ver, ouvir and sentir. Taking into account that the same verb may be used with different types of evidentiality, a number of syntactic, semantic and pragmatic aspects were considered in the description of each type. FDG model was very useful in the approach to the phenomenon under study. The syntactic description of the data showed... (Complete abstract click electronic access below)

Submitted by Aline Maria Miguel Kapp null (aline_kapp@hotmail.com) on 2017-10-26T11:03:01Z No. of bitstreams: 1 KAPP-BARBOZA, A. M. M. (2017) - USOS DO VERBO SABER E A EXPRESSÃO DA EVIDENCIALIDADE NO PORTUGUÊS BRASILEIRO.pdf: 2197142 bytes, checksum: 5e2dd5a434b2188e51dc499bf52cdb29 (MD5) / Approved for entry into archive by Luiz Galeffi (luizgaleffi@gmail.com) on 2017-10-26T15:45:52Z (GMT) No. of bitstreams: 1 kappbarboza_amm_dr_sjrp.pdf: 2197142 bytes, checksum: 5e2dd5a434b2188e51dc499bf52cdb29 (MD5) / Made available in DSpace on 2017-10-26T15:45:52Z (GMT). No. of bitstreams: 1 kappbarboza_amm_dr_sjrp.pdf: 2197142 bytes, checksum: 5e2dd5a434b2188e51dc499bf52cdb29 (MD5) Previous issue date: 2017-08-07 / Considerando que a ideia de conhecimento prevê uma acessibilidade entre objetos linguísticos e cognitivos que só ganha sentido mediante a noção de uso, já que o conhecimento não ocorre fora de experiências significativas do sujeito com as referências do mundo ou das ações simbólicas humanas, o objetivo deste trabalho é descrever o verbo saber como um verbo de múltiplas funções no português do Brasil, buscando confirmar a hipótese de que ele pode ser considerado forma de expressão da evidencialidade lexical. Basicamente classificado como verbo de cognição, o verbo saber apresenta outras funcionalidades, como o uso modal epistêmico, em que se avalia o (des)comprometimento do sujeito com relação à proposição que ele apresenta, e o uso modal facultativo, em que se avalia a capacidade em realizar um estado-de-coisas. Verifica-se também um uso evidencial reportativo de domínio comum desse predicado, por meio do qual uma informação é apresentada como um conhecimento compartilhado entre falante e ouvinte. Além desses usos, nossa hipótese geral é a de que o verbo saber serve também à expressão da evidencialidade inferencial e dedutiva, respeitando a hierarquia dos evidenciais proposta por Hengeveld e Hattnher (2015). Assim, tomando como embasamento teórico-metodológico a Gramática DiscursivoFuncional, investigamos os domínios pragmático-semânticos e morfossintáticos que envolvem a multifuncionalidade desse verbo. A aceitação desse aparato teórico direciona para a escolha de um corpus que considere usos reais da língua, aqui representados pelo Banco de Dados Iboruna, pelo Corpus do Português e por ocorrências retiradas da internet. Para analisar os diferentes usos desse verbo, recorre-se à organização em camadas, e seus diferentes níveis hierárquicos de análise, proposta pela Gramática Discursivo-Funcional (HENGEVELD; MACKENZIE, 2008, p.13). As análises confirmaram a existência de dois outros usos evidenciais do verbo saber ainda não descritos na literatura: inferencial e dedutivo. Além disso, esta pesquisa traz também como contribuição a distinção entre conhecimento e informação no que se refere às ocorrências com esse verbo. Demonstramos, portanto, ao longo desta investigação, que o verbo saber serve à expressão da evidencialidade lexical em português brasileiro, e, com isso, identificamos sete contextos de uso: cognitivo, informativo, modal epistêmico, modal facultativo, evidencial reportativo, evidencial inferencial e evidencial dedutivo. / Considering that the idea of knowledge provides an accessibility between linguistic and cognitive objects that is only meaningful through the notion of use, since the knowledge does not occur outside the significant experiences of the subject with the references from the world or from the human symbolic actions, the aim of this study is to describe the verb saber as a multiple functional verb in Brazilian Portuguese, seeking to confirm the hypothesis that it can be considered a form of expression of lexical evidentiality. Basically classified as a cognition verb, the verb saber presents other functionalities, such as the epistemic modality, in which the subject's commitment to the proposition he or she presents is evaluated; and the facultative modality, in which the ability of a participant to engage in the state-of-affairs is evaluated. There is also a reportative evidential use of the common knowledge of this predicate, in which the information is presented as shared knowledge between the speaker and the listener. Besides these uses, our hypothesis is that the verb saber can be also used to convey inferential and deductive evidentiality, observing the evidentials hierarchy proposed by Hengeveld and Hattnher (2015). Thus, based on the Functional Discourse Grammar theoretical-methodological approach we investigate the pragmatic-semantic and morphosyntactic domains involving the multifunctionality of this verb. The acceptance of this theoretical approach leads to the choice of a corpus which considers real uses of the language, represented here by Iboruna Database, Corpus do Português and by occurrences taken from the internet. In order to analyze the different uses of this verb, it is used the layered organization, and its different hierarchical levels of analysis, proposed by the Functional Discourse Grammar (HENGEVELD; MACKENZIE, 2008, p.13). In this hierarchical context, the different semantic expressions of the verb saber are classified. The analysis confirmed the existence of two other evidential uses of the verb saber not yet described in the literature: inferential and deductive. In addition, this research also contributes to the distinction between knowledge and information regarding occurrences with this verb. We demonstrate, therefore, throughout the investigation, that the verb saber expresses lexical evidentiality in Brazilian Portuguese, and, hence, we identify seven contexts of use: cognitive, informative, epistemic modality, facultative modality, reportative evidentiality, inferencial evidentiality and deductive evidentiality.

Orientador: Marize Mattos Dall'Aglio Hattnher / Banca: Kees Hengeveld / Banca: Rosane de Andrade Berlinck / Banca: Sebastião Carlos Leite Gonçalves / Banca: Sandra Denise Gasparini Bastos / Resumo: A evidencialidade corresponde à explicitação da fonte da informação contida em um enunciado. Em muitas línguas, ela é codificada por meios gramaticais que indicam se a informação transmitida foi obtida de maneira direta, por meio de percepção sensorial, ou de maneira indireta, por meio de um relato ou uma inferência. Apesar de existir um grande número de trabalhos que tratam da evidencialidade gramatical, poucos são os estudos que consideram a evidencialidade expressa lexicalmente. A presente pesquisa surge como uma proposta de descrição da evidencialidade lexical expressa por meio de verbos de percepção em língua portuguesa. Mais especificamente, busca-se investigar quais são os tipos de contextos sintático-semânticos em que os verbos de percepção ver, ouvir e sentir têm valor evidencial e quais sentidos evidenciais estão relacionados com cada contexto. A teoria que embasa este estudo é a da Gramática Discursivo-Funcional (GDF), um modelo de gramática que "procura entender como unidades linguísticas são estruturadas em termos da realidade que elas descrevem e das intenções comunicativas com as quais elas são produzidas, e molda isso em uma implementação dinâmica da gramática" (HENGEVELD; MACKENZIE, 2008, p.2). Dentro da GDF, a evidencialidade é descrita em termos de níveis ou camadas de acordo com as características semântico-pragmáticas e morfossintáticas que a estrutura evidencial apresenta. O material de análise desta pesquisa é composto por amostras do português brasileiro falado e escrito. A partir da análise dos dados, foi possível identificar quatro tipos evidenciais codificados pelos verbos analisados: evidencialidade reportativa, expressa pelos verbos ver e ouvir, evidencialidade inferida, deduzida e direta, esses três tipos codificados pelos verbos ver, ouvir e sentir. Uma vez que um mesmo verbo pode codificar diferentes... (Resumo completo, clicar acesso eletrônico abaixo) / Abstract: Evidentiality is understood as the marking of the source of evidence a speaker has for his/her statement. In many languages, it is expressed by grammatical means that indicate whether the information was obtained in a direct way, via sensorial perception, or in an indirect way, via report or inference. Although there are many studies on grammatical evidentiality, only a few take lexical evidentiality into account. This research aims at describing lexical evidentiality expressed by perception verbs in Portuguese by investigating the verbs ver (to see), ouvir (to hear) and sentir (to feel) in order to find out the evidential meanings related to them and the syntactic and semantic contexts in which they occur. The theoretical background of accounting perception verbs here rests on Functional Discourse-Grammar (FDG), a model of grammar which "aims to understand how linguistic units are structured in terms of the world they describe and the communicative intentions with which they are produced, and models this in a dynamic implementation" (HENGEVELD; MACKENZIE, 2008, p.2). Within FDG, evidentiality is described in terms of levels and layers according to the semantic, pragmatic and morphosyntactic features the evidential structure presents. The data analyzed here comprise samples of spoken and written Brazilian Portuguese. The results show that the three verbs are used to express four types of evidentiality: reportative evidentiality, expressed by ver and ouvir, inferential, deductive and direct evidentiality, all these three kinds expressed by ver, ouvir and sentir. Taking into account that the same verb may be used with different types of evidentiality, a number of syntactic, semantic and pragmatic aspects were considered in the description of each type. FDG model was very useful in the approach to the phenomenon under study. The syntactic description of the data showed... (Complete abstract click electronic access below) / Doutor