The purpose of this research was to test the predictions of the Accessibility Hierarchy (AH) theory (Keenan & Comrie, 1977) applying it to the Russian language. According to this theory, relative clauses (RC) are acquired in a fixed unidirectional order: from subject (S) - the highest (unmarked) and more susceptible to relativization position - to object of comparative (OCOM) - the lowest (marked) and less susceptible to relativization position. Since some researchers (Hamilton, 1994) claim that the AH is multidirectional rather than unidirectional, this study takes into consideration these findings as well. The present study attempts to determine (a) if learners of the Russian language are able to make generalizations about more unmarked RC positions after receiving instruction only on a relatively marked relative clause position (in this study it is OPR - object of a preposition), and (b) if instruction on unmarked relative clause position facilitates learners' ability to generalize that learning to marked relative clauses. Participants of the study were Brigham Young University students studying Russian as a second language. Two groups, the basic treatment group (BG) and the complex treatment group (CG) with a total of fifty-four subjects, completed pretests and posttests, each of which included two elicitation tasks: a combination test (CT) and a grammaticality judgment test (GJT). Both groups received instruction between the tests. The BG received instruction on the subject (S), the direct object (DO), and the indirect object (IO) RC positions. The CG received instruction only on the OPR position. Three types of error, incorrect adjacency, incorrect morphological RC ending, and pronoun retention, were analyzed separately. In addition, the CT investigated the acquisition of pied-piping structure in the OPR and GEN types. The results of the research support Hamilton's (1994) findings and suggest that generalization is clearly not unidirectional. Regardless of type of instruction the subjects of both groups generalized their learning in both directions.

Made available in DSpace on 2014-06-11T19:22:19Z (GMT). No. of bitstreams: 0 Previous issue date: 2011-07-28Bitstream added on 2014-06-13T19:27:29Z : No. of bitstreams: 1 oliveira_gm_me_sjrp.pdf: 694678 bytes, checksum: d0b469f7cacf82b729e428d0d985aa22 (MD5) / Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP) / O objetivo deste trabalho é investigar a atuação de restrições de ordem semântica ao lado das de ordem sintática para a determinação do processo de formação de orações relativas, em busca de evidências que confirmem a revisão da Hierarquia de Acessibilidade (HA) de Keenan e Comrie (1977) proposta por Dik (1997). As críticas à HA de Keenan e Comrie (1977) dizem respeito tanto a razões empíricas – pelo fato de que em algumas línguas não há nenhuma forma de se construir orações relativas; quanto a razões teóricas – pelo fato de as funções sintáticas não serem relevantes para todas as línguas do mundo. O corpus deste trabalho é composto por 30 línguas indígenas brasileiras e conta com descrições previamente feitas, como gramáticas, teses e outros tipos de manuais descritivos. Os dados coletados por este trabalho confirmam a hipótese de que é necessária uma revisão da HA. As lacunas na HA dizem respeito, principalmente, à função de Objeto Indireto. Para várias das línguas investigadas, essa função sintática não está acessível à relativização, mas outras funções, mais baixas na hierarquia do que essa, podem ser relativizadas, contrariando a HA de Keenan e Comrie (1977). Ao se analisarem as funções semânticas relativizadas, concluiu-se que as funções de Recipiente, Locativo e Tempo têm o mesmo estatuto nas línguas, hipótese confirmada pela teoria da Gramática Discursivo-Funcional. Em consequência disso, propomos neste trabalho uma outra hierarquia, baseada em critérios sintáticos e semânticos. Além disso, a nominalização desponta, neste trabalho, como a estratégia de relativização mais recorrente, apesar de não ser aceita como estratégia legítima para alguns autores de orientação formalista / This study aims at investigating the role of semantic and syntactic constraints to determine the process of relative clauses formation, in search of evidence that confirms the revision of the Accessibility Hierarchy (AH), by Keenan and Comrie (1977), proposed by Dik (1997). The criticism to the HA, by Keenan and Comrie (1977), concern both empirical reasons (the fact that there are languages that do not have any type of relative construction) and theoretical reasons (the fact that the authors take syntactic categories as universal without mentioning any theory that defines them and without taking into account that there are languages for which these functions are not relevant). The corpus of this work consists of 30 Brazilian indigenous languages and includes descriptions made previously, such as grammar books, theses and other types of descriptive manuals. Data collected by this study confirm the hypothesis that a review is needed for AH. Gaps in the AH relate mainly to the function of Indirect Object. For many languages, such function is not accessible to relativization, but other functions, lower in the hierarchy, can be relativized. When analyzing the relativized semantic functions, it was concluded that the semantic functions of Recipient, Locative and Time have the same status in the languages, a hypothesis that is also confirmed by the theory of Functional Discourse Grammar. In addition, in this study, nominalization happens to be the most recurrent strategy of relativization, although it is not accepted as a legitimate strategy for some more formalistic authors

El presente artículo constituye una aproximación psicolingüística al estudio de la ocurrencia de pronombres relativos (PR) en cláusulas relativas (CR) en español, tomando en consideración que este tipo de construcciones corresponden a variedades no estándares de la lengua y suelen ser censuradas en las gramáticas descriptivas. Dado que no existen estudios experimentales sobre el tema en español, este trabajo se basa parcialmente en un estudio llevado a cabo por Ferreira y Swets (2004) sobre un fenómeno similar en el inglés. La investigación tiene como objetivos determinar si los PR de objeto directo y oblicuos son una estrategia de reparación de una oración poco planificada y si los PR de sujeto pueden aparecer en situaciones controladas a partir del contraste de dos posibles antecedentes. / This article constitutes a psycholinguistic approach to the study of the occurrence of relative pronouns (RP) in relative clauses (RC) in Spanish, considering that this type of construction is common in nonstandard varieties of the language and are often censored in the descriptive grammars.Since there are no experimental studies on the subject in Spanish, this work is based partly on a study conducted by Ferreira and Swets (2004) on a similar phenomenon in English. The research aims to determine whether direct object and oblique RP are a repair strategy for insufficiently planned sentences and if the subject RP may appear in controlled situations derived from the contrast of two possible antecedents

Cette thèse étudie la syntaxe et la typologie des propositions relatives en arménien parlé. Elle propose une analyse syntaxique et une classification des stratégies de relativisation disponibles en arménien, dans le cadre des approches théoriques et typologiques existantes de la relativisation, ainsi qu’à identifier les facteurs décisifs pour le choix des différentes stratégies. Chaque stratégie ayant des équivalents dans d'autres langues de la région, le contact linguistique peut influencer leur choix. Il est également probable que le rôle de l'élément relativisé dans la proposition relative soit pertinent pour le choix de la stratégie. En particulier, les données présentées dans de précédentes études sur l'arménien suggèrent que la distribution des relatives participiales peut constituer une violation de la hiérarchie d'accessibilité de la relativisation si elle est envisagée comme opérant directement en termes de fonctions grammaticales syntaxiques. L’étude se concentre sur les parlers de diverses régions d’Arménie. Les propositions relatives ont été saisies dans une base de données d’environ 2000 exemples avec des filtres portant sur les paramètres pertinents.Les résultats montrent que l’accessibilité à la relativisation est déterminée par la saillance liée aux rôles sémantiques (affectivité) et pragmatiques (thematicité) et par de fréquents modèles d’association rôle-référence plutôt que directement par des fonctions grammaticales syntaxiques. Ceci fournit une explication cohérente des violations apparentes de la hiérarchie d'accessibilité trouvées en arménien, ainsi que d'autres phénomènes qui se sont révélés problématiques pour les interprétations de l'accessibilité à la relativisation basées directement sur la structure syntaxique. / This thesis presents a study of the syntax and typology of relative clauses in colloquial Armenian. It proposes a syntactic analysis and classification of the relativization strategies available in Armenian within the framework of existing syntactic theoretical and typological proposals concerning relative clauses, and to identify the decisive factors associated with the distribution of these different strategies. As each of the available strategies is paralleled in other languages of the area, it is possible that language contact will have an impact on the choice of strategy. There is also evidence that the role of the relativized element in the relative clause is relevant for the choice of strategy; in particular, evidence presented in previous studies of Armenian suggests that the distribution of participial RCs may violate the Relativization Accessibility Hierarchy if this is envisaged as operating directly in terms of syntactic grammatical relations.The study is mainly based on data from sound recordings of native speaker consultants from various areas of Armenia, both spontaneous speech and responses to stimuli designed to elicit relative clauses with particular properties that have been proposed to affect the choice of relativization strategy. The relative clauses are entered into a database with filters for relevant features, which contains approximately 2000 examples. The results show that accessibility to relativization is determined by semantic (affectedness) and pragmatic (topicality) role prominence, and by frequent role-reference association patterns, rather than directly by syntactic grammatical relations. This provides a coherent explanation of the apparent Accessibility Hierarchy violations found in Armenian, as well as other phenomena that have proved problematic for syntactic structure-based interpretations of accessibility to relativization.

Orientador: Roberto Gomes Camacho / Banca: Angel Humberto Cobera Mori / Banca: Giseli Cássia de Sousa / Resumo: O objetivo deste trabalho é investigar a atuação de restrições de ordem semântica ao lado das de ordem sintática para a determinação do processo de formação de orações relativas, em busca de evidências que confirmem a revisão da Hierarquia de Acessibilidade (HA) de Keenan e Comrie (1977) proposta por Dik (1997). As críticas à HA de Keenan e Comrie (1977) dizem respeito tanto a razões empíricas - pelo fato de que em algumas línguas não há nenhuma forma de se construir orações relativas; quanto a razões teóricas - pelo fato de as funções sintáticas não serem relevantes para todas as línguas do mundo. O corpus deste trabalho é composto por 30 línguas indígenas brasileiras e conta com descrições previamente feitas, como gramáticas, teses e outros tipos de manuais descritivos. Os dados coletados por este trabalho confirmam a hipótese de que é necessária uma revisão da HA. As lacunas na HA dizem respeito, principalmente, à função de Objeto Indireto. Para várias das línguas investigadas, essa função sintática não está acessível à relativização, mas outras funções, mais baixas na hierarquia do que essa, podem ser relativizadas, contrariando a HA de Keenan e Comrie (1977). Ao se analisarem as funções semânticas relativizadas, concluiu-se que as funções de Recipiente, Locativo e Tempo têm o mesmo estatuto nas línguas, hipótese confirmada pela teoria da Gramática Discursivo-Funcional. Em consequência disso, propomos neste trabalho uma outra hierarquia, baseada em critérios sintáticos e semânticos. Além disso, a nominalização desponta, neste trabalho, como a estratégia de relativização mais recorrente, apesar de não ser aceita como estratégia legítima para alguns autores de orientação formalista / Abstract: This study aims at investigating the role of semantic and syntactic constraints to determine the process of relative clauses formation, in search of evidence that confirms the revision of the Accessibility Hierarchy (AH), by Keenan and Comrie (1977), proposed by Dik (1997). The criticism to the HA, by Keenan and Comrie (1977), concern both empirical reasons (the fact that there are languages that do not have any type of relative construction) and theoretical reasons (the fact that the authors take syntactic categories as universal without mentioning any theory that defines them and without taking into account that there are languages for which these functions are not relevant). The corpus of this work consists of 30 Brazilian indigenous languages and includes descriptions made previously, such as grammar books, theses and other types of descriptive manuals. Data collected by this study confirm the hypothesis that a review is needed for AH. Gaps in the AH relate mainly to the function of Indirect Object. For many languages, such function is not accessible to relativization, but other functions, lower in the hierarchy, can be relativized. When analyzing the relativized semantic functions, it was concluded that the semantic functions of Recipient, Locative and Time have the same status in the languages, a hypothesis that is also confirmed by the theory of Functional Discourse Grammar. In addition, in this study, nominalization happens to be the most recurrent strategy of relativization, although it is not accepted as a legitimate strategy for some more formalistic authors / Mestre