The aim of this dissertation was to examine how Japanese learners of English in the United States developed their interpretation of antecedents of reflexive pronouns and their understanding of the direction of verb gapping in English over time. The two properties in question operate differently in English and Japanese, and neither property is normally taught. Therefore, the acquisition of these properties of English by Japanese learners sheds light on the role of Universal Grammar (UG) in second language acquisition (SLA).Fourteen Japanese learners answered three questionnaires, each containing a test assessing reflexives (the Binding Test) and a test assessing verb gapping (the Gapping Test) at three different times, approximately 12 weeks apart from each other. In addition, three groups of 20 native speakers of English provided their results on the same questionnaires.The overall performance of the learners on the Binding Test started out significantly inferior to that of native speakers, and it did not reach the level of native speakers in later sessions. In contrast, the overall performance of the learners on the Gapping Test started out as good as that of native speakers, and it maintained the same level relative to the native performance in later sessions. Furthermore, the overall performance of the learners in later sessions was not significantly better than that in earlier sessions. These findings seem to suggest that increased time spent in an English-speaking environment did not have any effects on the learners' acquisition of the two properties. However, when performance by subset was considered, significant improvement was observed.Examination of individual patterns of responses revealed that the learners showed the patterns consistent to English, Japanese, and even other languages and that some of the learners who previously showed a pattern other than the English pattern successfully demonstrated the English pattern in later sessions. These findings constitute evidence for the view that parameter-resetting is possible in SLA.The dissertation also includes discussion of the test instrument and explanation of the data in the light of recent theoretical predictions. The concluding chapter offers pedagogical implications as well as suggestions for future UG-based SLA research. / Department of English

Thesis (M.A.)--University of Hawaii at Manoa, 2005. / Includes bibliographical references (leaves 133-139). / ix, 139 leaves, bound ill. 29 cm

Sei G eine p-adische analytische gruppe, welche die direkte Summe einer torsionfreien p-adische analytische gruppe H mit zerfallender halbeinfacher Liealgebra und einer n-dimensionalen abelschen p-adische analytische gruppe Z ist. In Kapitel 3 zeigen wir folgenden Satz: Sei M ein endlich erzeugter Torsionmodul über der Iwasawaalgebra von G, welcher keine nichtrivialen pseudo-null-Untermoduln besitzt. Dann ist q(M), das Bild von M in der Quotientenkategorie Q, genau dann volltreu, wenn M als Modul über der Iwasawaalgebra von Z torsionsfrei ist. Hierbei bezeichne Q den Serre-Quotienten der Kategorie der Moduln über der Iwasawaalgebra von G nach der Serre-Unterkategorie der pseudo-null-Moduln. In Kapitel 4 zeigen wir folgenden Satz: Es bezeichne T die Kategorie, deren Objekte die endlich erzeugten Modulen über der Iwasawaalgebra von G sind, welche auch als Moduln über der Iwasawaalgebra von H endlich erzeugt sind. Seien M, N zwei Objekte von T. Wir nehmen an, dass M, N keine nichttrivialen pseudo-null-Untermoduln besitzen und q(M) in Q volltreu ist. Dann gilt: Ist [M]=[N] in der Grothendieckgruppe von Q, so ist das Bild von N ebenfalls volltreu. In Kapitel 5 zeugen wir folgenden Satz: Sei G eine beliebige p-adische analytische Gruppe, welche keine Element der Ordung p besitzt. Dann sind die Grothendieckgruppen der Algebra stetiger Distributionen und der Algebra beschränkter Distributionen isomorph zu c Kopien des Rings der ganzen Zahlen, wobei c die Anzahl der p-regulären Konjugationsklassen des Quotienten von G nach einer offenen uniformen pro-p-Untergruppe H bezeichnet. / Let G be a compact p-adic analytic group with no element of order p such that it is the direct sum of a torsion free compact p-adic analytic group H whose Lie algebra is split semisimple and an abelian p-adic analytic group Z of dimension n. In chapter 3, we show that if M is a finitely generated torsion module over the Iwasawa algebra of G with no non-zero pseudo-null submodule, then the image q(M) of M via the quotient functor q is completely faithful if and only if M is torsion free over the Iwasawa algebra of Z. Here the quotient functor q is the unique functor from the category of modules over the Iwasawa algebra of G to the quotient category with respect to the Serre subcategory of pseudo-null modules. In chapter 4, we show the following: Let M, N be two finitely generated modules over the Iwasawa algebra of G such that they are objects of the category Q of those finitely generated modules over the Iwasaw algebra of G which are also finitely generated as modules over the Iwasawa algebra of H. Assume that q(M) is completely faithful and [M] =[N] in the Grothendieck group of Q. Then q(N) is also completely faithful. In chapter 6, we show that if G is any compact p-adic analytic group with no element of order p, then the Grothendieck groups of the algebras of continuous distributions and bounded distributions are isomorphic to c copies of the ring of integers where c denotes the number of p-regular conjugacy classes in the quotient group of G with an open normal uniform pro-p subgroup H of G.

[pt] Esta dissertação focaliza a compreensão de reflexivos e pronominais de terceira pessoa por crianças adquirindo o Português Brasileiro (PB), examinando os efeitos das restrições dos princípios A e B da Teoria da Ligação, sob uma perspectiva psicolinguística, que leva em conta a acessibilidade relativa dos possíveis antecedentes dessas formas na memória de trabalho no processamento de sentenças/discurso. Toma-se como referência a teoria da reflexividade de Reinhart e Reuland (1993), por ser compatível com um procedimento de aquisição de formas reflexivas como variáveis ligadas via identificação de predicados reflexivos na interface semântica. A distribuição do reflexivo e de pronominais clíticos/tônicos no PB é considerada. Três experimentos de identificação de imagens com crianças (3-7 anos) e um experimento-piloto com adultos são relatados. Os resultados dos experimentos 1 e 2 sugerem o reconhecimento de se como sinalizador de um predicado reflexivo desde cedo; sua interpretação local em orações encaixadas aos 5 anos; e suscetibilidade à presença de um elemento interveniente no domínio de ligação. Os resultados do experimento 3 (com adultos e crianças) indicam que a marcação de reflexividade fora da grade verbal pela anáfora si, particularmente com o intensificador mesmo(a), e por ele(a) mesmo(a) favorecem a recuperação do sujeito local. Em relação ao tônico ele, o tópico do discurso foi o mais acessível para os adultos, e o sujeito local, para as crianças. Apresenta-se um tratamento teórico que integra a expressão de reflexividade com acessibilidade na interpretação do pronome tônico em posição de objeto no PB. / [en] This dissertation focuses on the comprehension of the 3rd person reflexive and referential pronouns by Brazilian Portuguese-speaking children. The constraints of the Principles A and B of the Biding Theory are considered from a psycholinguistic perspective, which takes into account the relative accessibility of potential antecedents of those forms in the working memory during sentence/discourse processing. The reflexivity theory (Reinhart and Reuland,1993) is a major reference, for being compatible with a language acquisition procedure whereby reflexive forms are represented as bound anaphors via the identification of reflexive predicates at the semantic interface. The distribution of anaphors and clitic/stressed pronominals in BP is considered. Three picture identification experiments with children (3-7 years old) and one with adults are reported. The results of the Experiments 1 and 2 suggest that se is recognized as a reflexivity marker at an early age; that 5 year olds interpret it locally in embedded clauses; and children s susceptibility to the presence of an intervening element in the binding domain. The results of Experiment 3 (with adults and children) suggest that the marking of reflexivity by si (a non-verbal argument anaphor, mainly with the intensifier mesmo), and ele(a) mesmo(a) favor local interpretation. Regarding the stressed pronoun, it was verified the main clause subject is the most accessible antecedent for adults and the local subject is the most accessible antecedent for children. A theoretical account is presented, which integrates reflexivity and the accessibility of the 3rd person object stressed pronoun in in BP.

Les analyses statistiques ont été réalisées avec le logiciels SPSS 11.0(Statistical Package for Social Sciences) et AMOS 6 (Analysis of Moment Structures. La base de données de l'étude a été crée et receuillie par Caroline Despatie en collaboration avec Dr. Dianne Casoni. / Considérant les effets négatifs sur le développement des enfants associées à l'utilisation de châtiments corporels, un ensemble considérable de la recherche a tenté de proposer une avenue pour interrompre le cercle vicieux de la violence intergénérationnelle. Toutefois, on pourrait argumenter qu`une limitation majeure dans la plupart de ces études réside dans le fait que très peu d`entre eux ont eu comme but d`investiguer l'importance d`une troisième variable comme le sentiment de mériter le châtiment, les explications reçues sur l`équité de la peine imposée ou les attitudes qu`une personne peut garder envers cette méthode éducative. En utilisant la modélisation d'équations structurelles, la présente étude a démontré que la relation apparente entre les variables subjectives et sociodémographiques liés à l`expérience de la punition corporelle et l`intention de faire recours à cette méthode éducative en tant que parent est en fait expliquée par le rôle de médiation joué par d'autres variables telles que les attitudes à l'égard de châtiments corporels, le type de châtiments corporels, la façon dont le châtiment a été perçu par le sujet et les raisons données par l'agent pour motiver la peine infligée. La thèse principale avancée par le présent étude est à l`effet qu`une fois le rôle d`une troisième variable est considéré dans l`équation (par exemple, l'attitude envers les châtiments corporels), l'apparente relation entre les variables objectives et les résultats final, soit l`intention de recourir a la punition corporelle, disparaît. Plus précisément, comme il en résulte de l’analyse de modélisation, ce sont les médiateurs qui expliquent toute association possible entre les variables objectives et l`intention d'utiliser les châtiments corporels. C'est-à-dire, un modèle qui ne contiennent aucune liaison directe entre les variables objectives, et directement mésurables (c'est-à-dire l`âgé lors de la punition, le fait d'être un immigrant, la fréquence de la peine, le nombre d'objets utilisés, le nombre d`agents punisseurs, le sexe) et l'intention d'utiliser les châtiments corporels correspond aux données. De surcroît, parmi les quatre médiateurs pris en compte dans notre modèle, ce sont les opinions à l`égard de la punition corporelle qui possèdent la meilleure capacité prédictive en ce qui a trait à l`intention de recourir aux châtiments corporels. Finalement, afin d'étudier le profil des différences entre les sujets en fonction de leurs expériences passées avec les châtiments corporels, les explications reçues pour la peine, et leurs attitudes, une analyse par groupe a été employée. Bref, une solution en quatre groupes faisant des distinctions significatives entre les groupes a été choisie. / Considering the negative developmental findings associated with the use of parental corporal punishment, a considerable body of research has attempted to suggest an avenue for interrupting this vicious cycle of intergenerational violence. However, a major limitation in most of these studies has to do with the fact that very few of them has looked at the influence of a third variable such as the use of explanation, perceived fairness and deservedness of punishment, or the attitudes one has towards corporal punishment. By employing structural equation modelling (SEM), the present study has demonstrated that the apparent relationship between subjective and socio-demographic variables related to corporal punishment and latter intention to use it as a parent is in fact explained by the mediating role of other variables such as attitudes towards corporal punishment, type of corporal punishment, perceived reasons for punishment and reasons given for punishment by the agent. Our main contention was that when a third variable (e.g., attitude towards corporal punishment) is included in the equation, the apparent relationship between punishment variables and outcome disappears. More specifically, as it has resulted from the SEM analysis, the mediating variables explain all of the association between the objective variables and intention to use corporal punishment. That is to say, a model containing no direct paths from the independent variables (i.e., number of ages punished, being an immigrant, frequency of punishment, means of administration of corporal punishment , number of punishing agents, sex) to Intention to Use Corporal Punishment fits the data. Moreover, among the four mediators considered in our model only the attitudinal measures make a significant contribution to the prediction of intention to use corporal punishment. In addition, in order to explore the profile of differences between subjects on their past experiences with corporal punishment, explanations for punishment, and present attitudes, cluster analysis was employed. Finally, a four cluster solution that made meaningful distinctions between groups was selected for consideration.

Considérant les effets négatifs sur le développement des enfants associées à l'utilisation de châtiments corporels, un ensemble considérable de la recherche a tenté de proposer une avenue pour interrompre le cercle vicieux de la violence intergénérationnelle. Toutefois, on pourrait argumenter qu`une limitation majeure dans la plupart de ces études réside dans le fait que très peu d`entre eux ont eu comme but d`investiguer l'importance d`une troisième variable comme le sentiment de mériter le châtiment, les explications reçues sur l`équité de la peine imposée ou les attitudes qu`une personne peut garder envers cette méthode éducative. En utilisant la modélisation d'équations structurelles, la présente étude a démontré que la relation apparente entre les variables subjectives et sociodémographiques liés à l`expérience de la punition corporelle et l`intention de faire recours à cette méthode éducative en tant que parent est en fait expliquée par le rôle de médiation joué par d'autres variables telles que les attitudes à l'égard de châtiments corporels, le type de châtiments corporels, la façon dont le châtiment a été perçu par le sujet et les raisons données par l'agent pour motiver la peine infligée. La thèse principale avancée par le présent étude est à l`effet qu`une fois le rôle d`une troisième variable est considéré dans l`équation (par exemple, l'attitude envers les châtiments corporels), l'apparente relation entre les variables objectives et les résultats final, soit l`intention de recourir a la punition corporelle, disparaît. Plus précisément, comme il en résulte de l’analyse de modélisation, ce sont les médiateurs qui expliquent toute association possible entre les variables objectives et l`intention d'utiliser les châtiments corporels. C'est-à-dire, un modèle qui ne contiennent aucune liaison directe entre les variables objectives, et directement mésurables (c'est-à-dire l`âgé lors de la punition, le fait d'être un immigrant, la fréquence de la peine, le nombre d'objets utilisés, le nombre d`agents punisseurs, le sexe) et l'intention d'utiliser les châtiments corporels correspond aux données. De surcroît, parmi les quatre médiateurs pris en compte dans notre modèle, ce sont les opinions à l`égard de la punition corporelle qui possèdent la meilleure capacité prédictive en ce qui a trait à l`intention de recourir aux châtiments corporels. Finalement, afin d'étudier le profil des différences entre les sujets en fonction de leurs expériences passées avec les châtiments corporels, les explications reçues pour la peine, et leurs attitudes, une analyse par groupe a été employée. Bref, une solution en quatre groupes faisant des distinctions significatives entre les groupes a été choisie. / Considering the negative developmental findings associated with the use of parental corporal punishment, a considerable body of research has attempted to suggest an avenue for interrupting this vicious cycle of intergenerational violence. However, a major limitation in most of these studies has to do with the fact that very few of them has looked at the influence of a third variable such as the use of explanation, perceived fairness and deservedness of punishment, or the attitudes one has towards corporal punishment. By employing structural equation modelling (SEM), the present study has demonstrated that the apparent relationship between subjective and socio-demographic variables related to corporal punishment and latter intention to use it as a parent is in fact explained by the mediating role of other variables such as attitudes towards corporal punishment, type of corporal punishment, perceived reasons for punishment and reasons given for punishment by the agent. Our main contention was that when a third variable (e.g., attitude towards corporal punishment) is included in the equation, the apparent relationship between punishment variables and outcome disappears. More specifically, as it has resulted from the SEM analysis, the mediating variables explain all of the association between the objective variables and intention to use corporal punishment. That is to say, a model containing no direct paths from the independent variables (i.e., number of ages punished, being an immigrant, frequency of punishment, means of administration of corporal punishment , number of punishing agents, sex) to Intention to Use Corporal Punishment fits the data. Moreover, among the four mediators considered in our model only the attitudinal measures make a significant contribution to the prediction of intention to use corporal punishment. In addition, in order to explore the profile of differences between subjects on their past experiences with corporal punishment, explanations for punishment, and present attitudes, cluster analysis was employed. Finally, a four cluster solution that made meaningful distinctions between groups was selected for consideration. / Les analyses statistiques ont été réalisées avec le logiciels SPSS 11.0(Statistical Package for Social Sciences) et AMOS 6 (Analysis of Moment Structures. La base de données de l'étude a été crée et receuillie par Caroline Despatie en collaboration avec Dr. Dianne Casoni.