No description available.

The syntactic and semantic treatment of Chinese Bare Conditionals is a topic of much debate (Cheng and Huang 1996; Lin 1996; Chierchia 2000). This dissertation investigates the nature of Chinese Bare Conditionals in three aspects: quantification and modal implications as compared to English free relatives with –ever, and pronoun occurrence. With regard to quantification, I propose to treat the anteceding wh-phrase and its anaphoric element (pronoun/wh-word) uniformly as a definite description denoting a maximal plural entity similar to Jacobson (1995). This entity can be an atomic entity resulting in a singular definite reading, or an entity consisting of more than one atom deriving a universal-like reading. Concerning modal implication, I propose to capture the agent’s/speaker’s indifference reading of bare conditionals with von Fintel (2000). Indifference reading in his analysis is interpreted against a counterfactual modal base which predicts a causal link. His analysis is needed for the interpretation of Chinese bare conditionals but may not be applied directly to whatever, given that a causal link is necessarily present in a bare conditional, but not required in an English whatever-sentence. I argue that the use of a pronoun in a bare conditional is not subject to a uniqueness and existence condition as claimed in Lin (1996). Although bare conditionals typically contain two identical wh-words, they may occur naturally with a pronoun that links bare conditionals with other sentences into a piece of coherent discourse. This account bears an important implication for the study of Chinese wh-phrases and third person pronouns in being able to predict the existence of anaphoric definite wh-phrases and bound-variable pronouns in the language. It also improves on existing accounts of Chinese bare conditionals in being able to capture the details of the form and meaning of this construction. Chinese bare conditionals are structurally related to ruguo ‘if’-conditionals and Hindi left-adjoined correlatives and their meaning is similar to, and yet not quite the same as that of whatever. / text

Orientadores: Jairo Morais Nunes, Mary Aizawa Kato / Tese (doutorado) - Universidade Estadual de Campinas, Instituto de Estudos da Linguagem / Made available in DSpace on 2018-11-09T16:30:17Z (GMT). No. of bitstreams: 1 LessadeOliveira_AdrianaStellaCardoso_D.pdf: 1270774 bytes, checksum: e4fffae5f8011e501b903a3b031139ce (MD5) Previous issue date: 2008 / Resumo: Este estudo investiga a aquisição das estratégias relativas em português brasileiro (PB). Os corpora investigados são constituídos de dados de fala naturalísticos-longitudinais de três crianças, dos 1;6 aos 3;6 anos de idade, bem como de dados da fala de seus interlocutores adultos. Assumindo a proposta de Kato (1993) e Kato e Nunes (2007), segundo a qual estruturas de deslocamento à esquerda estão na base da derivação de relativas não-padrão (cortadora) em PB, registro duas generalizações que não foram notificadas na literatura sobre sentenças relativas em PB: a) relativas apositivas são adquiridas cedo mas não são permitidas como estratégia não-padrão; e b) relativas livres também não tardam a aparecer e admitem a estratégia não-padrão. Quanto às relativas restritivas, mostro que, enquanto as relativas não-preposicionais e preposicionais não-padrão aparecem cedo na fala infantil, a aquisição das relativas com pied-piping requer ensino formal. Assumindo a proposta de Hornstein (2007), de acordo com quem derivações que recorrem a movimento (compreendido em termos de cópia) são mais econômicas que aquelas que recorrem a pronome resumptivo, proponho que a estratégia de relativização padrão é adquirida antes da estratégia não-padrão, no caso das relativas não preposicionais, porque estão a opção mais econômica. Com base na noção de Economia de Roeper (2003) em termos de profundidade mínima (computada por número de nódulos da sonda ao alvo), argumento que a dificuldade na aquisição das relativas com pied-piping se deve a uma complexidade inerente operação de pied-piping / Abstract: This study investigates the acquisition of relativization strategies in Brazilian Portuguese (BP). The corpora investigated are constituted by naturalistic longitudinal data from three children, from 1;6 to 3;6 years old, as well as data from their caretakers. Assuming Kato¿s (1993) and Kato and Nunes¿s (2007) proposal that left dislocation structures underlie the derivation of non-standard (¿PP-chopping¿) relatives in BP, I document two generalizations which have not been noticed in the literature on BP relative clauses: a) appositive clauses do not allow the non-standard strategy, but are acquired early; and b) free relatives admit the nonstandard strategy, and are also acquired early . As for restrictive relative clauses, the data show that the non-prepositional and the non-standard prepositional relatives appear early, whereas pied-piping relatives require formal teaching. Assuming Hornstein¿s (2007) proposal that derivations that resort to movement (understood in terms of copying) are more economical than those that resort to resumption, I propose that the standard relativization strategy is acquired before the non-standard strategy in the case of the non-prepositional relatives because it relies on the more economical option. Moreover, using Roeper¿s (2003) Economy notion in terms of minimal depth (number of nodes from the probe to the goal), I argue that the difficulty in the acquisition of pied-piping relatives is due to the inherent complexity of the pied-piping operation / Doutorado / Linguistica / Doutor em Linguística