This dissertation argues for the existence of functional attitude predicates, light attitudes, such as light say and light see. Two phenomena are identified as functional attitudes: evidentiality and logophoric say-complementizers. I propose that reportative evidential markers and logophoric licensing complementizers are cross-linguistic variations of overt morphosyntactic realizations of the same light attitude: a functionalized predicate say. The parallel between evidentiality and logophoricity drawn here highlights their properties that have not been discussed or formally accounted for until now, and explains why they are in a typological complementary distribution across the world's languages. At the same time, direct and reportative evidentials even within the same language exhibit a number of syntactic and semantic differences that have not been noticed in the literature before. I derive them from the analysis of reportative and direct evidentials as different kinds of functional predicates: say and perceive, respectively. After establishing the nature of evidentials, I develop their syntactic and semantic properties formally. I claim that light attitudes are hosted by a projection cP, which selects CP and has properties similar to that of the light verb projection vP, such as argument structure, thematic roles, and `flavors'. The semantic composition of light attitudes is based on that of lexical attitudes, for which I am following and expanding ideas from de-compositional semantics. This allows for a simple and conceptually motivated analysis that does not need any additional theoretical primitives. I develop novel methodology to test for evidential challengeability and newness of evidentials that take the between-evidential differences into account. The results support the representation of the evidential contribution as a presupposition.

This work investigates a series of phenomena that shed light on the analysis of attitudes de se. We adopt Lewis’ (1979) proposal that attitudes de se involve self-ascription of a property, and investigate how this view of mental content is reflected in natural language. The implementation favored is a strong version of Lewis’ position: root and embedded clauses are uniformly treated as being of property type. Our approach elaborates Chierchia’s (1990) view that de se construals arise via binding by an abstraction operator in the clausal left periphery. Part I develops an argument that such operators occur in root as well as embedded clauses. This is contrasted with the view that the evaluation index incorporates an individual parameter, a prominent version of which treats the behavior of predicates of taste such as tasty as evidence that truth is relativized to individuals (Lasersohn, 2005; Stephenson, 2007a, 2007b). Chapter 2 argues against this view, defending a semantics for taste predicates that requires no appeal to an individual parameter. Chapter 3 employs an argument from Moore’s Paradox to motivate the proposal that root clauses bear individual abstractors in their left periphery, while Chapter 4 identifies phenomena that the system accounts for. Part II concerns two elements whose distribution is confined to embedded clauses: controlled PRO and the logophoric pronoun in the Niger-Congo language Ewe. Chapters 5 and 6 investigate the semantics of partial control, a variety of control where the controller denotes a proper subset of the understood subject. The view that control complements express properties lends itself to a principled account of which predicates license partial control. Chapter 7 presents novel data regarding the logophoric pronoun in Ewe. We show that, contrary to what had been assumed in the absence of the necessary fieldwork, Ewe logopohors are not obligatorily de se. We propose an account of this finding that is compatible with the implementation of the property view that we favor. Chapter 8 closes the dissertation by considering why it should be that certain expressions, such as PRO, are obligatorily de se while others, like the Ewe logophor, can be de re. / Linguistics

Submitted by Morgana Silva (morgana_linhares@yahoo.com.br) on 2016-10-11T17:46:11Z No. of bitstreams: 1 arquivototal.pdf: 1246064 bytes, checksum: 48d673716d069f78f8353ea6c93230d5 (MD5) / Made available in DSpace on 2016-10-11T17:46:11Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 1246064 bytes, checksum: 48d673716d069f78f8353ea6c93230d5 (MD5) Previous issue date: 2016-02-19 / Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - CAPES / This study aims to investigate whether logophoric structures are acceptable or not for the speakers of Brazilian Portuguese. Much of the issue literature on the subject (Reinhart and Reuland (1993), Menuzzi (1997)) points out that logophoric anaphora exists separately from syntactic anaphora in Universal Grammar. The syntactic anaphoras are guided by syntactic factors such as location and c-command. But logophor may or may not notice these syntactic conditions. These studies also demonstrate that where there is logophoricity, the pronoun and anaphora are interchangeable, so, either one or the other can be used in the same position in the sentence. The conclusion we reached after the literature review is that anaphora resolution in logophorics contexts seems to involve not only the influence of structural information but also and above all, the influence of pragmatic and discursive information, as seen in pronouns. Based on these theoretical assumptions, we examine the acceptability of logophorics sentences in PB in order to know if they are present in the grammar of the speakers of that language. Two acceptability judgment experiments were conducted in order to know if the structures with the logophoric anaphor “ele mesmo” and the logophoric pronoun “ele” are acceptable in BP, based on the English structures presented by Reinhart e Reulnad (1993). Our hypothesis was that both structures are acceptable to speakers of BP and that in contexts where there is a logophoric anaphor, there can also be a logophoric pronoun, as shown by Reinhart e Reuland (1993). The results showed an acceptance of both structures in BP, supporting the formulated hypothesis. With these results, we obtained evidence that the structures with the logophoric anaphora “ele mesmo” and the logophoric pronoun “ele” observed in this study are acceptable in Brazilian Portuguese, as in English. / Este trabalho busca investigar se as estruturas logofóricas são aceitáveis ou não para os falantes do português brasileiro (PB). Grande parte da literatura sobre o assunto (Reinhart e Reuland (1993), Menuzzi (1997), dentre outros) aponta que a anáfora logofórica existe separadamente da anáfora sintática na Gramática Universal. As anáforas sintáticas são guiadas por fatores sintáticos como localidade e c-comando. Já a logofórica pode ou não observar essas condições sintáticas. Esses estudos também demonstram que nos casos em que há logoforicidade, a anáfora e o pronome são intercambiáveis, ou seja, tanto um quanto o outro podem ser usados na mesma posição na frase. A conclusão a que chegamos após a revisão de literatura é que a resolução de anáforas em contextos logofóricos parece envolver não apenas a influência de informações estruturais, mas também, e sobretudo, a influência de informações pragmáticas e discursivas, tal como se observa nos pronomes. Apontamos que a Teoria da Reflexividade parece ser mais coerente para a explicação desse fenômeno. Com base nesses pressupostos teóricos, examinamos a aceitabilidade de construções logofóricas em PB com a finalidade de saber se elas estão presentes na gramática dos falantes dessa língua. Foram realizados dois experimentos de julgamento de aceitabilidade no intuito de saber se as estruturas com a anáfora logofórica ‘ele mesmo’ e com o pronome logofórico ‘ele’ são aceitáveis em PB, tomando como base as estruturas em inglês apresentadas por Reinhart e Reuland (1993). A nossa hipótese foi de que ambas as estruturas são aceitáveis para os falantes do PB e que nos contextos em que há uma anáfora logofórica também pode haver um pronome logofórico, conforme aponta Reinhart e Reuland (1993). Os resultados encontrados evidenciaram uma aceitação de ambas as estruturas em PB, corroborando a hipótese formulada. Com estes resultados, obtivemos evidências de que as estruturas com a anáfora logofórica ‘ele mesmo’ e com o pronome logofórico ‘ele’ observadas nesse estudo são aceitas em PB, como em inglês.

This thesis is an investigation of the structural and formal semantic properties of copy raising constructions in English, as well as their expletive counterparts. The first main claim is that contrary to what has been previously assumed, the perceiver of the event (i.e. the Pgoal in Asudeh & Toivonen's 2012 terms) is an obligatory syntactic and semantic argument of the matrix verb. I argue that the identification of the Pgoal is not left to pragmatics, but rather that is represented as a silent pronoun in the structure: one that picks up a logophoric antecedent. The result of this is that the material in the embedded clause is semantically interpreted with respect to the Pgoal's perspective. The second major claim of the thesis is that this perspective-sensitivity is most appropriately captured using a modal semantic framework (Kratzer, 1977, 1981 von Fintel & Heim, 2002). Specifically, I argue that each of the different copy raising verbs encodes a different accessibility relation between possible worlds or situations, while the Pgoal's information state provides the relevant domain of worlds. Using these insights, I propose truth conditions for these constructions, which ultimately are sensitive to a kind of stereotypical ordering, and account for inter-speaker variability. Finally, I discuss the anomalous class of copy raising constructions with non-thematic subjects, and argue that overlapping discourse functions may have resulted in a shift away from modal semantics in these cases. / Thesis / Master of Science (MSc) / This thesis investigates the linguistic meaning associated with the "copy raising" sentence construction, e.g. "Your cat looks like she wants to go outside." I argue that the interpretation of these sentences is dependent on establishing the individual whose perspective is conveyed in the sentence, which does not need to be the speaker. After examining the range of contexts in which various different copy raising constructions can be used, I propose an analysis of their core meaning that draws on the philosophical idea of possible worlds, and the psychological notion of stereotypicality. I also address the question of whether these constructions are related to the phenomenon of evidentiality, a property of certain languages which allows the speaker to linguistically mark the source of evidence for their claim.