| Name: | Description: | Size: | Format: | |
|---|---|---|---|---|
| 2.21 MB | Adobe PDF |
Authors
Abstract(s)
Diferentemente do espanhol e do alemão, que possuem duas conjunções coordenativas adversativas distintas (pero e sino; e aber e sondern, respetivamente), o português europeu (PE) possui apenas o morfema mas, veiculando quer o valor de negação de expectativas quer o de correção.
Nestas línguas que têm apenas um mas, como o PE e o francês (veja-se Anscombre & Ducrot, 1977), assume-se classicamente que o valor de correção está principalmente associado a construções do tipo não X mas Y (cf. Horn, 2001). Por outro lado, frases com a estrutura X mas não Y, têm sido frequentemente descuradas quanto à definição do seu valor. No entanto, Vicente (2010) e Steindl (2017) assumem que em espanhol e alemão é possível nesta construção ocorrerem as conjunções adversativas pero e aber, típicas de contextos de negação de expectativas, com partículas de polaridade no segundo termo coordenado, veiculando o valor de correção.
Apresentámos evidência empírica de que as frases adversativas devem ser analisadas da seguinte forma: (i) as frases com negação frásica (no primeiro termo) devem ser analisadas como coordenação de dois constituintes subfrásicos, sendo que apenas o primeiro está sob o escopo da negação, por questões de localidade e também pelo facto de mas bloquear a negação para o segundo termo; (ii) quanto aos casos com negação de constituintes, a análise adotada será de coordenação de sintagmas (sem elipse) mesmo quando são CPs; (iii) por fim, para frases com negação frásica no segundo termo coordenado, seguindo Saab (2008) para espanhol e Matos (2017) para o PE, assumimos que estas construções são casos de Elipse de TP com itens de polaridade, como também foi assumido por Vicente (2010 ) para frases com pero + itens de polaridade, que, na sua investigação, correspondem a coordenação frásica + elipse.
Quanto aos valores da conjunção adversativa, nesta investigação, de forma a complementar a análise teórica, construímos duas tarefas de Compatibilidade Contexto-Frase, em que os participantes eram expostos a uma frase-alvo e dois contextos: um com valor de correção e outro de negação de expectativas. Na tarefa, foi-lhes solicitado que avaliassem qual o contexto mais compatível com a frase-alvo ou, alternativamente, se ambos eram.
A primeira tarefa incluiu frases com negação frásica e de constituintes, no primeiro termo coordenado. Os dados obtidos sugerem uma preferência pela leitura cumulativa de correção e negação de expectativas, e não apenas de correção. Os dados também revelaram que, diferentemente do que é referido na literatura, a negação de constituintes não facilita a interpretação corretiva.
Na segunda tarefa, as frases-alvo foram manipuladas e a negação (frásica ou de constituintes) surgia no segundo termo coordenado, correspondendo, assim, a frases com Elipse de TP e Contraste Sintagmático. Os resultados indiciam uma preferência pela leitura de negação de expectativas, em ambas as construções. Porém, a leitura corretiva não foi rejeitada, provando, assim, que é possível ter um elemento negativo no segundo termo coordenado e obter uma interpretação de correção.
Estes dados analisados nesta investigação mostram que a ligação assumida por alguns autores entre os valores da conjunção e a configuração sintática não é robusta. Vicente (2010), por exemplo, afirma que frases corretivas correspondem a coordenação frásica + elipse (cf. Clause-Only Coordination Hypothesis), diferentemente de construções com valor de negação de expectativas que podem ser analisadas como coordenação subfrásica ou frásica. Quanto às frases corretivas, Toosarvandani (2013) e Steindl (2017) consideram que é possível aceitar uma análise de coordenação frásica + elipse, embora os argumentos não sejam suficientemente robustos para excluir a hipótese de coordenação subfrásica.
Os nossos resultados mostram que não há uma ligação direta entre estrutura sintática e valor semântico (num sentido amplo e estrito). Os participantes do nosso estudo revelaram uma preferência por uma leitura cumulativa ou de negação de expectativas, e estas leituras não se correlacionam com a sintaxe.
Nesta tese também exploramos construções incoordenadas, que se caracterizam pela presença de mas e de uma frase, que ocorre na posição do segundo termo, e que possui o valor de negação de expectativas (cf. Kuteva et al, 2017). Retomamos o trabalho de Kuteva et al e propomos uma análise para frases incoordenadas e para o estatuto do conector coordenativo que nelas ocorre.
Propomos a existência de dois tipos de frases incoordenadas, dependendo de como é legitimada a sua ocorrência. Distinguimos entre verdadeiras incoordenadas e pseudo-incoordenadas: as primeiras são legitimadas pela sua conexão com um contexto não-linguístico; as últimas têm um primeiro termo omitido, que pode ser recuperado pelo conteúdo linguístico do fragmento discursivo anterior.
Quanto ao estatuto de mas nestas construções, levantamos a hipótese de que mantém a sua natureza conjuncional, e que possui o traço [± coordenativo]. Este traço será valorado como [+ coordenativo] em frases coordenadas ou pseudo-incoordenadas, por ter, nestas construções, uma função coordenativa; e o valor [- coordenativo] será associado às verdadeiras incoordenadas, revelando sua natureza não coordenativa. Também propomos que mas incoordenado não verifica todas as propriedades associadas às conjunções, por, na verdade, algumas serem propriedades da coordenação.
No geral, este trabalho lança luz sobre a interação entre sintaxe, discurso e valores semânticos/pragmáticos associados a construções adversativas. Além disso, acreditamos que nossa investigação sobre a incoordenação mostra que este processo estende a sintaxe às estruturas discursivas.
Differently from Spanish and German, which have two distinct adversative coordinating conjunctions (pero and sino; and aber and sondern, respectively), European Portuguese (EP) only has the mas morpheme conveying both the counterexpectational and corrective meanings. In languages that only have one adversative morpheme, as EP and French (see Anscombre & Ducrot, 1977), it is classically assumed that the value of correction is mainly associated to the construction type not X but Y (cf. Horn 2001). On the other hand, sentences with a X but not Y structure, have often been neglected regarding the definition of their value. However, Vicente (2010) and Steindl (2017) assume that in Spanish and German it is possible to have the X but not Y construction with the adversative conjunctions pero and aber, typical of counterexpectational contexts, occurring with polarity particles in the second conjunct to convey correction. We presented empirical evidence that adversative sentences should be analyzed as follows: (i) for sentences with sentential negation (in the first conjunct) we assume the coordination of two subclausal constituents, being that only the first one is under the scope of negation due to locality and also the fact that mas blocks the scope of negation to the second conjunct; (ii) as for cases with constituent negation, the analysis adopted will be of phrasal coordination (without ellipsis) even when the phrases are CPs; (iii) finally, for sentences with sentential negation in the second conjunct, following Saab (2008) for Spanish and Matos (2017) for EP, we assume these constructions to be cases of TP ellipsis with polarity items, as it was also assumed by Vicente (2010) for sentences with pero + polarity items, which, in his investigation, correspond to clausal coordination + ellipsis. In this investigation, we built two Context-Sentence Compatibility tasks, in which the participants were presented with a target sentence and contexts associated to corrective and counterexpectational values. Participants were asked to assess the which context is more appropriate or, alternatively, if both are. The first task dealt with sentences with sentential and constituent negation occurring in the first conjunct. The data obtained showed a preference for a cumulative reading of correction and counterexpectation, and not only correction. The data also revealed that, differently from what is proposed in the literature, constituent negation does not facilitate a corrective interpretation. In the second task, the target sentences were manipulated and (sentential or constituent) negation occurred in the second conjunct, yielding TP Ellipsis and Phrasal Contrast sentences. The results showed a preference for the counterexpectational reading, in both constructions. However, the corrective reading was not rejected thus proving that it is possible to have a negative element in the second conjunct and obtain correction. The data we analyzed in this investigation allowed us to determine that the connection assumed by some authors between the values of the adversative conjunction and the syntactic configuration is not robust. Vicente (2010), for example, claims that corrective sentences correspond to full clause coordination + ellipsis (cf. Clause-Only Coordination Hypothesis), differently from counterexpectational constructions that can be analyzed as subclausal or clausal coordination. As for corrective sentences, Toosarvandani (2013) and Steindl (2017) consider that it is possible to accept clausal coordination + ellipsis, but there are no arguments robust enough to exclude subclausal constituent coordination. Our results showed that there is no direct connection between syntactic structure and semantic value (in a broad or strict sense). The participants in our study revealed a preference for either a cumulative reading or a counterexpectational one, and these readings do not correlate to syntax. In this investigation we also explored incoordinate constructions, which are characterized by the presence of but and a sentence, occurring in the position of the second conjunct, and bear a counterexpectational value. We took up on the work of Kuteva et al, proposed an analysis for incoordinate sentences and for the status of the coordinating connective that occurs in them. We proposed the existence of two different types of incoordinate sentences, depending on how its occurrence is legitimated. We made a distinction between true incoordinates and pseudo-incoordinates: the former is legitimated by its connection to a previous non-linguistic situation; the latter has an omitted first conjunct, which can be recovered by the linguistic content of the previous discourse. Regarding the status of but in these constructions, we raise the hypothesis that it maintains its conjunctional nature, and that it bears a [± coordinating] feature. This feature will be valued as [+ coordinating] in coordinate or pseudo-incoordinate sentences as it has, indeed, a coordinating function; and the value [- coordinating] will be associated to true incoordinate, revealing its non-coordinating nature. We also proposed that incoordinate but does not verify all the properties associated to coordinating conjunctions because some of these properties should be seen as characteristics of coordination. Overall, this work sheds light on the interplay between syntax, discourse, and semantic / pragmatic values associated to adversative constructions. Moreover, we believe our investigation on incoordination has the purpose of showing that this process extends syntax to discoursive structures.
Differently from Spanish and German, which have two distinct adversative coordinating conjunctions (pero and sino; and aber and sondern, respectively), European Portuguese (EP) only has the mas morpheme conveying both the counterexpectational and corrective meanings. In languages that only have one adversative morpheme, as EP and French (see Anscombre & Ducrot, 1977), it is classically assumed that the value of correction is mainly associated to the construction type not X but Y (cf. Horn 2001). On the other hand, sentences with a X but not Y structure, have often been neglected regarding the definition of their value. However, Vicente (2010) and Steindl (2017) assume that in Spanish and German it is possible to have the X but not Y construction with the adversative conjunctions pero and aber, typical of counterexpectational contexts, occurring with polarity particles in the second conjunct to convey correction. We presented empirical evidence that adversative sentences should be analyzed as follows: (i) for sentences with sentential negation (in the first conjunct) we assume the coordination of two subclausal constituents, being that only the first one is under the scope of negation due to locality and also the fact that mas blocks the scope of negation to the second conjunct; (ii) as for cases with constituent negation, the analysis adopted will be of phrasal coordination (without ellipsis) even when the phrases are CPs; (iii) finally, for sentences with sentential negation in the second conjunct, following Saab (2008) for Spanish and Matos (2017) for EP, we assume these constructions to be cases of TP ellipsis with polarity items, as it was also assumed by Vicente (2010) for sentences with pero + polarity items, which, in his investigation, correspond to clausal coordination + ellipsis. In this investigation, we built two Context-Sentence Compatibility tasks, in which the participants were presented with a target sentence and contexts associated to corrective and counterexpectational values. Participants were asked to assess the which context is more appropriate or, alternatively, if both are. The first task dealt with sentences with sentential and constituent negation occurring in the first conjunct. The data obtained showed a preference for a cumulative reading of correction and counterexpectation, and not only correction. The data also revealed that, differently from what is proposed in the literature, constituent negation does not facilitate a corrective interpretation. In the second task, the target sentences were manipulated and (sentential or constituent) negation occurred in the second conjunct, yielding TP Ellipsis and Phrasal Contrast sentences. The results showed a preference for the counterexpectational reading, in both constructions. However, the corrective reading was not rejected thus proving that it is possible to have a negative element in the second conjunct and obtain correction. The data we analyzed in this investigation allowed us to determine that the connection assumed by some authors between the values of the adversative conjunction and the syntactic configuration is not robust. Vicente (2010), for example, claims that corrective sentences correspond to full clause coordination + ellipsis (cf. Clause-Only Coordination Hypothesis), differently from counterexpectational constructions that can be analyzed as subclausal or clausal coordination. As for corrective sentences, Toosarvandani (2013) and Steindl (2017) consider that it is possible to accept clausal coordination + ellipsis, but there are no arguments robust enough to exclude subclausal constituent coordination. Our results showed that there is no direct connection between syntactic structure and semantic value (in a broad or strict sense). The participants in our study revealed a preference for either a cumulative reading or a counterexpectational one, and these readings do not correlate to syntax. In this investigation we also explored incoordinate constructions, which are characterized by the presence of but and a sentence, occurring in the position of the second conjunct, and bear a counterexpectational value. We took up on the work of Kuteva et al, proposed an analysis for incoordinate sentences and for the status of the coordinating connective that occurs in them. We proposed the existence of two different types of incoordinate sentences, depending on how its occurrence is legitimated. We made a distinction between true incoordinates and pseudo-incoordinates: the former is legitimated by its connection to a previous non-linguistic situation; the latter has an omitted first conjunct, which can be recovered by the linguistic content of the previous discourse. Regarding the status of but in these constructions, we raise the hypothesis that it maintains its conjunctional nature, and that it bears a [± coordinating] feature. This feature will be valued as [+ coordinating] in coordinate or pseudo-incoordinate sentences as it has, indeed, a coordinating function; and the value [- coordinating] will be associated to true incoordinate, revealing its non-coordinating nature. We also proposed that incoordinate but does not verify all the properties associated to coordinating conjunctions because some of these properties should be seen as characteristics of coordination. Overall, this work sheds light on the interplay between syntax, discourse, and semantic / pragmatic values associated to adversative constructions. Moreover, we believe our investigation on incoordination has the purpose of showing that this process extends syntax to discoursive structures.
Description
Keywords
Coordenação (Linguística) Proposições adversativas Conjunções (Linguística) Marcadores discursivos Língua portuguesa - Sintaxe Teses de doutoramento - 2023