Repeated names, overt pronouns, and null pronouns in Spanish

Abstract

In two self-paced, sentence-by-sentence reading experiments we examined the difference in the processing of Spanish discourses with repeated names, overt pronouns, and null pronouns in emphatic and non-emphatic contexts. In Experiment 1, repeated names and overt pronouns caused a processing delay when they referred to salient antecedents in non-emphatic contexts. In Experiment 2, both processing delays were eliminated when an emphatic cleft-structure was used. The processing delay caused by overt pronouns referring to salient antecedents in non-emphatic contexts in Spanish contrasts with previous findings in Chinese, where null and overt pronouns elicited similar reading times. We explain both our Spanish findings and the Chinese findings in a unified framework based on the notion of balance between processing cost and discourse function in line with the Informational Load Hypothesis.Coherent discourse often includes repeated reference to a small number of concepts. Such repetition can be achieved by means of anaphors, that is, referential expressions that co-refer with an antecedent found earlier in the same discourse (Ducrot & Todorov, 2003). Anaphors can take many forms, including pronouns (e.g., he, she), definite expressions (e.g., the boy, those girls), and repeated names (e.g., Mary). Coreferentiality relations between anaphors and their antecedents serve an important role in establishing and maintaining coherence in discourse. The fact that the same expression, for example a proper name, can be used anaphorically or non-anaphorically precludes a taxonomic account of anaphoric devices that is based strictly on the form of the referential expressions. Indeed, the literature suggests that a complete explanation of anaphoric processing and use should consider, in addition to the form of the anaphoric expression, other factors including the syntactic configuration in which the anaphor appears (Chambers & Smith, 1998;Crawley, Stevenson & Kleinman, 1990;Frederiksen, 1981), the discourse pragmatics (Almor, 1999;Ariel, 1990;Prince 1978,) and the related memory processes (Almor, 1999;Gernsbacher, 1989;Sanford & Garrod, 1981).One theory that focuses on the role of memory in anaphoric processing is Gernsbacher's Structure Building framework (1989), which views anaphoric expressions as memory retrieval cues. The theory predicts that repeated names should be processed faster than pronouns because they provide more information that can more effectively activate appropriate memory representations. For example, once we have mentioned Evita, it may seem more efficient to refer to her again as Evita rather than she because the interpretation of the anaphoric pronoun she relies on a small number of properties such as gender, number and animacy that only provide a weak retrieval cue. By the same explanation, using non repetitive expressions such as the queen of the poor or the thief, which add new information by revealing the speaker's attitude toward this political figure, should be less effective as

Language processing, acceptability, and statistical distribution: A study of null and overt subjects in Brazilian Portuguese

“…This ''repeatedname penalty" (RNP) has also been extended to Mandarin Chinese (Yang, Gordon, Hendrick, & Wu, 1999) and Spanish (Gelormini-Lezama & Almor, 2011, suggesting that this might be a universal phenomenon.…”

“…Despite this difference in initial processing cost, more informative referential forms can be processed more easily overall (as reflected in whole sentence reading times) when their additional cost is justified by some discourse function that could not have been achieved by a less informative expression. Gelormini-Lezama and Almor (2011) applied this explanation to the OPP in Spanish and argued that the OPP reflects the use of an overly informative expression (the overt pronoun) when it serves no discourse function.…”

“…The OPP was first reported in Spanish (Gelormini-Lezama & Almor, 2011) in which overt pronouns can serve the function of either retrieving a non-salient antecedent, or marking a contrast or emphasis. In non-contrastive contexts, an overt pronoun referring back to a salient antecedent represents a case where the richer semantic content of the overt pronoun in comparison to a null pronoun is unnecessary.…”

Referenciação e técnicas experimentais: aspectos metodológicos na investigação do processamento correferencial em português brasileiro

“…Refeita essa importante consideração quanto à necessidade de um rigoroso controle dos estímulos utilizados em experimentos psicolinguísticos, é também digno de nota que eles, em especial os realizados no paradigma de leitura autocadenciada, têm sido amplamente empregados nos estudos sobre referenciação, tanto em português brasileiro (e.g., CORRÊA, 1998; CUNHA LIMA In: KOCH, MORATO, BENTES, 2005;LEITÃO, 2005), quanto em outras línguas, como o inglês (e.g., GORDON et al, 1993;ALMOR, 1999), o chinês (e.g., YANG et al, 1999) e o espanhol (e.g., GELORMINI-LEZAMA, 2008). A realização de experimentos nesse campo se vincula à tentativa de elucidação dos aspectos psicológicos relacionados à questão da referência, isto é, à investigação de como as pessoas identificam o referente de uma expressão nominal/pronominal qualquer e de como esse referente se mantém ou se desativa na memória dos falantes.…”

“…E os objetivos específicos foram (i) verificar a existência ou não de RNP e OPP em PB e (ii) oferecer uma análise do fenômeno da correferência em PB que pudesse ser diretamente comparada aos trabalhos anteriormente mencionados sobre o inglês, o chinês e o espanhol. 6 De modo a cumprir com os objetivos apresentados, replicamos, em colaboração com o autor, o primeiro experimento de leitura autocadenciada descrito em sua dissertação (GELORMINI-LEZAMA, 2008). Tal replicação se deu mediante tradução de todos os itens experimentais (do espanhol para o PB) empregados no seu teste de leitura autocadenciada, dos quais um exemplo pode ser conferido no quadro 1.…”