On the relation between consciousness presentation and exempt reflexives

The literary practice of reflexivising the pronoun designating fictional consciousness has long been noted in stylistics. Despite their varying details, existing studies generally agree upon one pattern: the more immediate the presentation of fictional consciousness, the higher the likelihood of observing a reflexive [1,2]. Although this phenomenon partly overlaps the so-called exempt reflexives discussed in psycholinguistics (reflexives which are not in complementary distribution with their pronominal counterpart) [3,4], to date, it has yet to be properly explored and represents a case in which manner of narration might interact with perspective taking in creating possible contexts for the licensing of exempt reflexives. To test whether readers’ expectations about the appropriateness of reflexivisation are sensitive to the degree of immediacy involved in the presentation of a given fictional consciousness, we created two types of narratives in Experiment 1 (see Table 1) and manipulated the presence vs. absence of typical literary devices that signal immediacy of consciousness presentation (e.g., exclamations; [5]). Both texts can be read as Percy’s inner turmoil. However, in (a), the employment of the highlighted literary devices is expected to obscure the presence of the narrator and create for readers an illusion of being granted direct access to Percy’s consciousness. In contrast, in (b), the absence of such devices is supposed to increase the level of narrator involvement and induce readers to perceive the narrative as a summary/analysis of Percy’s inner turmoil. Two options were provided in the critical sentence (himself/him); participants were asked to click on the word that they felt were more natural/suitable within the context. Based on the generalisation proposed in prior literary studies, we hypothesised that the suitability of the exempt reflexive should improve when the context has guaranteed a reading of immediate consciousness presentation, compared to when the presence of the narrator can be distinctly felt throughout the rendering of the same event. The results confirmed our hypothesis: there was an increase in readers’ disposition to consider the exempt reflexive to be a suitable alternative for the textual manifestation of the character in (a), when compared to (b) (z = 2.56, p = .011; see Fig.1). Experiment 2 examined whether our manipulation had indeed elicited a difference in the level of immediacy as perceived by readers. The same materials were adopted, with the critical sentence removed from both conditions (see Table 2). Participants were asked to indicate the extent to which they felt they were inside the character’s/narrator's mind when reading the narrative. The results corroborated our hypothesis by showing that participants tended to feel that they were deeper inside the character’s mind in the subjective condition, when compared to the omniscient condition (z = 10.05, p < .001; see Fig.2). Our findings demonstrate readers’ sensitivity towards the relation between the use of exempt reflexives and the manner in which a given fictional consciousness is presented.