Representação espacial de papéis temáticos: natureza ou cultura?

Abstract

This thesis investigates the interdependence of the cognitive modules of language and space through two experiments that investigate the relationship between linguistic and spatial notions. We delimited spatial notions on the Cartesian x and y axes in order to better understand the effects of literacy on human spatial cognition. In our first experiment (sentence-image matching task), we analyzed how the spatial representation of the thematic roles of agent and patient occurs in sentences that present action verbs with left-to-right and right-to-left direction. We investigated how the spatialization of thematic roles occurs in individuals who do not have written linguistic competence, in order to verify how the bias of languages that follow the left-to-right order - such as Portuguese - can influence spatial representation tasks. Based on the work carried out by Chatterjee et al. (1999) and Maass and Russo (2003), who did not reach conclusive results on the origin of a specific biological or cultural bias, we verified how this spatial representation occurs in illiterate participants. Considering that understanding the existence and causes of a specific behavior in tasks of this nature has not yet been clarified for the group of illiterate individuals, the aim of this thesis is to identify the factor that influences the spatial bias in literate people, by studying the way of spatializing of those who have not been literate (illiterate). To verify the factors that determine this behavior, we proposed a sentence-image matching task using the eye tracker, in a comparative study between a group of 30 literate participants (control) and 10 illiterate participants (experimental), to assess whether it is possible to dissociate the spatial representation of action verbs in the active voice from the left-right direction of the individuals' literacy. Our data provides evidence in favor of the Spatial Primitives Hypothesis (of a biological nature). In our second experiment (conceptualization task using prime), we observed the semantic and spatial relationship between prime words and target pictures in the spatial congruency effect. Our aim was to investigate whether the components prime type and congruency, i.e. the correspondence between prime word and target image, influence the processing of words that have implicit spatial semas. Based on the study carried out by Ostarek et al. (2018), we adapted the conceptualization experiment with the use of prime in order to compare the effect of online processing of words with implicit spatial semas in saccadic trajectory and to compare the performance of literate and illiterate people. Since there are no studies on the saccadic performance of people who can neither read nor write, in experiments of this nature, we consider it relevant to analyze the influence of literacy on the Cartesian y-axis. We believe that the findings could give us insights into human spatial cognition. To achieve our goal, we tested the same participants as in experiment 1 and the same experimental eye-tracking technique. The data revealed a significant result only for the dependent variable saccade speed. For this variable, we found that the type of prime and the type of figure influence saccade speed. Our data provides evidence that could help confirm hypotheses four and five.