Frame Semantics as a Framework to account for the Foreign Language Effect

Abstract

Linguistic meaning can be expressed in multiple languages. One may assume that equivalent texts/ utterances in two languages by means of translation generate equivalent meanings in their readers/ hearers. This follows if we assume that meaning calculation is solely objective in nature. However, research in language and cognition is building up to show otherwise. Meaning calculated from semiotic input is not objective but is influenced by and grounded in experience of the language acquisition process and the habitual interaction of the speaker with the referents of linguistic content. In this dissertation, I address a phenomenon that exposes the subjectivity of meaning called the Foreign Language Effect (FLE). It refers to the finding that late bilinguals exhibit different decision-making patterns when language content of emotional nature is presented to them in their native (L1) versus non-native (L2) language. I adopt Pavlenko’s (2012) account where she hypothesizes that this behavior reflects disembodiment of L2. I construct a semantic representation of embodied language processing through unifying two theories: The Embodied Simulation Hypothesis (Bergen, 2015a; 2015b) and Frame Semantics (Fillmore, 1976) resulting in a cognitive model of meaning simulation: the Embodied Simulation Frame Semantic blueprint model (ES-FS blueprint). I implement it as an algorithm that calculates an information structure to serve as a representation of embodied meaning simulation yielding an insight to semantic memory with an embodied and grounded lens. The simulation blueprint is composed of frames retrieved from FrameNet: an implementation of Frame Semantics as a network of background knowledge concepts (Ruppenhofer et al., 2016) called frames which depict total experiential situations indexed by words. I test my model on empirical data from the Semantic Priming Project (Hutchison et al., 2013) and find support for it in the L1 English. I then run a semantic priming experiment on L1 and L2 speakers of English to conduct a comparison of meaning processing across the two nativeness conditions. I provide preliminary support to Pavlenko’s account for the FLE from experiential grounding of language, a major factor of disparity in language acquisition and use between the L1 and the L2 in the late bilingual.