PhD Talks

Presentation 1: Mind-wandering during reading, Chen Chen

Presentation 2: The impact of the semantic content of L1 and L2 maskers on L1 target transcription, Emily Rice

Listening to a talker is more challenging when the language of the talker matches that of a competing (masker) talker compared to when the two languages differ. The Target-Masker Similarity Hypothesis attributes this to acoustical overlaps between languages, making it harder to hear them as separate speech streams. For bilinguals, ignoring a first-language (L1) compared to a second-language (L2) masker may also be more challenging when the target is L1, because L1 has higher activation levels, so is more interfering. Additionally, some research suggests that meaningful maskers (e.g., “Rice is often served in round bowls”) may impact target perception more than semantically anomalous maskers (e.g., “The great car met the milk”). This study investigated how L1 target perception is affected by the semantic content of L1 vs. L2 maskers. Spanish-English bilinguals (N = 120) heard a Spanish (L1) target simultaneously with either a Spanish or English (L2) masker that was semantically related or unrelated to the target. Results showed greater listening accuracy when the masker was L2, supporting the Target-Masker Similarity Hypothesis. However, masker semantic relatedness had no effect on listening accuracy, indicating that, for bilinguals at least, target-masker interference does not take place at the semantic level.

Presentation 3: Schema-effects in Temporal Order Memory, Adam Curtis

Schemas are associative knowledge structures that represent statistical regularities that occur across multiple episodes. Schemas can represent information about what you might experience (content) as well as when you might experience it (temporal order). Moreover, schematic temporal order may allow for temporal predictions of the future that, in turn, modulate encoding. Memory retention may be enhanced when the temporal order of events is highly congruent or incongruent with our existing schemas, relative to situations that are unrelated to such knowledge. We designed a temporal order memory paradigm where participants learned the order of a series of exemplars from distinct categories. Participants were required to retrieve category- and exemplar-level information in a cued temporal order task. Participants pre-learned schemas that related to the order of the categories in a sequence, allowing them to make predictions about the category of upcoming exemplars. Sequences of exemplars in the main study were then either congruent, incongruent, or unrelated to the pre-learned schemas. We show clear effects of congruency, with greater memory accuracy in the congruent and incongruent relative to unrelated condition. However, these effects might be more related to violations in content rather than temporal order, providing insight into the representational content of schemas.