Abstract
This study investigated how different modes of micro-level emotion have different impacts on foreign language (LX) memory formation in shallow/perceptual processing. Participants were instructed to orally imitate the words they heard and saw while trying to replicate the emotional tone of the presented sound. Each word corresponded to either positive (LexVal+), neutral (LexVal=), or negative (LexVal-). The valence data for each word were retrieved from the proto-ANEW-JLE (Kanazawa, 2016b; for LX) and ANEW database (Bradley & Lang, 1999; for L1). The emotional prosody of each auditory prime voice clip was either positive (PercVal+), unemotional (PercVal=), or negative (PercVal-). The test session consisted of a free recall memory test, where the numbers of correct responses (dependent variables) were calculated according to (a) PercVal and (b) LexVal (independent variables). It was revealed that (a) PercVal- had a significant facilitatory effect compared to PercVal=; whereas the facilitatory effect of PercVal+ was not statistically significant. (b) LexVal+ and LexVal- were significantly better recalled than LexVal=. Contrary to Kanazawa’s (2016b) positivity effect under the deep/semantic condition, the present results were more congruent with the negativity effect (Bąk, 2016). The results and the rationale further corroborated the Deep Positivity Hypothesis (Kanazawa, 2018; 2020a).
