Semantic knowledge of words is necessary to produce an incidental self-reference effect.

Encoding information in relation to the self produces a memory advantage compared with other encoding methods. Termed the self-reference effect (SRE), this self-memory advantage has typically been demonstrated when people explicitly evaluate the self-relevancy of to-be-remembered target stimuli. Notably, recent works showed that simply indicating the location of target stimuli in relation to a self-relevant versus other-relevant cue (e.g., one's own or another person's name) under a non-evaluative, non-referential encoding context produces a self-memory advantage. What accounts for this incidental SRE (iSRE)? In the current study, by varying the semantic properties of target stimuli, we asked whether the iSRE mainly results from (a) spontaneous evaluative self-referencing, (b) spontaneous semantic self-referencing, or (c) perceptual processing of target stimuli in spatio-temporal relation to a self-relevant cue. During encoding, participants indicated whether trait adjectives, concrete nouns, or pseudowords were presented to the left or right of their own, their friend's, or a stranger's name. In a subsequent source memory test, we found better source memory for words paired with one's own name versus their friend's and a stranger's name for trait adjectives and concrete nouns but not for pseudowords, with the magnitude of the iSRE similar for trait adjectives and concrete nouns. No self-memory advantage with pseudowords suggests that perceptual processing of target stimuli in relation to a self-cue is insufficient to produce an iSRE, but rather the iSRE draws on a basic semantic connection between the self and a simultaneously presented target stimulus. Potential mechanisms underlying the iSRE, implications, and future research directions are discussed.