Cortical processing of human and emoji faces: an ERP analysis

TLDR

Social networks increasingly rely on emojis as non‑verbal emotional cues, often replacing face‑to‑face interaction. The study aimed to determine whether human and emoji faces are processed similarly in the cortex. Thirty participants had P100, N170, and LPP event‑related potentials recorded while viewing happy, neutral, and angry expressions of both human and emoji faces. Human faces produced larger P100 and LPP amplitudes, whereas emoji faces elicited larger N170; angry expressions amplified P100 and LPP, and the overall neural time‑course was comparable, with human faces driving greater attentional orientation, arousal, and engagement and emoji faces enhancing structural encoding—suggesting that emojis evoke neural responses akin to face‑to‑face communication.

Abstract

Social networks have become an effective tool of interaction, frequently replacing face-to-face communication by using non-verbal emotional cues, such as emojis. The aim of the present study was to investigate whether human and emoji faces are analogously processed cortically. We recorded P100, N170, and LPP event-related potentials (ERPs) in response to happy, neutral, and angry expressions of human and emoji faces in 30 participants. The results showed that P100 and LPP amplitudes were larger in response to human faces but emoji faces generated larger N170 amplitude. Angry faces elicited significantly larger P100 and LPP amplitudes. The neural time-course of processing human and emoji faces was similar, however, human faces generate greater attentional orientation response, emoji faces increase the structural encoding processes, and human faces elicited greater arousal and attentional engagement. These results suggest that the use of emoji faces in computer-mediated communication produces neural responses that are similar to those that are observed in face-to-face communication.