Behavioral and Neural Mechanisms Underlying the Filtering of Emoji Distractors in Working Memory: An Event-Related Potential Study

Abstract

Emojis, as emerging paralinguistic cues in computer-mediated communication, are increasingly integrated into daily digital interactions and are known to be efficiently stored as targets in working memory (WM). Despite their pervasive use, the cognitive mechanisms underlying the filtering of emojis as distractors remains unclear. The present study combined behavioral measures and event-related potentials to investigate how emojis are filtered in WM. Participants performed a WM task in which emojis served as distractors. The results showed that emojis, compared with other types of distractors, could be efficiently filtered, as evidenced by reduced unnecessary storage (US) and lower contralateral delay activity (CDA) amplitudes. Moreover, a positive correlation between US and CDA emerged only in the emoji distractor condition, especially among high-frequency emoji users (r = 0.517, p = 0.023), suggesting that prior experience with emojis modulates the link between behavioral and neural indices of filtering. These findings provide preliminary evidence that emojis can be effectively filtered in WM and underscore the modulatory role of usage experience in shaping cognitive processing during digital communication.

Keywords

contralateral delay activity emoji filtering efficiency unnecessary storage working memory

Introduction

With the advent of digital communication, reduced paralinguistic cues have prompted the use of digital symbols as compensatory tools.¹ Emojis, in particular, serve as visual strategies to restore these cues in computer-mediated communication (CMC).² They are vertical, human-like icons that convey emotions through exaggerated facial features.^3–6 Emojis convey affective states, enhance communicative effectiveness,^7,8 and can even elicit social orienting,⁹ reflecting individuals’ inherent sensitivity to social cues.¹⁰ In 2017, WeChat users sent over 600 million emojis daily,¹¹ and globally, >70 billion are exchanged each day,¹² reflecting their widespread role in digital communication.

Given their prominence in online communication,⁸ research on emojis has grown rapidly since 2015.¹³ However, they also pose challenges. Emojis’ generalized emotional functions and inherent ambiguity can hinder understanding,^14,15 whereas inappropriate or excessive use may reduce message credibility in formal contexts¹⁶ or amplify negative emotions.¹⁷ Misinterpretations of negative content may even spark public disputes.¹⁸ In this context, understanding how emojis are processed becomes particularly important.

Derived from basic facial features, emojis function like human faces^7,8 but are recognized more accurately and processed more rapidly.¹⁹ Individuals prioritize attention to social stimuli, which is described and explained by Social Motivation Theory as “social orienting.”^10,20 In our previous study, participants’ working memory (WM) for emojis was better than for faces and comparable to simple shapes across 200, 1,000, and 2,000 ms.⁹ It indicates that emojis are regarded as social cues that evoke social orienting. From a cognitive perspective, examining emojis via WM is appropriate, as WM is a fundamental, continuously active cognitive function.²¹

WM is specialized for the short-term maintenance and storage of information, forming the basis of human thought.²² However, it has limited capacity.²³ Distractors that are irrelevant to the task, such as differently colored squares or faces, can impair WM performance.^24–26 For example, repeating an irrelevant sound (e.g., the word “the”) reduces task performance by interfering with WM.²⁷ When WM is disrupted, individuals experience greater difficulty performing cognitive activities that depend on it, including thinking, reasoning, learning, and comprehension.²² Environmental information far exceeds processing capacity, making distractors almost inevitable.²⁶ Therefore, the filtering ability of irrelevant information is critical,^28,29 because it determines whether WM capacity is allocated to targets.²⁴ Activating filtering via cues reduces distractor intrusion, improving performance without extra attentional load,³⁰ indicating that enhancing filtering efficiency can improve WM independently of capacity. Neural substrates differ: filtering relies on the bilateral insula, right occipital cortex, right brainstem, and right cerebellum, whereas storage engages the bilateral posterior parietal cortex, left ventromedial prefrontal cortex, and right precuneus.³¹ These findings highlight a functional dissociation; however, filtering efficiency has received far less attention than storage in previous research.³²

Colored squares and geometric shapes, which are commonly used as distractors in WM filtering studies, help isolate and quantify core cognitive processes such as storage and attentional allocation.³³ Individual differences in filtering were observed by manipulating the spatial locations of shapes.²⁴ However, such stimuli have limited ecological validity because real-world distractors, such as human faces, convey rich social information and evolutionary significance.³⁴ Although some studies have used facial stimuli,^25,26 related research remains scarce. With social interactions increasingly occurring in digital environments, emojis offer suitable experimental stimuli. When used as WM targets with simple shapes, emojis receive processing priority.⁹ However, it is unclear whether emojis, as distractors, are actively filtered, a key question for understanding WM.

This question can be examined within a theoretical framework. Within this framework, WM is shaped by the interaction between top-down and bottom-up attentional processes.³⁵ It suggests that top-down attention is guided by task demands or internal goals and is supported by WM.^35–38 Accordingly, we previously enhanced the mnemonic priority of emojis by directing goal-oriented attention toward them.⁹ Bottom-up attention, in contrast, is driven by the salience of stimuli.^{35–37,39,40} However, attentional capture can be attenuated when salient but task-irrelevant information is suppressed at the encoding stage.⁴⁰ In the present study, emojis were presented as distractors rather than targets, which may promote their suppression during the encoding stage. Thus, we hypothesized the following: H1:

Emoji distractors can be effectively filtered and excluded from WM maintenance.

In the real world, individuals selectively attend to socially relevant stimuli, interpret cues, and adjust their behavior based on experience.⁴¹ Experience profoundly shapes social face recognition,⁴² as in infancy, where exposure refines recognition toward frequently encountered face types,⁴³ adults exhibit a heightened capacity for automatic recognition of familiar faces. Emojis, as digital paralinguistic cues, share the social and emotional properties of human faces,^9,44,45 and individuals vary in their exposure to them.⁴⁶ However, whether such experiences shape cognitive processing, particularly WM filtering, remains unclear. Therefore, our study examined the role of emoji usage frequency in this process.

Filtering influences WM capacity by allowing unnecessary storage (US) of irrelevant items.^47,48 Therefore, US is commonly used as a behavioral index of filtering efficiency.^29,49,50 In this study, US served as the behavioral measure of individual filtering ability. At the neural level, the contralateral delay activity (CDA) is a well-established electrophysiological marker of distractor filtering in WM.^24,25,28,51 CDA is a negative slow-wave event-related potential (ERP) component reflecting visual WM maintenance, emerging around 200 ms after memory array onset and persisting throughout the retention interval. It represents the number of items retained in WM during the delay period.^{23,25,52–56} Larger CDA amplitudes have been observed in individuals with lower WM capacity when filtering socially neutral or angry face distractors.²⁵ The present study therefore examined whether and how emoji distractors presented during the maintenance phase similarly elicit this component.

Despite growing interest, the cognitive mechanisms of WM filtering remain unclear,³² particularly regarding how individuals suppress distracting or negative stimuli in digital rather than face-to-face communication. Using emojis as digital distractors provides a novel means to examine WM filtering in online contexts, thereby extending existing findings to computer-mediated settings. This approach advances the mechanistic understanding of filtering efficiency and offers a theoretical basis for WM research in digital communication.

Materials and Methods

Experimental design and participants

The experiment used a single-factor, within-subjects design with six conditions. Two conditions had no distractors: one with two targets (T2D0) and one with four targets (T4D0). The other four conditions included two targets and two distractors (T2D2), which differed by distractor type: emoji, scrambled emoji (Semoji), face, and shape. Trials from all conditions were randomly mixed within each block. The dependent variables included multiple behavioral and neural measures assessed across the six levels of the independent variable.

An a priori power analysis was conducted to estimate the required sample size using G*Power 3.1.9.7.⁵⁷ Assuming a medium effect size (f = 0.25) for a one-way repeated-measures analysis of variance (ANOVA) (α = 0.05), a minimum of 28 participants was required to achieve 0.95 power. Ultimately, 50 healthy participants (28 females; mean age = 20.64 ± 2.35 years; range = 18–29 years) participated. All participants were Han Chinese from Chengdu, China, had normal or corrected-to-normal vision, and reported no color blindness or neurological disorders. None reported using psychotropic medication. Sex was recorded by self-report. The study was conducted in accordance with the Declaration of Helsinki and was approved by the Research Ethics Committee of Nanjing University. Written informed consent was obtained from all participants, who were fully debriefed after the experiment.

Materials

Four types of distractors from a previous study were used,⁹ including emojis, scrambled emojis, human faces, and yellow simple shapes. Emoji distractors were taken from WeChat, a popular Chinese social media platform. Scrambled emojis were created by disrupting the facial features of these emojis. Human face distractors were selected from the Chinese Facial Affective Picture System⁵⁸ and standardized to circular shapes with restored original colors to approximate the physical characteristics of emojis.⁹ Simple yellow shapes contained no social information but matched the emojis in color. Blue geometric shapes served as targets, providing a clear color contrast with the distractors. All stimuli were standardized in size (Figure 1 (a)).

FIG. 1.

Stimuli used in the current study (a) and a schematic of a formal working memory task trial (b). The seven human face distractors and seven emoji distractors separately represent anger, disgust, fear, happiness, calmness, sadness, and surprise from left to right. The ID codes of the human faces in the Chinese Facial Affective Picture System were AM1, DM1, FM3, HM97, NEM15, SAM2, and SUM37. They resemble emojis in appearance but lack the symbolic characteristics that define emojis. Scrambled emojis were generated by rearranging emoji features while preserving visual elements protruding outside the face frame and avoiding meaningful structures. Simple yellow shapes were colored light yellow (RGB: 249, 173, 20) to closely match the physical features of emojis, without conveying any social information. The targets were blue shapes (RGB: 21, 96, 130). (b) Illustrates a formal trial of the working memory task in the Emoji distractor condition. In this trial, the correct response for the picture test was J (J = “no change”; F = “change”), indicating that the blue circle in the test phase was one of the targets participants were instructed to maintain in working memory (i.e., the target presented in the memory array phase). The background of the trial was light gray (RGB: 128, 128, 128).

Procedure

Working memory task

A one-sided change-detection task was used to assess WM performance.²³ Participants viewed a memory array and memorized the stimuli indicated by an arrow cue. Before the task, they reported their emoji usage frequency by answering two questions: (a) “How often do you encounter emojis online and in daily life?” and (b) “How often do you use emojis online and in daily life?”⁴⁶

Participants received task instructions and confirmed their understanding. Each trial began with a double-arrow cue presented for 300 ms on a 1,920 × 1,080 display, indicating the side of the memory array to be memorized. After a blank interval of 150–350 ms, the memory array appeared for 200 ms, with items arranged along an invisible ellipse centered on the screen. Stimuli presented on the opposite side of the arrow were of the same type to ensure correspondence. Following a 1,000 ms retention interval, a probe stimulus was presented. Participants were asked to press F if the probe was absent or J if it was present, within 3 seconds. They were instructed to fixate on the central cross throughout the task (Figure 1(b)).

Participants completed six blocks of 192 trials each, for a total of 1,152 trials. After each block, participants could rest before proceeding at their own pace. To ensure task comprehension, they first completed 24 practice trials and were required to reach at least 80 percent accuracy. The full session lasted ∼2 hours.

EEG recording, Behavioral analysis, EEG preprocessing, EEG analysis, and EEG–behavior correlation analysis

See Supplementary Data for details.

Results

Behavioral results

The descriptive statistics for K across various conditions are presented in Table 1. For the T2D2 conditions, US showed a statistically significant main effect of condition, F (3, 117) = 118.874, p < 0.001, η_p² = 0.753. The Face distractor condition (M = 0.734 ± 0.239) was significantly higher than the Emoji distractor condition (M = 0.076 ± 0.222, 95 percent confidence interval [CI]: 0.551–0.766), the Semoji distractor condition (M = 0.047 ± 0.214, 95 percent CI: 0.578–0.796), and the Shape distractor condition (M = 0.183 ± 0.310, 95 percent CI: 0.401–0.702), all with p < 0.001 (Figure 2).

FIG. 2.

Results of a one-way repeated-measures ANOVA for unnecessary storage across the four distractor conditions. The X-axis represents the four distractor conditions, and the Y-axis indicates the magnitude of unnecessary storage for each condition. Error bars represent standard errors. “Semoji” refers to scrambled emojis. ***p < 0.001, **p < 0.01. ANOVA, analysis of variance.

Table 1.

Results of Descriptive Statistics for K in This Study

Condition	n	Mean	SD	SE	Coefficient of variation
2-Target	40	1.372	0.237	0.038	0.173
4-Target	40	1.301	0.583	0.092	0.448
Face distractor	40	0.638	0.293	0.046	0.459
Emoji distractor	40	1.296	0.308	0.049	0.238
Semoji distractor	40	1.325	0.316	0.050	0.238
Shape distractor	40	1.190	0.393	0.062	0.330

Note: n, the number of participants. SD, standard deviation; SE, standard error. “Semoji”, scrambled emoji.

The Face distractor condition (r = −0.618, 95 percent CI: −0.780to −0.380), Emoji distractor condition (r = −0.643, 95 percent CI: −0.796 to −0.415), Semoji distractor condition (r = −0.660, 95 percent CI: −0.806 to −0.439), and Shape distractor condition (r = −0.797, 95 percent CI: −0.888 to −0.645) were all significantly correlated with K within their respective conditions (all p < 0.001).

EEG results

Contralateral delay activity

There was a statistically significant main effect of condition for CDA, F(5, 195) = 5.907, p < 0.001, η_p² = 0.132. The T2D0 condition (M = −0.392 ± 0.853) was significantly different from the T4D0 condition (M = −0.870 ± 0.708, 95 percent CI: 0.034–0.923, p = 0.026), the Face distractor condition (M = −0.851 ± 0.845, 95 percent CI: 0.101–0.818, p = 0.004), the Semoji distractor condition (M = −0.966 ± 0.646, 95 percent CI: 0.219–0.929, p < 0.001), and the Shape distractor condition (M = −0.893 ± 0.759, 95 percent CI: 0.101–0.901, p = 0.005). No significant difference was found between the T2D0 condition and the Emoji distractor condition (M = −0.747 ± 0.799, 95 percent CI: 0.101–0.901, p = 0.180), as shown in Figures 3 and 4.

FIG. 3.

N2pc and CDA waveforms (in μV) across all conditions. The first translucent gray rectangle indicates the time window used to calculate the mean N2pc amplitude for each condition (200–300 ms). The second rectangle indicates the time window used to calculate the mean CDA amplitude for each condition (400–1,000 ms). Both ERP components were calculated as the difference between contralateral and ipsilateral amplitudes at parietal (P3/P4) and parietal–occipital (PO7/PO8, PO3/PO4) electrodes. The X-axis represents time, and the Y-axis represents amplitude. The gray solid line represents the 2-target condition, and the gray dashed line represents the 4-target condition. Red, yellow, green, and blue solid lines represent the face, emoji, scrambled emoji, and shape distractor conditions, respectively. “Semoji” refers to scrambled emoji. CDA, contralateral delay activity; ERP, event-related potential.

FIG. 4.

Results of a one-way repeated-measures ANOVA on CDA amplitude across all conditions. The X-axis represents six conditions, and the Y-axis indicates the CDA amplitude for each condition. Error bars represent standard errors. “Semoji” refers to scrambled emoji. ***p < 0.001, **p < 0.01, *p < 0.05.

N2-posterior-contralateral

A significant main effect of condition was observed for N2-posterior-contralateral (N2pc), F(5, 195) = 17.04, p < 0.001, η_p² = 0.304 (Table 2).

Table 2.

Results of N2pc Across All Conditions in This Study

Condition	MD	95% CI	SE	t	d	p
2-Traget
4-target	0.527	(0.119, 0.934)	0.130	4.044	0.619	0.004^**
Face distractor	0.522	(0.146, 0.899)	0.120	4.336	0.614	0.001^**
Emoji distractor	0.619	(0.258, 0.981)	0.116	5.362	0.729	<0.001^***
Semoji distractor	0.885	(0.543, 1.228)	0.110	8.082	1.041	<0.001^***
Shape distractor	0.788	(0.356, 1.220)	0.138	5.707	0.927	<0.001^***
4-Target
Semoji distractor	0.359	(0.028, 0.689)	0.106	3.392	0.422	0.024^*
Face distractor
Semoji distractor	0.363	(0.070, 0.656)	0.094	3.875	0.427	0.006^**
Emoji distractor
Semoji distractor	0.266	(0.012, 0.520)	0.081	3.272	0.313	0.034^*

Note: MD, mean difference; 95 percent CI, confidence interval; t, t value obtained from post hoc comparisons; d, Cohen’ s d effect size. “Semoji”, scrambled emoji.

*p < 0.05, **p < 0.01, ***p < 0.001.

EEG–behavior correlations

Only the CDA amplitude in the Emoji distractor condition was significantly correlated with US (r = 0.419, 95 percent CI: 0.124–0.646, p = 0.007). After dividing participants based on the frequency of emoji usage, a significant correlation between CDA amplitude under the Emoji distractor condition and US was only found in the high-frequency group of emoji usage (r = 0.517, 95 percent CI: 0.082–0.787, p = 0.023).

Discussion

This study examined whether emojis, as distractors, are filtered in WM. Both behavioral (US) and neural (CDA) results indicated effective filtering, supporting our hypothesis. A positive correlation emerged only in the emoji condition and was driven by individuals with higher emoji use. These findings highlight the unique nature of emojis as digital symbols.

Higher US values indicate lower filtering efficiency. Face distractors showed the highest US, reflecting greater cognitive demands and lowest filtering efficiency. Prior research has established that faces are special stimuli that selectively engaging regions such as the occipital face area, superior temporal sulcus, and fusiform face area compared to other objects.^59–61 In contrast, emojis elicited lower US values, indicating that they were efficiently filtered despite conveying emotional content similar to that of faces.⁴⁵ Furthermore, despite being physically more complex than simple shapes, their filtering efficiency did not differ significantly. Overall, these findings align with previous evidence indicating that emojis consume relatively few cognitive resources in WM.⁹

Both capacity and filtering are fundamental for assessing the functional integrity of WM.^32,62 In this study, filtering efficiency, indexed by US, correlated negatively with K across the four T2D2 conditions, indicating that higher filtering efficiency was associated with greater WM capacity. This replicates prior findings^47,50,56 and underscores the role of filtering efficiency in individual WM difference.^63,64

Analysis of the CDA revealed significantly stronger negative amplitudes in the T4D0, Face, Semoji, and Shape distractor conditions than in the T2D0 condition, indicating that these distractors were encoded and maintained in WM. Notably, CDA amplitude in the Emoji distractor condition did not differ significantly from that in the T2D0 condition and was the weakest among all T2D2 conditions, suggesting that emojis required minimal suppression and elicited reduced neural activity.⁶² Stimuli that initially capture attention are not necessarily obligatorily maintained in WM when task goals require suppression. Previous research has shown that highly discriminable features are automatically encoded into WM, whereas irrelevant fine-grained features are filtered out.⁶⁵ To some extent, emojis convey emotions in a relatively coarse and generalized manner,⁶⁶ making it difficult to infer individuating identity characteristics.⁶⁷ Moreover, emojis within the same category share highly similar visual features and lack discriminability, which may reduce their likelihood of being prioritized and maintained in WM.

N2pc was initially interpreted as reflecting the suppression of surrounding irrelevant information,⁶⁸ but it was later considered as a marker of selective attention.⁶⁹ It corresponds to the beginning of target localization in the ventral visual pathway.⁷⁰ In this study, participants were required to focus on targets while ignoring distractors; thus, the emergence of N2pc was expected. Across both the T4D0 condition and the four T2D2 conditions, N2pc amplitudes were significantly stronger than those in the T2D0 condition. Interestingly, the N2pc amplitude in the Semoji distractor condition exceeded that in the T4D0, Face distractor, and Emoji distractor conditions. Because the N2pc analysis was conducted subsequent to the CDA analysis, these results should be interpreted with caution.

A positive correlation was found between US and CDA amplitude in the Emoji distractor condition, indicating that stronger negative CDA waves were associated with lower US values. This relationship emerged only among high-frequency emoji users. As WM represents a form of explicit memory⁷¹ and CDA indexes the active suppression of distractors, the results indicate that extensive emoji experience may strengthen inhibitory control over emoji interference during WM maintenance.⁶² Such experience-dependent modulation aligns with evidence linking object familiarity to improved discrimination.⁷² These findings may explain the reduced cognitive resource demands associated with emojis, evident at both neural and behavioral levels.

Limitations

This study investigated the filtering mechanisms of emoji distractors in this study. Although demographic variables such as gender and age are known to influence emoji use and comprehension,^3,46,73,74 the present study did not examine these factors in depth. Future research should address this limitation. Cross-cultural studies should be adopted in the future to account for potential cultural differences.⁷⁵ Although the sample size was statistically adequate a priori, future research should employ larger cross-sectional samples to validate the findings.

Moreover, personality traits may influence emoji use and perception. For example, individuals high in neuroticism tend to use emojis more frequently to manage social interactions,⁷⁶ whereas trait anxiety may impair inhibitory control in WM.^77,78 These variables should be considered in future research, as they may jointly affect emoji processing and filtering efficiency.

We only used EEG technology to examine the temporal resolution of the filtering process. Future research could employ other neuroimaging techniques, such as functional magnetic resonance imaging, to assess brain activity and functional connectivity with higher spatial resolution. Because distractor filtering represents a dynamic, multi-stage process that engages distinct brain regions.^32,64,79,80

Frequent switching of filtering settings can reduce efficiency, causing more irrelevant information to enter WM.⁸¹ In this study, the randomized presentation of target positions and distractor types may have lowered filtering efficiency, suggesting stricter control in future work.

Significance

Previous work quantified the storage capacity for emojis.⁹ Here, we assessed their filtering efficiency in WM, highlighting another key aspect of this cognitive function. It highlights the fundamental importance of filtering irrelevant distractors for WM performance, offering robust empirical support for the WM filtering model.²⁴ By including emojis rather than traditional stimuli such as shapes and colors,³³ the paradigm gains ecological validity. This approach advances the understanding of WM in social contexts and provides a basis for future studies on attention,⁸² executive functions,⁸³ and related cognitive processes in digital communication.

The present study adopts the theoretical framework of top-down and bottom-up attentional interactions in WM to investigate the cognitive mechanisms underlying the processing of social cues, specifically emojis. The findings demonstrate that emojis do not inevitably penetrate attentional filters to gain access to WM. Instead, such distractors can be effectively suppressed under the constraints of explicit task goals. These results support the theoretical position that WM representations are not purely driven by bottom-up stimulus salience; rather, they are jointly modulated by goal-directed top-down control and stimulus-driven bottom-up processes.^35–37,39 It also provides a comprehensive account of how emojis function in CMC.¹

The present study revisits this isolated approach in basic cognitive research, addressing a theoretical gap and enabling broader applications. Beyond semiotics,² emojis are increasingly applied in medicine,⁸⁴ biology,⁸⁵ and so on. For instance, they can facilitate communication of scientific health information, supporting public health initiatives.⁸⁶

Training programs targeting both WM capacity and filtering efficiency can enhance overall performance.⁴⁹ Such approaches could be applied into future WM interventions to improve the filtering of irrelevant information and support cognitive function.⁸⁷ This perspective deepens our understanding of the interplay between storage and processing in WM⁶² and informs the design of more effective memory-training and cognitive-enhancement strategies.

Individuals with autism spectrum disorder (ASD) often show difficulties in perceiving social cues,⁸⁸ such as faces.⁸⁹ Interestingly, children with ASD may attend more to cartoon faces, particularly focusing on the eyes,⁹⁰ in contrast to the “eye avoidance” commonly observed with real faces.⁹¹ Given their semiotic and pictorial nature,² emojis may attract attention in individuals with ASD and facilitate engagement with social cues. Studying emoji use could therefore provide valuable insights into social cognition in ASD and inform strategies to improve daily interactions and quality of life.

Conclusions

This study demonstrates that emojis, as a distinct class of stimuli, are efficiently filtered in WM. Compared with other distractors, emojis elicited smaller US and the lowest CDA amplitudes, reflecting reduced cognitive demands. A positive correlation between US and CDA was observed and was subsequently found only among high-frequency emoji users, highlighting the modulatory role of experience in shaping filtering performance. These findings clarify experience-dependent mechanisms in social information processing and inform future research on cognitive functioning in digital contexts.

Authors’ Contributions

D.Z.: Data acquisition, formal analysis, data interpretation, and writing—original draft. D.G.: Conceptualization, methodology, and funding acquisition. R.T.: Review, supervision, and overall responsibility.

Ethical Statement

The study was conducted in accordance with the Declaration of Helsinki and approved by the Research Ethics Committee of Nanjing University.

Footnotes

Acknowledgments

The authors thank Ruyi Liu for assistance with programming, as well as Lin Luo and Jing Wang for assistance with data collection. The authors also thank Tinghao Tang and acknowledge the use of artificial intelligence (AI) tools (ChatGPT, OpenAI) for grammar correction and language refinement during article preparation. All content was reviewed and verified by the authors.

Author Disclosure Statement

No competing financial interests exist.

Funding Information

This research is supported by the National Natural Science Foundation of China (Grant Nos.: 31571131 to R.T.).

Supplemental Material

References

1. Dresner

, Herring

. Functions of the nonverbal in CMC: Emoticons and illocutionary force. Commun Theory 2010;20(3):249–268; doi: 10.1111/j.1468-2885.2010.01362.x

2. Kerslake

, Wegerif

. The semiotics of emoji: The rise of visual language in the age of the internet. MaC 2017;5(4):75–78; doi: 10.17645/mac.v5i4.1041

3. Dalle Nogare

, Cerri

, Proverbio

. Emojis are comprehended better than facial expressions, by male participants. Behav Sci (Basel) 2023;13(3):278; doi: 10.3390/bs13030278

4. Kaye

, Rodriguez-Cuadrado

, Malone

, et al. How emotional are emoji?: Exploring the effect of emotional valence on the processing of emoji stimuli. Comput Hum Behav 2021;116:106648; doi: 10.1016/j.chb.2020.106648

5. Kaye

, Darker

, Rodriguez-Cuadrado

, et al. The emoji spatial stroop task: Exploring the impact of vertical positioning of emoji on emotional processing. Comput Hum Behav 2022;132:107267; doi: 10.1016/j.chb.2022.107267

6. Pavalanathan

, Eisenstein

. More emojis, less :) the competition for paralinguistic function in microblog writing. FM 2016; doi: 10.5210/fm.v21i11.6879

7. Chatzichristos

, Morante

, Andreadis

, et al. Emojis influence autobiographical memory retrieval from reading words: An fMRI-based study. PLoS One 2020;15(7):e0234104; doi: 10.1371/journal.pone.0234104

8. Jin

, Deng

, Wu

, et al. Emoji image symbol’s social function and application. Adv Psychol Sci 2022;30(5):1062–1077; doi: 10.3724/SP.J.1042.2022.01062

9. Li

, Zhang

, Liu

, et al. Processing facial emojis as social information: Evidence from visual working memory for facial emojis, simple shapes, human faces, and their relations to theory of mind. Comput Hum Behav 2024;153:108106; doi: 10.1016/j.chb.2023.108106

10.

10. Chevallier

, Kohls

, Troiani

, et al. The social motivation theory of autism. Trends Cogn Sci 2012;16(4):231–239; doi: 10.1016/j.tics.2012.02.007

11.

11. Liu

, Gao

, Wang

. An empirical study on the communication and usage psychology of emoji in Wechat. OBM Neurobiol 2022;06(04):1–18; doi: 10.21926/obm.neurobiol.2204142

12.

12. George

DAS

, George

ASH

, Baskar

. Emoji unite: Examining the rise of emoji as an international language bridging cultural and generational divides. Partn Univers Int Innov J 2023;1(4):183–204; doi: 10.5281/zenodo.8280356

13.

13. Bai

, Dan

, Mu

, et al. A systematic review of emoji: Current research and future perspectives. Front Psychol 2019;10:2221; doi: 10.3389/fpsyg.2019.02221

14.

14. Weissman

, Tanner

. A strong wink between verbal and emoji-based irony: How the brain processes ironic emojis during language comprehension. PLoS One 2018;13(8):e0201727; doi: 10.1371/journal.pone.0201727

15.

15. Kutsuzawa

, Umemura

, Eto

, et al. Classification of 74 facial emoji’s emotional states on the valence-arousal axes. Sci Rep 2022;12(1):398; doi: 10.1038/s41598-021-04357-7

16.

16. Dai

, Gao

, Leng

, et al. Congruent or conflicting? The interaction between emoji and textual sentence is not that simple!. Heliyon 2024;10(12):e32984; doi: 10.1016/j.heliyon.2024.e32984

17.

17. Khalid

, Bukhari

, Maqsood

. The impact of emojis on text perception: A descriptive analysis of WhatsApp users. Pak Lang Humanit Rev 2024;8(3):326–338; doi: 10.47205/plhr.2024(8-III)30

18.

18. Nguyen

GBH

. How to Be Impolite with Emojis: A Corpus Analysis of Vietnamese Social Media Posts. PhD Thesis. Purdue University Graduate School; 2023.

19.

19. Dalle Nogare

, Proverbio

. Emojis vs. facial expressions: An electrical neuroimaging study on perceptual recognition. Soc Neurosci 2023;18(1):46–64; doi: 10.1080/17470919.2023.2203949

20.

20. Keifer

, Dichter

, McPartland

, et al. Social motivation in autism: Gaps and directions for measurement of a putative core construct. Behav Brain Sci 2019;42:e95; doi: 10.1017/S0140525X18002418

21.

Logie

Working memory: Beyond Baddeley and Hitch. SAGE Publications; 2015; pp. 86–104

22.

22. Baddeley

. Working memory: Looking back and looking forward. Nat Rev Neurosci 2003;4(10):829–839; doi: 10.1038/nrn1201

23.

23. Vogel

, Machizawa

. Neural activity predicts individual differences in visual working memory capacity. Nature 2004;428(6984):748–751; doi: 10.1038/nature02447

24.

24. Vogel

, McCollough

, Machizawa

. Neural measures reveal individual differences in controlling access to working memory. Nature 2005;438(7067):500–503; doi: 10.1038/nature04171

25.

25. Ye

, Xu

, Liu

, et al. The impact of visual working memory capacity on the filtering efficiency of emotional face distractors. Biol Psychol 2018;138:63–72; doi: 10.1016/j.biopsycho.2018.08.009

26.

26. Ye

, Xu

, Pan

, et al. The differential impact of face distractors on visual working memory across encoding and delay stages. Atten Percept Psychophys 2024;86(6):2029–2041; doi: 10.3758/s13414-024-02895-6

27.

27. Baddeley

. The episodic buffer: A new component of working memory? Trends Cogn Sci 2000;4(11):417–423; doi: 10.1016/S1364-6613(00)01538-2

28.

28. Jost

, Bryck

, Vogel

, et al. Are old adults just like low working memory young adults? Filtering efficiency and age differences in visual working memory. Cereb Cortex 2011;21(5):1147–1154; doi: 10.1093/cercor/bhq185

29.

29. Lee

E-Y

, Cowan

, Vogel

, et al. Visual working memory deficits in patients with Parkinson’s disease are due to both reduced storage capacity and impaired ability to filter out irrelevant information. Brain 2010;133(9):2677–2689; doi: 10.1093/brain/awq197

30.

30. Allon

, Luria

. Filtering performance in visual working memory is improved by reducing early spatial attention to the distractors. Psychophysiology 2019;56(5):e13323; doi: 10.1111/psyp.13323

31.

31. Vellage

, Becke

, Strumpf

, et al. Filtering and storage working memory networks in younger and older age. Brain Behav 2016;6(11):e00544; doi: 10.1002/brb3.544

32.

32. Zhang

, Zhang

, Gong

. The filtering efficiency in visual working memory. Adv Psychol Sci 2021;29(4):635–651; doi: 10.3724/SP.J.1042.2021.00635

33.

33. Luck

, Vogel

. The capacity of visual working memory for features and conjunctions. Nature 1997;390(6657):279–281; doi: 10.1038/36846

34.

34. Öhman

, Lundqvist

, Esteves

. The face in the crowd revisited: A threat advantage with schematic stimuli. J Pers Soc Psychol 2001;80(3):381–396; doi: 10.1037/0022-3514.80.3.381

35.

35. Zheng

, Sun

, Wu

, et al. The interaction of top–down and bottom–up attention in visual working memory. Sci Rep 2024;14(1):17397; doi: 10.1038/s41598-024-68598-y

36.

36. Allen

. Prioritizing targets and minimizing distraction within limited capacity working memory. J Cogn 2019;2(1):32; doi: 10.5334/joc.75

37.

37. Oberauer

. Working memory and attention – response to commentaries. J Cogn 2019;2(1):30; doi: 10.5334/joc.79

38.

38. Gaspelin

, Luck

. “Top-down” does not mean “voluntary”. J Cogn 2018;1(1); doi: 10.5334/joc.28

39.

39. Hitch

, Allen

, Baddeley

. Attention and binding in visual working memory: Two forms of attention and two kinds of buffer storage. Atten Percept Psychophys 2020;82(1):280–293; doi: 10.3758/s13414-019-01837-x

40.

40. Gaspelin

, Luck

. The role of inhibition in avoiding distraction by salient stimuli. Trends Cogn Sci 2018;22(1):79–92; doi: 10.1016/j.tics.2017.11.001

41.

41. Camacho

, Deshpande

, Perino

. The cognitive–affective social processing and emotion regulation (CASPER) model. Neuropsychopharmacology 2026;51(1):136–152; doi: 10.1038/s41386-025-02224-x

42.

42. Young

, Burton

. Are we face experts? Trends Cogn Sci 2018;22(2):100–110; doi: 10.1016/j.tics.2017.11.007

43.

43. Pascalis

, De Haan

, Nelson

. Is face processing species-specific during the first year of life? Science 2002;296(5571):1321–1323; doi: 10.1126/science.1070223

44.

44. Erle

, Schmid

, Goslar

, et al. Emojis as social information in digital communication. Emotion 2022;22(7):1529–1543; doi: 10.1037/emo0000992

45.

45. Fischer

, Herbert

. Emoji as affective symbols: Affective judgments of emoji, emoticons, and human faces varying in emotional content. Front Psychol 2021;12:645173; doi: 10.3389/fpsyg.2021.645173

46.

46. Prada

, Rodrigues

, Garrido

, et al. Motives, frequency and attitudes toward emoji and emoticon use. Telemat Inform 2018;35(7):1925–1934; doi: 10.1016/j.tele.2018.06.005

47.

47. Awh

, Vogel

. The bouncer in the brain. Nat Neurosci 2008;11(1):5–6; doi: 10.1038/nn0108-5

48.

48. Emrich

, Busseri

. Re-evaluating the relationships among filtering activity, unnecessary storage, and visual working memory capacity. Cogn Affect Behav Neurosci 2015;15(3):589–597; doi: 10.3758/s13415-015-0341-z

49.

49. Geiter

, Mazreku

, Foresti

, et al. Distractor filtering and task load in working memory training in healthy older adults. Sci Rep 2024;14(1):24583; doi: 10.1038/s41598-024-76376-z

50.

50. Spronk

, Vogel

, Jonkman

. Electrophysiological evidence for immature processing capacity and filtering in visuospatial working memory in adolescents. PLoS One 2012;7(8):e42262; doi: 10.1371/journal.pone.0042262

51.

51. Owens

, Koster

EHW

, Derakshan

. Impaired filtering of irrelevant information in dysphoria: An ERP study. Soc Cogn Affect Neurosci 2012;7(7):752–763; doi: 10.1093/scan/nsr050

52.

52. Allon

, Balaban

, Luria

. How low can you go? Changing the resolution of novel complex objects in visual working memory according to task demands. Front Psychol 2014;5:265; doi: 10.3389/fpsyg.2014.00265

53.

53. Luria

, Balaban

, Awh

, et al. The contralateral delay activity as a neural measure of visual working memory. Neurosci Biobehav Rev 2016;62:100–108; doi: 10.1016/j.neubiorev.2016.01.003

54.

54. McCollough

, Machizawa

, Vogel

. Electrophysiological measures of maintaining representations in visual working memory. Cortex 2007;43(1):77–94; doi: 10.1016/S0010-9452(08)70447-7

55.

55. Salahub

, Lockhart

, Dube

, et al. Electrophysiological correlates of the flexible allocation of visual working memory resources. Sci Rep 2019;9(1):19428; doi: 10.1038/s41598-019-55948-4

56.

56. Spronk

, Vogel

, Jonkman

. No behavioral or ERP evidence for a developmental lag in visual working memory capacity or filtering in adolescents and adults with ADHD. PLoS One 2013;8(5):e62673; doi: 10.1371/journal.pone.0062673

57.

57. Faul

, Erdfelder

, Lang

A-G

, et al. G*power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behav Res Methods 2007;39(2):175–191; doi: 10.3758/BF03193146

58.

58. Xu

, Huang

Y-X

, Yan

, et al. Revision of the Chinese facial affective picture system. Chin Ment Health J 2011;25:40–46.

59.

59. Hoehl

, Peykarjou

. The early development of face processing—What makes faces special? Neurosci Bull 2012;28(6):765–788; doi: 10.1007/s12264-012-1280-0

60.

60. Kamps

, Hendrix

, Brennan

, et al. Connectivity at the origins of domain specificity in the cortical face and place networks. Proc Natl Acad Sci U S A 2020;117(11):6163–6169; doi: 10.1073/pnas.1911359117

61.

61. Kosakowski

, Cohen

, Takahashi

, et al. Selective responses to faces, scenes, and bodies in the ventral visual pathway of infants. Curr Biol 2022;32(2):265–274.e5; doi: 10.1016/j.cub.2021.10.064

62.

62. Feldmann-Wüstefeld

, Vogel

. Neural evidence for the contribution of active suppression during working memory filtering. Cereb Cortex 2019;29(2):529–543; doi: 10.1093/cercor/bhx336

63.

63. Luck

, Vogel

. Visual working memory capacity: From psychophysics and neurobiology to individual differences. Trends Cogn Sci 2013;17(8):391–400; doi: 10.1016/j.tics.2013.06.006

64.

64. McNab

, Klingberg

. Prefrontal cortex and basal ganglia control access to working memory. Nat Neurosci 2008;11(1):103–107; doi: 10.1038/nn2024

65.

65. Zhou

, Yin

, Chen

, et al. Visual working memory capacity does not modulate the feature-based information filtering in visual working memory. PLoS One 2011;6(9):e23873; doi: 10.1371/journal.pone.0023873

66.

66. Częstochowska

, Gligorić

, Peyrard

, et al. On the context-free ambiguity of emoji. ICWSM 2022;16:1388–1392; doi: 10.1609/icwsm.v16i1.19393

67.

67. Robertson

, Magdy

, Goldwater

. Identity Signals in Emoji Do Not Influence Perception of Factual Truth on Twitter. 2021; doi: 10.48550/arXiv.2105.03160

68.

68. Luck

, Hillyard

. Spatial filtering during visual search: Evidence from human electrophysiology. J Exp Psychol Hum Percept Perform 1994;20(5):1000–1014; doi: 10.1037/0096-1523.20.5.1000

69.

69. Eimer

. The N2pc component as an indicator of attentional selectivity. Electroencephalogr Clin Neurophysiol 1996;99(3):225–234; doi: 10.1016/0013-4694(96)95711-9

70.

70. Stoletniy

, Alekseeva

, Babenko

, et al. The N2pc component in studies of visual attention. Neurosci Behav Physi 2022;52(8):1299–1309; doi: 10.1007/s11055-023-01359-y

71.

71. Persuh

, LaRock

, Berger

. Working memory and consciousness: The current state of play. Front Hum Neurosci 2018;12:78; doi: 10.3389/fnhum.2018.00078

72.

72. Reeder

, Stein

, Peelen

. Perceptual expertise improves category detection in natural scenes. Psychon Bull Rev 2016;23(1):172–179; doi: 10.3758/s13423-015-0872-x

73.

73. Gallud

, Fardoun

, Andres

, et al. A Study on How Older People Use Emojis. In: Proceedings of the XIX International Conference on Human Computer Interaction Association for Computing Machinery: New York, NY, USA; 2018; pp. 1–4; doi: 10.1145/3233824.3233861

74.

74. Garcia

, Țurcan

, Howman

, et al. Emoji as a tool to aid the comprehension of written sarcasm: Evidence from younger and older adults. Comput Hum Behav 2022;126:106971; doi: 10.1016/j.chb.2021.106971

75.

75. Togans

, Holtgraves

, Kwon

, et al. Digitally saving face: An experimental investigation of cross-cultural differences in the use of emoticons and emoji. J Pragmat 2021;186:277–288; doi: 10.1016/j.pragma.2021.09.016

76.

76. Liu

, Sun

. To express or to end? Personality traits are associated with the reasons and patterns for using emojis and stickers. Front Psychol 2020;11:1076; doi: 10.3389/fpsyg.2020.01076

77.

77. Qi

, Ding

, Li

. Neural correlates of inefficient filtering of emotionally neutral distractors from working memory in trait anxiety. Cogn Affect Behav Neurosci 2014;14(1):253–265; doi: 10.3758/s13415-013-0203-5

78.

78. Stout

, Shackman

, Larson

. Failure to filter: Anxious individuals show inefficient gating of threat from working memory. Front Hum Neurosci 2013;7:58; doi: 10.3389/fnhum.2013.00058

79.

79. Liesefeld

, Liesefeld

, Zimmer

. Intercommunication between prefrontal and posterior brain regions for protecting visual working memory from distractor interference. Psychol Sci 2014;25(2):325–333; doi: 10.1177/0956797613501170

80.

80. Peverill

, McLaughlin

, Finn

, et al. Working memory filtering continues to develop into late adolescence. Dev Cogn Neurosci 2016;18:78–88; doi: 10.1016/j.dcn.2016.02.004

81.

81. Jost

, Mayr

. Switching between filter settings reduces the efficient utilization of visual working memory. Cogn Affect Behav Neurosci 2016;16(2):207–218; doi: 10.3758/s13415-015-0380-5

82.

82. Posner

, Petersen

. The attention system of the human brain. Annu Rev Neurosci 1990;13(1):25–42; doi: 10.1146/annurev.ne.13.030190.000325

83.

83. Miyake

, Friedman

, Emerson

, et al. The unity and diversity of executive functions and their contributions to complex “frontal lobe” tasks: A latent variable analysis. Cogn Psychol 2000;41(1):49–100; doi: 10.1006/cogp.1999.0734

84.

84. Szeto

, Barber

, Ranpariya

, et al. Emojis and emoticons in health care and dermatology communication: Narrative review. JMIR Dermatol 2022;5(3):e33851; doi: 10.2196/33851

85.

85. Mammola

, Falaschi

, Ficetola

. Biodiversity communication in the digital era through the emoji tree of life. iScience 2023;26(12):108569; doi: 10.1016/j.isci.2023.108569

86.

86. Lin

, Luo

. Emoji and visual complexity in health information design: A moderated serial mediation model. Telemat Inform 2023;85:102065; doi: 10.1016/j.tele.2023.102065

87.

87. Schweizer

, Hampshire

, Dalgleish

. Extending brain-training to the affective domain: Increasing cognitive and affective executive control through emotional working memory training. PLoS One 2011;6(9):e24372; doi: 10.1371/journal.pone.0024372

88.

88. Klin

, Jones

, Schultz

, et al. Visual fixation patterns during viewing of naturalistic social situations as predictors of social competence in individuals with autism. Arch Gen Psychiatry 2002;59(9):809–816; doi: 10.1001/archpsyc.59.9.809

89.

89. Schultz

. Developmental deficits in social perception in autism: The role of the amygdala and fusiform face area. Int J Dev Neurosci 2005;23(2–3):125–141; doi: 10.1016/j.ijdevneu.2004.12.012

90.

90. Liu

, Jiang

, Wang

, et al. The specificity of faces processing in children with autism: Evidences from eye movement study. Adv Psychol 2024;14:400; doi: 10.12677/ap.2024.144234

91.

91. Tanaka

, Sung

. The “eye avoidance” hypothesis of autism face processing. J Autism Dev Disord 2016;46(5):1538–1552; doi: 10.1007/s10803-013-1976-7

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.02 MB

0.00 MB