Encoding In Social Context Enhances And Biases Long-Term Recognition Memory: Behavioral And Electrophysiological Evidence

Encoding often occurs in social contexts, yet research has hardly addressed their role in verbal memory. In three experiments, we investigated the behavioral and neural effects of encoding context on memory for positive, negative, and neutral adjectives, contrasting a social-feedback group (N=24) with an explicit verbal-learning (N=24) and a levels-of-processing group (N=24). Participants in the social-feedback group were not aware of a recognition session one week later, but their memory was better than the explicit learning or the levels-of-processing groups’. However, they also exhibited the strongest response bias, particularly for positive words. Brain event-related potentials (ERPs) revealed largest early negativities (EPN) and late positivities (LPP) in the social-feedback group. Only in the subsequent slow-wave did the explicit learning group show higher amplitudes than the other two groups, suggesting reliance on strategic rather than automatic processes. Still, context-driven incidental encoding outweighed explicit instructions, specifying a decisive role of social factors in memory.


Introduction
Humans are a social species. Therefore, human memory formation, as much other cognitive activity, often occurs in social contexts. Yet, scienti c memory research has paid limited attention to the social situatedness of memory encoding, taking memory operations devoid of contextual embedding as its point of departure (Ebbinghaus (1885(Ebbinghaus ( , 2013, thereby delineating many key principles. Meanwhile, contextual binding in space and time has proven crucial for successful episodic memory formation . The social encoding context has, by comparison, received little attention, although recently social in uences on memory systems have attracted scienti c interest. These range from investigations into brain mechanisms of social memory conformity (Edelson et al., 2011) to the nding that the hippocampus, a critical region for spatial navigation and episodic memory formation, also represents a "social space" of relationships . Such ndings lend momentum to the hypothesis that the social context in which items are encoded affects their subsequent remembering.
Memory research also initially focused on content-general mechanisms. However, human memory does not treat all contents equally. Ample evidence indicates that emotionally relevant material is remembered differently and often better than neutral material (Dolcos & Cabeza, 2002;Dolcos et al., 2017), which may be due to both higher accuracy and response bias (Kensinger & Corkin, 2003;Ochsner, 2000). Emotion effects may be driven by arousal, equally affecting positive and negative contents, or by positive or negative valence (Dolcos & Cabeza, 2002; Kensinger & Corkin, 2004).
The content's self-reference also affects memory (Rogers, 1977;. Self-referential processing is experimentally induced by a corresponding processing instruction on individually presented words, which enhances recognition or recall. This "self-reference effect" is attributed to elaborative stimulus processing and deep encoding , similar to other levels of processing manipulations, such as concreteness decisions (Craik & Lockhart, 1972). Setting self-referential processing apart from other encoding manipulations, recent theories propose that the "self" acts as a hub, enabling e cient integration and storage of self-relevant information . Self-referential processing in social contexts is so far hardly understood. However, being confronted with social evaluations from somebody else likely creates a powerful social context, inducing self-referential updating to detect discrepancies with one's self-concept, adapt it, or reject the information (see . In healthy people, such updating processes are often positively-biased, inducing more elaborative processing of self-serving (rather than merely self-relevant) information . Self-reference and emotion appear to interact and engage at least partly overlapping neural systems (Gutchess & Kensinger, 2018).
Event-related brain potentials have been used to study stimulus encoding under various processing demands, ranging from free viewing and incidental encoding (Kissler et al., 2007), over levels of processing instructions (Rugg & Curran, 2007), and stimulus appraisals (Dolcos & Cabeza, 2002), to self-referential processing (Herbert et al., 2011) or processing in evaluative social contexts (Schindler et al., 2015). Early posterior negative (EPN) potentials appearing over visual cortex around 200 ms post-stimulus generally index feature-based attention and conceptual stimulus encoding, which can gate episodic encoding (Schupp et al., 2006). EPN potentials have been found to co-vary with task-driven attention deployment , emotionally motivated attention (Junghöfer et al., 2001), self-reference (Herbert et al., 2011), but also contextually induced social relevance (Schindler et al., 2015). A series of later surface-positive potentials occurring over frontal and parietal brain areas index episodic memory encoding proper (e.g., see Friedman & Johnson, 2000). The parietal part of these effects, typically known as late positive potential (LPP) and occurring from about 400 ms post stimulusonset, is also sensitive to explicit emotional appraisal (Dolcos and Cabeza, 2002) and self-reference (Herbert et al., 2011). Fronto-parietal slow waves after about 800 ms are generally assumed to re ect strategic verbal memory encoding (e.g., see Bosch et al., 2001). Social evaluative contexts have been shown to considerably enhance both early negative and late positive brain potentials when identical trait adjectives are presented as personality feedback in virtual interaction setups, with larger EPN and LPP brain potentials and higher appropriateness ratings when feedback seems to come from more relevant interaction partners (e.g., see Schindler et al., 2015Schindler et al., , 2019. However, no previous studies investigated possible long-term memory consequences.
The present research aims to ll this gap within the larger context of social memory formation. We study the effects of a social-evaluative encoding context on long-term recognition memory of positive, negative, and neutral trait adjectives. We present the exact same word lists in three different encoding contexts, measuring event-related brain potentials at encoding and recognition memory one week later. Speci cally, we compare incidental encoding in a social-evaluative feedback context with intentional learning and incidental encoding via self-referential or semantic processing. We expect recognition memory and EPN and LPP amplitudes to be enhanced for emotional content and self-reference to be positively biased and speci cally examine how recognition memory and ERPs are affected by the encoding context.

Results
Behavior during the encoding task Social-feedback group (feedback acceptance). A main effect of emotional content (F ( participants were more likely to accept feedback from the more relevant ("human") sender. No interaction was found (F (2,46) = 0.48, p = .622, η p 2 = .020). = .571), with participants selecting signi cantly more positive than both neutral (p < .001) and negative (p < .001) words as self-descriptive, neutral and negative words not differing (p = 1.0). In the concreteness task, there was no effect of emotional content on concrete/abstract decisions (F (2,46) = 1.05, p = .359, η p 2 = .044).
Recognition Memory (one week later) Recognition memory data are shown in Fig. 1 and further detailed in Table 1 and Supplement A.
The social-feedback group was signi cantly more accurate than both the verbal-learning (p = .007) and the levels-of-processing group (p < .001). The latter did not differ (p = .403). Group and emotion did not interact (F (4,130) = 0.45, p = .776, η p 2 = .014), but a three-way interaction between group, encoding condition, and emotion (F (4,130) = 20.53, p < .001, η p 2 = .387) indicated that depending on the experiment, emotion effects differed between conditions (see Table 1). Within the levels-of-processing group, positive (p = .003) but not negative words (p = .113) from the self-descriptiveness task were recognized more accurately, while from the concreteness task neutral words (p = .001) were recognized best, whereas emotion effects did not depend on "sender" or "block" in the other two groups.

Encoding ERPs
Early EPN (200-300) For the early EPN, an effect of experimental group was found (F (2,69) = 9.47, p < .001, η p 2 = .215; see Fig. 2a). The social-feedback group had a larger EPN than both the verbal learning (p < .001) and the levels-of-processing group (p = .037). The latter also showed a larger EPN than the verbal-learning group (p = .029  A further four-way interaction occurred (F (4,138) = 3.48, p = .010, η p 2 = .092) which is resolved and illustrated in the Supplement -Section C.

Discussion
Humans typically acquire information in social contexts. Therefore, we investigated its role for long-term recognition memory and encoding ERPs for positive, negative, and neutral words. We compared a group for whom a social context was induced via a self-introduction followed by evaluative feedback ("social-feedback" group) with a verbal-learning and a standard levels-of-processing group. Although the social-feedback group was not aware of the recognition session one week later, this group showed superior recognition accuracy, but also an increased response bias, particularly for positive adjectives. Encoding ERPs revealed the highest EPN and LPP amplitudes in the social-feedback group. Only in the slow-wave window were ERPs most positivegoing in the verbal-learning group. ERPs from the levels-of-processing group fell in between the other two. Together, these ndings reveal a distinct effect of evaluative social context, enhancing long-term recognition memory and encoding ERPs beyond both explicit learning and standard levels of processing manipulations, while also inducing speci c biases. Therefore, encoding in the social feedback and levels-of-processing groups seems to have relied more on automatic mechanisms, whereas the verbal-learning group recruited more strategic processes-exploratory correlations accord with this interpretation (see Supplementary Materials Section B).
As expected, signi cant emotion effects on brain potentials were also observed, although they were smaller than the context-effects. Overall, we show that social context during encoding enhances long-term recognition memory beyond explicit learning or typical deeper incidental encoding tasks (self-descriptiveness or concreteness decisions). Notably, although both the social feedback and self-descriptiveness task probably recruited self-referential processing, effects in the feedback condition, where any self-reference was socially contextualized, by far exceeded those of task-driven self-referential processing alone. Although the retention interval leaves ample time for consolidation or spontaneous rehearsal, which we cannot directly assess, the immediate repetition run revealed no signi cant effects, suggesting that effects were either due to initial encoding, where corresponding ERP modulations occurred, or arose considerably later.
Our focus was on the between-groups analysis, which de-emphasizes potentially interesting effects in individual conditions. Still, in Experiment 1, some "sender" differences occurred at encoding (see Supplement C), replicating previous research (e.g., see Schindler et al., 2019). However, these had little effect on long-term recognition, suggesting that participants integrated both blocks from experiment 1 into one episode whose items were considerably more memorable than those from the other two experiments. Interactions in the levelsof-processing group indicate that encoding task modulates emotion effects in long-term recognition memory, only self-descriptiveness and not concreteness decisions resulting in an emotion effect on recognition accuracy, in line with the suggestion of interacting memory effects of self and emotion (see Gutchess & Kensinger, 2018).
In sum, although we used identical stimuli and presentation parameters across three experiments, the psychological encoding contexts elicited pronounced between-group differences in recognition memory and ERPs: The social-feedback context enhanced recognition memory beyond both explicit verbal-learning and selfdescriptiveness/concreteness judgments. It also induced the largest EPN and LPP amplitudes, followed by levels-of-processing and verbal learning. The most positive-going slow wave amplitudes in the verbal learning group suggest that this group might have relied particularly on strategic memory processes. Our ndings specify some important social factors in human long-term memory. They resonate with the fact that humans typically acquire information in social contexts (e.g., see Light & Perret-Clermont, 1991), social evaluation being a particularly salient factor in human life (e.g., see Moor et al., 2010). Finally, these results have implications for memory formation in educational settings or memory assessment in legal contexts, where the extent to which individuals were exposed to an evaluative social context will affect their memory for elements of the initial episode.

Participants
Three groups of twenty-four participants each were recruited at Bielefeld University. Participants provided written informed consent and received 10 Euros per hour for participation. The Ethics Committee at Bielefeld University approved the study and the study was performed in accordance with the regulations of the Declaration of Helsinki. All subjects were right-handed, had normal or corrected-to-normal vision, and were free from a self-reported neurologic or psychiatric disorder. They were tested twice with a lag of about one week (T1-T2 difference M = 7.13 days, Min = 6 days, Max = 9 days, see Table 1).

Stimuli
The stimulus set had been rated by 22 other students on nine-point Likert-type scales in terms of valence and arousal using the Self-Assessment Manikin (Bradley & Lang, 1994) and a similarly constructed concreteness scale. The selected 270 adjectives (90 negative, 90 neutral, 90 positive) were assigned to three separate lists. Neutral adjectives were allowed to deviate from emotional adjectives on both arousal and concreteness since truly neutral trait adjectives are rare in an interpersonal evaluative contexts. The three lists did not differ in relevant emotional and lexical properties, and the list-condition assignment was counterbalanced (see Table 3).

Procedure
An outline of the experimental setup for the three groups is presented in Fig. 4.
Social-feedback group. Upon arrival, participants were told that another, unknown person, would evaluate them based on their self-presentation. In a second condition, a randomly operating computer algorithm was supposedly going to give them feedback. All subjects underwent both "sender" conditions. Condition sequence and word lists were counterbalanced. Importantly, random feedback was presented in both conditions. Participants were instructed to brie y describe themselves in a structured interview in front of a camera. They were informed that the video of their self-description would be presented to another person. During EEG preparation, participants completed a demographic questionnaire and the BDI (Hautzinger et al., 2006) and STAI  to characterize the sample. A research assistant left the testing room 15 minutes ahead of the ctitious feedback to ensure face validity, guiding an "unknown person" to a laboratory room next to the testing room.
Stimuli were presented by software described as "Interactional Behavioral Systems", supposedly allowing instant online communication. Participants were told that the unknown other person would select adjectives describing the participant, which would be subsequently presented. In the other condition, feedback was denoted as random computer feedback. Feedback-adjectives were presented for 1500 ms, after which the participants had to indicate via button-press whether or not they agree with this evaluation. After the response, a xation cross was presented for 2000 to 3500 ms. For each sender, 30 negative, 30 neutral, and 30 positive adjectives (one of the three lists, see Table 2) were presented twice. The desktop environment and stimulus presentation were created using Presentation (www.neurobehavioralsystems.com). Afterwards, participants were debriefed that no social evaluation had taken place and presented a passive viewing run of all 60 negative, 60 neutral, and 60 positive adjectives again to test for any immediate post-processing effects of the manipulation.
Participants were informed that an unrelated test would be conducted in a second session one week later, where an unexpected recognition test took place. All adjectives from the rst session and a set of new items were shown for 1500 ms each, followed by an old/new decision and a xation cross for 1500 to 2000 ms.
Verbal-learning group. The very same stimuli and presentation parameters were used. In order to keep the visual input constant, the same background presentation setup (Interactive Behavioral Systems) was used but never referred to. The self-introduction phase was omitted, and no social signi cance was assigned to the stimulation. Instead, participants were instructed to memorize all adjectives for recognition testing in the next session one week later. Adjectives were presented for 1500 ms, followed by a variable xation cross presented for 2000 to 3500 ms. To mimic the "social-feedback" experiment, two counterbalanced conditions were presented (block "A" and block "B"), using the same material and number of repetitions (see Fig. 4b, Fig. 4d, Table 2). Condition order was counterbalanced. A passive viewing run nished the experimental session.
Levels-of-processing group. The same materials and presentation parameters as in the previous two experiments were used. As in experiment 1, active word evaluation was instructed, but no social context was created. Participants were told to either decide via button press on the self-descriptiveness of the presented words or, in condition two, perform a binary concreteness judgment on the words, classifying them as either abstract or concrete. This closely re ected the encoding trial structure for the social-feedback group. Adjectives were presented for 1500 ms, after which the binary decision was requested. Material and experimental parameters were the same as in the social feedback group (see Fig. 4c, Fig. 4d, Table 2). Again, the session was nished with a passive viewing run. EEG recording and analyses EEG was recorded from 128 BioSemi active electrodes (www.biosemi.com) at 1024Hz. Two separate electrodes were used as ground electrodes, a Common Mode Sense active electrode (CMS) and a Driven Right Leg passive electrode (DLR), which form a feedback loop to measure the average potential close to the reference in the A/Dbox (www.biosemi.com/faq/cms&drl.htm). Four additional electrodes (EOG) measured horizontal and vertical eye movement.
Pre-processing and statistical analyses were performed using BESA (www.besa.de) and EMEGS (Peyk et al., 2011). O ine, data was re-referenced to an average reference and ltered with a high-pass forward lter of 0.16 (6 db/oct) and a 30 Hz low-pass zero-phase lter (24 db/oct). Filtered data were segmented from 200 ms before word onset until 1500 ms after stimulus presentation. The 200 ms before stimulus onset were used for baseline correction. Eye-movements were corrected using the automatic correction method implemented in BESA (Ille et al., 2002). Remaining artifacts were rejected based on an absolute threshold (< 120 µV), signal gradient (< 75 µV/∂T), and low signal (i.e., the SD of the gradient, > 0.01 µV/∂T). Noisy EEG sensors were interpolated using a spline interpolation procedure.

Statistical analyses
For all data, we calculated three (group: social-feedback, verbal-learning, levels-of-processing) by two (condition: human sender/block A /self-reference, computer sender/block B /concreteness) by three (emotion: positive, negative, neutral) repeated-measures ANOVAs. For memory data, the discrimination index (P r = hits-false alarms), and the response bias (B r = false alarms/(1-P r )) were calculated according to Snodgrass  EPN effects of A) experimental group and B) emotion. Scalp topographies depict the mean amplitude differences for the respective interval. ERPs show the time course averaged from highlighted sensors. Betweengroup difference potentials contain 95% bootstrap con dence intervals. Between-emotion difference-potentials contain 95% bootstrap con dence intervals of intra-individual differences. For bar charts, error bars show 95% con dence intervals.
Page 18/19 Figure 3 LPP and slow-wave effects of A) experimental group and B) emotion. Scalp topographies depict the mean amplitude differences for the respective interval. ERPs show the time course from averaged highlighted sensors. Respective difference potentials contain 95% bootstrap con dence intervals of the group differences. Emotion difference waves contain 95% bootstrap con dence intervals of intra-individual differences. For bar charts, error bars show 95% con dence intervals.