---

Directions

(from Semester Projects: Descriptions)

For this project, you will write a 3-page minimum paper. To do this you will:

• Work with your summary and the main points of the article. It is important that you close the article while you are doing this so you can be sure you are not plagiarizing (changing a couple words or switching phrases around still constitutes plagiarism).
• Write down a 3 paragraph description of the article including why the authors performed this study, what tools and participants they used, and what they found.
• Compare the information you read to the topic information found in your textbook along the lines of whether this is the same information, or different. Did the article extend on the stuff that is found in your text? Does it contradict it in some way? Can you reconcile the two? Are there any questions that are still left unanswered? What further research would you do if you were the author of the article? What other questions would you like answers to?
• Create APA citations for your article and the pages you referred to in your textbook. To review how to do APA citations, please refer to the resources below.

This is the article

Title:

Forgetting unrelated episodic memories through suppression-induced amnesia.

Authors:

Zhu, Zijian. School of Psychology, Shaanxi Normal University, Beijing, China, zhuzijian0203@snnu.edu.cn
Wang, Yingying. Academy for Advanced Interdisciplinary Studies, Peking University, China, zhuzijian0203@snnu.edu.cn

Address:

Zhu, Zijian, School of Psychology, Shaanxi Normal University, 199 Changannan Road, Xi’an, China, 710062, zhuzijian0203@snnu.edu.cn

Source:

Journal of Experimental Psychology: General, Vol 150(3), Mar, 2021. pp. 401-413.

NLM Title Abbreviation:

J Exp Psychol Gen

Publisher:

US : American Psychological Association

Other Journal Titles:

Journal of Experimental Psychology

Other Publishers:

US : Psychological Review Company

ISSN:

0096-3445 (Print)
1939-2222 (Electronic)

Language:

English

Keywords:

forgetting, retrieval suppression, TNT paradigm, amnesic shadow, independent cue

Abstract:

Cognitively suppressing the retrieval of an unwanted memory causes its forgetting and, in the meantime, disrupts hippocampal functions. The present study investigated whether retrieval suppression induces virtual amnesia, which disturbs any existing memories that are reactivated in the temporal vicinity but are otherwise unrelated to the targets of suppression. Participants performed retrieval suppression on a set of memories while cues of an unrelated set of memories were briefly presented near in time to the suppression trials. Results showed that retrieval suppression impaired the retrieval of both the directly suppressed content and the reactivated unrelated memory. This amnesic shadow functioned in both the forward and backward temporal directions, and its forgetting effect was revealed by independent cues that were not presented in the shadow. Remarkably, a negative memory could be impaired simply by presenting it between the suppression episodes of an unrelated neutral memory. These findings provide support for systemic influence of retrieval suppression on hippocampal functions and offer a way to disrupt existing episodic memory strategically. (PsycInfo Database Record (c) 2021 APA, all rights reserved)

Document Type:

Journal Article

Subjects:

*Amnesia; *Episodic Memory; *Forgetting; *Hippocampus; *Human Information Storage; Cues; Suppression (Defense Mechanism); Test Construction

PsycInfo Classification:

Neurological Disorders & Brain Damage (3297)

Population:

Human
Male
Female

Location:

China

Age Group:

Adolescence (13-17 yrs)
Adulthood (18 yrs & older)
Young Adulthood (18-29 yrs)

Tests & Measures:

TNT Pair Task
Think/No-Think Task–Adapted

Grant Sponsorship:

Sponsor: National Natural Science Foundation of China, China
Grant Number: 31421003
Recipients: No recipient indicated

Sponsor: China Postdoctoral Science Foundation, China
Grant Number: 2019M663621
Recipients: No recipient indicated

Methodology:

Empirical Study; Quantitative Study

Supplemental Data:

Tables and Figures Internet

Format Covered:

Electronic

Publication Type:

Journal; Peer Reviewed Journal

Publication History:

First Posted: Aug 13, 2020; Accepted: Mar 20, 2020; Revised: Mar 15, 2020; First Submitted: Aug 26, 2019

Release Date:

20210311

Correction Date:

20210311

Copyright:

American Psychological Association. 2020

Digital Object Identifier:

http://dx.doi.org.ezp.pasadena.edu/10.1037/xge0000782; http://dx.doi.org.ezp.pasadena.edu/10.1037/xge0000782.supp(Supplemental)

PMID:

32790459

PsycARTICLES Identifier:

xge-150-3-401

Accession Number:

2020-59521-001

Persistent link to this record (Permalink):

https://login.ezp.pasadena.edu/login?url=https://search.ebscohost.com/login.aspx?direct=true&db=pdh&AN=2020-59521-001&site=ehost-live&scope=site

Cut and Paste:

Forgetting” class=”redactor-linkify-object”>https://login.ezp.pasadena.edu/login?url=https://s… unrelated episodic memories through suppression-induced amnesia.

Database:

APA PsycArticles

---

Forgetting Unrelated Episodic Memories Through Suppression-Induced Amnesia

By: Zijian Zhu
School of Psychology, Shaanxi Normal University;
Yingying Wang
Academy for Advanced Interdisciplinary Studies, Peking University;

Acknowledgement: This work was supported by the National Natural Science Foundation of China (31421003) and a China Postdoctoral Science Foundation Grant (2019M663621). Zijian Zhu and Yingying Wang developed the study concept and contributed to the study design together. Zijian Zhu performed the programming and data collection. Yingying Wang performed the data analysis. Both authors wrote the article and approved the final version of the manuscript for submission. The authors declare no competing financial interests. The experiments in this article were not formally preregistered. The data are posted on the Open Science Framework at https://osf.io/gxbfc/.

Following a traumatic experience, memories of the event often intrude. To deal with this, people often try to exclude the unwanted memory from awareness. Such attempts not only protect people from being hurt by the ongoing intrusions but also diminish the memory in the long run (Anderson & Hanslmayr, 2014; Depue, Banich, & Curran, 2006; Gagnepain, Henson, & Anderson, 2014). Research on motivated forgetting indicates that this process, known as retrieval suppression, is mediated by the downregulation of hippocampal activity (Anderson et al., 2004; Benoit & Anderson, 2012; Gagnepain, Hulbert, & Anderson, 2017) and is related to a higher concentration of inhibitory neurotransmitters in the hippocampus in the resting state (Schmitz, Correia, Ferreira, Prescot, & Anderson, 2017). Recent research found that this downregulation causes a “virtual lesion’” on the hippocampus, which disrupts the encoding of new unrelated experiences occurring near in time to retrieval suppression efforts (Hulbert, Henson, & Anderson, 2016; Hulbert, Hirschstein, Brontë, & Broughton, 2018). Such systemic influence on hippocampus-dependent processes mimics the effect of hippocampal amnesia and is referred to as the amnesic shadow. In this sense, it should be possible to forget any hippocampally dependent memories merely by exposing it to the amnesic shadow.

Here, we further hypothesized that the amnesic shadow not only influences the encoding of novel information but also is capable of influencing memories that have been encoded into the hippocampus. To test this hypothesis, the present study measured the influence of retrieval or suppression on the later accessibility of previously acquired memories that happened to be cued in the temporal vicinity but were otherwise unrelated to the targets of suppression. To induce retrieval suppression, we used an adapted think/no-think (TNT) procedure (Anderson & Green, 2001). In the TNT procedure, people were presented with a reminder of a past event and cued either to retrieve the associated memory (think trials) or to suppress its retrieval (no-think trials). At the behavioral level, the target item that has been suppressed is impaired in memory (Anderson & Green, 2001), as well as in its affective (Gagnepain et al., 2017), perceptual (Gagnepain et al., 2014; Kim & Yi, 2013), and conceptual (Wang, Luppi, Fawcett, & Anderson, 2019) representations. Recently, Hulbert and colleagues (2016) modified the TNT procedure to a hippocampal modulation (HM) paradigm by innovatively inserting innocent bystanders, fresh information that had never been studied, between think or no-think trials. Their findings that retrieval suppression disrupted memory of innocent bystanders presented near in time to suppression provided the first evidence that retrieval suppression induces an amnesic shadow that harms the encoding function of the hippocampus.

In the present study, we adapted the HM paradigm to explore the amnesic shadow effect on an unrelated existing memory. We trained participants to initiate retrieval suppression on a set of memories using the TNT task (Anderson & Green, 2001). Different from Hulbert et al. (2016), innocent bystanders embedded between think or no-think trials were cues of a different and unrelated set of existing memories. The bystander cues were presented to the participants to reactivate the unrelated memory in order to engage the hippocampus. Buffer trials were inserted before and after bystanders to ensure that the same task always preceded/followed bystanders. Unbeknownst to participants, targets for the bystanders would be tested after the TNT phase. Critically, we employed a double-cue/one-target procedure for the bystander memories (Zhu et al., 2016; Zhu, Wang, Jia, & Wu, 2019). That is, the target memory of each bystander cue was tested with both the bystander cue and an independent cue, the latter of which was not presented in amnesic shadow and thus provided a cleaner measure of target memory change. Considering that the amnesic shadow effect is specific to hippocampally dependent memory processes (Hulbert et al., 2016; Hulbert et al., 2018), we used a cued-recall task with both trained and independent cues to test the memory changes on the bystander memories.

We predicted that retrieval suppression would impair an unrelated episodic memory once it was reactivated within its amnesic shadow. Due to the nature of retrieval suppression, some characteristics about the shadow effect were expected. First, retrieval suppression would work on the to-be-suppressed memory rather than particular cue-target associations (Anderson & Green, 2001; Wang, Cao, Zhu, Cai, & Wu, 2015); thus, disruptions on the bystander target should be on the target item itself and persist when retrieved by independent cues. Second, because the amnesic shadow occurs due to inhibited hippocampal activity by retrieval suppression (Hulbert et al., 2016), the degree of shadow-induced forgetting should be predicted by the degree of suppression-induced forgetting. Third, although emotional memories might rely on some different brain regions than neutral memories, due to their common reliance on the hippocampus (Phelps, 2004), the amnesic shadow induced by the suppression of a neutral memory might be able to disrupt an unrelated emotional memory. Here, we provided evidence for the above predictions and confirmed that the amnesic shadow could be exploited to disrupt existing memory. This mechanism provides a way to treat unwanted episodic memories without the need to directly suppress them.

Experiment 1: Reduction in Episodic Memories Due to Amnesic Shadow

We first tested whether retrieval suppression could influence the retrieval performance of an unrelated memory that was reactivated close in time to retrieval suppression. We used the classical TNT paradigm for retrieval suppression (Anderson et al., 2004) and inserted bystander cues of an unrelated memory between two TNT trials. Influence on the bystander memory was examined with cues that were presented during retrieval suppression (trained cues) and with cues that were independent of reactivation or retrieval suppression (independent cues).

Method

Participants

Thirty native Chinese speakers (age range 17–21 years, 11 male) participated in exchange for monetary compensation. No statistical methods were used to predetermine sample size for Experiment 1, but our sample size was chosen to be similar to those reported in previous publications from the lab and in the literature that focused on suppression-induced forgetting (e.g., Gagnepain et al., 2017). All participants had normal or corrected-to-normal vision. They had no reported history of head injury, neurological disease, or learning disability and no red/green color-blindness. Participants provided written consent before participation. Experimental procedures were reviewed and approved by the Human Subject Review Committee of the Shaanxi Normal University.

Materials

The stimuli for the main experiment contained 30 critical TNT pairs and 30 bystander pairs. The nature of the TNT pairs followed previous work (Zhu et al., 2016). The constituent members of each TNT pair were two 2-character Chinese words (e.g., “PORT-SURFACE”) that were emotionally neutral and semantically unassociated with each other. A special design of double-cue/one-target was used for bystander pairs (Wang et al., 2015; Zhu et al., 2016; Zhu et al., 2019). That is, each target picture was paired with two different cue words, which formed two series of bystander pairs, in the form of A-X and B-X. The two series were studied and trained separately during the experiment. Critically, cues from one series would be embedded between TNT trials (trained cues), while cues from the other series would not appear in the TNT phase (independent cues). The three members of each bystander pair were emotionally neutral and semantically unassociated with each other. Items from different TNT pairs were also semantically unassociated with items from the bystander pairs.

Critical TNT pairs were divided into three subsets (10/group). For each participant, cues from one subset would appear in the think task, cues from a second subset would appear in the no-think task, and cues from the remaining subset did not reappear during the TNT phase and served as the control group. Critical bystanders were also divided into three subsets (10/group): One subset would be presented between two think trials, one subset would be presented between two no-think trials, and the remaining subset would not appear during the TNT phase (served as the baseline). Each group of bystanders was always used for a fixed group of TNT pairs. The arrangement of experimental conditions for the three TNT (along with its bystander) groups was counterbalanced across participants.

Procedure

The modified HM paradigm consisted of three phases: the study phase, the TNT phase, and the test phase. The procedure is shown in Figure 1.

Figure 1. The paradigm and predictions for cognitively induced amnesia. Participants first study two series of word-object pairs (in the form of double-cue/one-target) and one series of word-word pairs until they self-report that they could correctly memorize each pair (Phase 1). Participants then perform trials on the third series, which require that they either retrieve (green) or suppress (red) retrieval of the target word in each pair, given the first word (Phase 2). Inserted between these trials are “bystander” words, which are cues from one word-object pair series. Participants judge whether they recognize these cues as old or new. Even/odd buffer judgments are performed before and after bystanders to match the immediate task context across retrieve and suppress trials. We hypothesized that surrounding bystanders with suppression trials affects later memory for bystander targets, causing an amnesic shadow. This is assessed in Phase 3, in which the trained bystander cue and the independent cue that was associated with the bystander target but has not been presented during Phase 2 are used to retrieve the bystander target separately. Memory impairment is predicted for bystander targets under the retrieval of both cues. The cued-recall test assesses the memorability of the learned associations from across the three series, and participants respond by verbally reporting the target object or word.

Study phase

Participants studied three series of cue-target pairs, including two bystander pair series and one TNT pair series. The three series were always studied in a fixed order: first the trained-cue bystander pairs, then the independent-cue bystander pairs, and finally the TNT pairs. This arrangement ensures comparative memory strength of pairs within the same series and avoids memory integration of pairs across different series.

First, the trained-cue bystander pair series was studied. Thirty word-picture pairs were presented to the participants sequentially, each for 3 s (interstimulus interval = 1 s). Test-feedback cycles followed, in which each cue word was presented alone for up to 5 s and participants judged whether they could retrieve the corresponding target picture or not by pressing one of two keys. Upon key pressing or when the response window expired, the target picture was given at the right side of the cue word for up to 5 s. During this period, participants judged whether they had retrieved the target picture correctly or not by pressing one of two keys. Test-feedback cycles continued on pairs that could not be retrieved until participants indicated that they had successfully recalled 100% of the cue target pairs. Next, the independent-cue bystander pair series was studied, using the same procedure as above. Participants were informed that the target pictures in the first series would be studied again and paired with different cue words. They were explicitly instructed to study the new series without thinking of the first series and to not integrate the three items (i.e., two cue words and one common target picture) into a single memory. Finally, the 30 TNT word-word pairs were studied, using the same procedure as for the bystander pairs.

TNT phase

Cue words from two subsets of the TNT pairs appeared in the TNT phase. Each of the 20 cue words was presented for 6 s either in green (think trial) or in red (no-think trial) font. For think trials, participants were instructed to recall the associated target picture as vividly as possible and to rehearse it silently for the 6-s duration of the trial. For no-think trials, participants were required to avoid thinking about the associated target picture while sustaining their attention on the cue word for the 6-s trial. The standard direct suppression instructions were used (Benoit & Anderson, 2012; Wang et al., 2019): Participants were asked not to replace the target picture with any other diversionary thoughts or images but simply to stop themselves from retrieving the target. Cues from the third subset did not appear in the think or no-think task and served as the control for the TNT manipulation.

To test whether voluntary suppression on one memory affected a different memory, we presented bystander cues between the TNT trials. Notably, only cues from the first bystander series were presented during the TNT phase; they were thus called the trained cues. Cues from the other bystander series did not appear during the TNT phase; they were independent of the TNT manipulation and were thus called the independent cues. The 10 trained cues from one bystander subset were presented between two think trials, serving as probes for retrieval shadow; the 10 trained cues from a second bystander subset were presented between two no-think trials, serving as probes for suppression shadow; the remaining 10 trained cues, from the third bystander subset, did not reappear in the TNT phase and thus served as the baseline for the shadow effect. Each trained bystander cue was presented for 2 s, during which participants internally judged whether they had studied this cue word or not. Notably, no novel words were included as bystanders. Participants were not instructed to retrieve or suppress the target picture for the bystander cue.

Before and after each bystander, we inserted “buffer” intervals during which participants viewed a series of 2–4 digits and decided whether each digit was odd or even. The buffer task ensured that the same task was performed before and after each bystander cue so that a similar transition was made from one task to another (see Hulbert et al., 2016). Therefore, a whole trial was composed of one TNT task, a buffer task, a bystander cue recognition task, another buffer task, and another TNT task. The TNT phase was divided into three sessions, each with four repetitions of the 20 TNT items; because one bystander cue was embedded between two TNT items, each session contained two repetitions of the 20 trained bystander cues. A given bystander was randomly embedded between two TNT items in each repetition.

Test phase

Following the TNT phase, we tested the suppression-induced forgetting effect on the critical TNT pairs and its shadow effect on bystanders. Memory related to the bystanders was tested first. Because the independent cues were not presented during the TNT phase, they should provide a cleaner measurement of target memory change than the trained cues. We therefore tested participants’ memory for the bystander targets using a block design. A cued-recall test was first given to the independent bystander cues. In each trial, the independent cue word was presented on the screen, instructing participants to verbally report its associated target picture within 5 s. After being tested with all independent cues, participants were tested on their memory for the bystander targets with the trained cues, using the same procedure as for the independent-cue test. Finally, the TNT word pairs were assessed via a cued-recall task, with each cue being presented on the screen for 5 s and participants verbally reporting the associated target word. The order of stimuli corresponding to the think, no-think, and control conditions in each word-word/picture series was counterbalanced within and across participants.

Results

Effects of retrieval suppression on the TNT associates

Data from all experiments are available on the Open Science Framework (https://osf.io/gxbfc/). We first tested the suppression-induced forgetting (SIF) effect on the TNT pairs. The percentages of correctly recalled target words (Figure 2A and Table S2) were submitted to a one-factor repeated measures analysis of variance (ANOVA) with three levels (i.e., think, no-think, vs. control). Results showed a significant main effect of suppression status, F(2, 58) = 5.64, p = .006, ?p2 = 0.16. Comparisons between the no-think condition and the control condition revealed significant suppression-induced forgetting, t(29) = ?2.36, p = .025, Cohen’s d = 0.40, significant after Bonferroni correction, replicating the classical findings that retrieval suppression impaired memory of the target items. However, no memory improvement was detected for the think condition, think vs. control: t(29) = 0.78, p = .445, Cohen’s d = 0.11. This result verified that the manipulation of retrieval suppression was successful.

Figure 2. Results for Experiment 1. Panel A: Percentage of targets recalled for the think/no-think (TNT) pairs. Decreased memory performance was found in the no-think condition relative to the control condition, showing the suppression-induced forgetting effect. Panel B: Percentage of targets recalled for the two series of bystander pairs. Significant memory impairment was found in the no-think condition when compared with baseline condition, under independent-cue retrieval. Panel C: Significant correlation between the suppression-induced forgetting effect and the overall amnesic shadow effect of the trained- and independent-cue retrieval. * p < .05. ** p < .01. (Two-tailed t test.) Error bars indicate standard error of the mean.

Shadow effect of retrieval suppression on bystander associates

We then tested the influence of retrieval suppression on the target memory of the bystander cues. A 2 (Cue Type: trained cue vs. independent cue) × 3 (Suppression Status: think, no-think, and baseline) repeated measures ANOVA was performed on the recall accuracy of the bystander targets (Figure 2B; see Table S1 for detailed results for each condition). Results showed a significant main effect of suppression status, F(2, 58) = 8.11, p = .001, ?p2 = 0.22. Consistent with the pattern of the SIF effect, the overall recall accuracy for bystanders embedded between two no-think trials decreased significantly when compared with the baseline (45.33% vs. 51.83%), t(29) = ?3.24, pcorrected = .009, and think (45.33% vs. 53.67%), t(29) = 3.54, pcorrected = .004, conditions; yet no memory improvement was observed for bystanders inserted between two think trials (think vs. baseline: 53.67% vs. 51.83%), t(29) = 0.85, pcorrected = 1.00. Meanwhile, the main effect of cue type was significant, F(1, 29) = 9.63, p = .004, ?p2 = 0.25, with better overall performance in the independent- than trained-cue condition, which might be due to the recency effect and less retroactive interference as the independent-cue group was always studied after the trained-cue group.

Results showed no significant interaction effect of the two factors, F(2, 58) = 0.76, p = .47, ?p2 = 0.03, suggesting similar shadow effect in the two cue types. However, considering that the trained and independent cues received different treatments—the trained cues were presented in the suppression shadow while the independent cues were not—we tested the effect of each cue type separately. We found a significant main effect for independent-cue retrieval, F(2, 58) = 5.68, p = .006, ?p2 = 0.16, but not for trained-cue retrieval, F(2, 58) = 1.29, p = .282, ?p2 = 0.04. Further analysis on the simple effect of the independent-cue group revealed significant memory impairment for no-think bystanders when compared with the baseline bystanders, t(29) = ?2.94, p = .006, Cohen’s d = 0.46. No memory improvement was found for think bystanders, think vs. baseline: t(29) = 0.50, p = .620, Cohen’s d = 0.10. Taken together, retrieval suppression does not only impair the directly suppressed memory but also disrupts other memory via amnesic shadow.

An important concern is that the shadow-induced forgetting on the bystander items might simply be due to distractions following the suppression task. It was necessary to examine whether the shadow-induced forgetting on the bystanders was caused by a different amount of distractions following suppression and retrieval trials. To do this, we analyzed the odd/even judgment task performance on the first and second buffer digits, which were presented right after the think and no-think tasks and before the bystander cues (Tables S3 and S4). The buffer task was not influenced by the think or no-think conditions as the performance on it differed neither in speed, first buffer: t(29) = ?1.49, p = .148, Cohen’s d = 0.27; second buffer: t(29) = ?0.91, p = .372, Cohen’s d = ?0.17, nor accuracy, first buffer: t(29) = 1.48, p = .149, Cohen’s d = 0.27; second buffer: t(29) = 0.10, p = .918, Cohen’s d = 0.02, as a function of whether it was performed after a retrieval or suppression trial. There was a weak trend of slower response after the no-think than after the think task