Directed Forgetting: Active Or Passive?

ObjectiveForgetting is often viewed as memory failure. In many cognitive models of memory, forgetting is treated as a failure to encode, maintain, or retrieve information; and it has been hypothesized to result from various processes including passive decay, interference, interrupted consolidation, and retrieval failure. However, Ribot, a memory psychologist, consider that forgetting may be not a disease; the appropriate forgetting is conducive to health. Therefore, efficient memory functioning involves not only the successful remembering of previously presented or learned material but also the successful forgetting of irrelevant, invalid, or out-of-date information. This is called the intentional forgetting or directed forgetting. The research paradigm of directed forgetting includes list-cueing method and item-cueing method. The difference between them is mainly embodied in learning phase. An explicit cue will be provided to indicate which presented items are to-be-remembered (TBR) and which are to-be-forgotten (TBF). In the list-method, the cue is presented after an entire list of items. In contrast in the item-method, each study item is accompanied by a cue. The requirements of the two methods in the test phase are the same. Subjects need free recall the studied item or do the task of old/new recognition. TBR performance is always better than that of TBF, which is referred to as the directed forgetting effect.For the item-method, the predominant explanation of directed forgetting effect has emphasized selective rehearsal theory and attention inhibition theory. Bjork et al. purposed that studied items be initially maintained in short-term memory with rote rehearsal until the presentation of Remember/Forget cues. Rote rehearsal is terminated in response to a Forget cue, whereas elaborative rehearsal processes are allocated to the TBR items in response to a Remember cue. By contrast, Zacks et al. suggested that following the Forget cue, the TBF items become irrelevant information, subsequently inhibited actively and prevented from entering into working memory. So that TBF performance is always worse than that of TBF. Paller et al. demonstrated ERPs in the time window of 200-600 ms elicited by TBR cues were more positive than that of TBF cues at the posterior scalp areas. They argued that these ERPs differences could have been caused by differential rehearsal of the words. Paz-Caballero et al. reported ERPs of TBR cues were more positive than those of TBF cues at the anterior scalp areas, this difference was related to inhibition mechanism. The research of functional magnetic resonance imaging of Wylie et al. supports the view that active cognitive processes are engaged by an TBF instruction, they found that greater activity for TBR cue was found in middle frontal gyrus, middle cingulate gyrus, middle temporal gyrus, and parahippocampal gyrus. The evidence also comes from patients with frontal damage, Conway et al. have reported that patients with right frontal lobe damage have no directed forgetting effect, this supported frontal lobe was involved in inhibition process.In previous researches of directed forgetting, TBR item and TBF item have been compared directly without the control group of neither remembering or forgetting. This caused the difficulty of distinguishing the effects between directed (active) forgetting and natural (passive) forgetting. In this research, we add a control group consisting of the items of only to-be-viewed (TBV) to the classic paradigm and study the neural mechanism of directed forgetting by event-related potentials (ERP).MethodsSixteen college students participated in the experiment. Data of 4 subjects was removed because of large artifacts. The rest 12 (7 females) subjects aged between 23 and 26 years (24.25±0.75 years). All is right-handed without neurological or mental disorders.720 double-words with medium word frequency were selected from the book "Modern Chinese Frequency Dictionary", the stroke ranged from 14 to 22 (17.89±1.75). The words with obvious emotional color and the same pronunciation were excluded.360 words randomly were selected as the studied items (old items) were randomly divided into TBR, TBF and TBV groups. The rest 360 words as new items presented in test phase. Different colors (red, green, blue) circular picture indicated different clues requirements respectively.The item-method has two phases. (1) Study phase:Each word presented 800 ms followed by a cue. And each cue presented 400 ms. The interval between two words was 2000 ms. Subjects needed to do a simple button task:pressed the right, left, any button when either TBR, or TBF or TBV cue presented respectively. (2) Test phase: After 4 minutes break, one will get into test phase. Both the old and the new words were displayed. Subjects were instructed to memorize all the old words in the old-new recognition test. Every word presented 800 ms, the interval between two words was 1800 ms. Words were arranged in a randomized order (no more than 3 consecutive presentations of the same word category).EEG was recorded using an ERP system developed in our lab. It was sampled from 19 electrodes mounted in an elastic cap according to international 10-20 system referenced to both earlobes. Prefrontal midline site was grounding. The ERP epochs were extracted off-line and included pre-stimulus activity of 100 ms and post-stimulus activity of 1000 ms for the cues (study phase) and the word (test phase). Baseline was from -100 ms to 0 ms. Artifacts above 70 V were eliminated automatically.One-way ANOVA of repeated-measurement and paired comparisons were carried out for the task factor. Statistical parametric mapping F-values [SPM(F)] and t-values [SPM(t)] were gained from interpolation calculated by each channel's F-values or t-values respectively. The significant level was 0.05. Results1 Behavior data:The accuracy effect was significant [F(2,22)=48.22, P=0.000]. TBR group (67.28±9.45%) were significantly superior to the other two groups (P=0.000). No significant behavioral effect was found between the TBF group and the TBV group.The reaction time effect was significant [F(1.23,13.53)=4.770, P=0.041]. TBR group (862.48±111.14 ms) were significantly superior to the other two groups (P=0.000). No significant behavioral effect was found between the TBF group and the TBV group.2 ERP and SPM dataThe study phase:The spatiotemporal pattern of statistical parametric mapping suggested the task effect occur in bilateral frontal poles (200-300 ms), the parietal regions (300-400 ms), the left fronto-parietal and the right occipital region (450-700 ms). The effects of the TBR cue were presented in bilateral frontal poles (200-300 ms) and the parietal- occipital regions (300-400 ms), left frontal and the right occipital region (500-600 ms). ERP effect occurring in the right frontal elicited by the TBF cue had the significant trend (350-600 ms) and reached to the significant level together with the left parietal effect (500-525 ms).The test phase:the difference between the hit item (old word) of three kinds of cues and the item of corrected rejection (new word) were analyzed respectively. The old/new effects of the TBR item were presented in left fronto-temporal and right frontal pole (375-450 ms), left temporo-parietal and right frontal pole (475-625 ms), central fronto-parietal regions (700-775 ms). The old/new effects of the TBF item occurred in left fronto-temporal and right frontal pole (300-450 ms), left temporo-parietal (450-500 ms), central fronto-parietal regions (650-750 ms). The old/new effects of the TBV item were seen in left fronto-temporal and right frontal pole (325-475 ms) and central fronto-parietal regions (600-750 ms).ConclusionOur research has the similar behavioral performance result to the previous researches:The behavioral performance of TBR group was significantly superior to the other two groups. In the study phase, the effect of bilateral frontal pole (200-300 ms) and left parietal, bilateral occipital region (300-400 ms) suggest the activation of the two selective attention subsystem respectively:top-down and bottom-up. The effect of left fronto-parietal and right occipital (450-700 ms) reflect enhanced rehearsal. Therefore, the two theories of directed forgetting are not contradictory. Attention inhibition theory explains the early process, while the selective rehearsal theory explains the later.In the test phase, our results show the recognition of the TBV item is based on familiarity. The retrieve of the TBR is base on recollection, and the TBF is between the two conditions. The depth of processing of the word corresponding to TBR, TBF and TBV cues decreased gradually.Our results suggest the TBR instruction can improve the memory performance of words by enhancing the rehearsal function of left hemisphere. In comparison with the TBV instruction, the TBF has the similar behavioral performance though the activation of the right frontal lobe and the left parietal revealed the efforts of active inhibition.