Think/no-think

The think/no-think paradigm models the kind of forgetting that occurs when we intentionally suppress or avoid remembering in response to strong reminders of a particular event (Anderson & Green, 2001; Levy & Anderson, 2002). Imagine a man who associates a particular song with an unhappy love affair. Each time he hears the song, he tries to avoid thinking of the failed relationship, and over time, he remembers less.

In this paradigm, participants learn a series of cue-target pairs (e.g. “ambition- ballet”, “ordeal-roach”, “fuss-poodle”). Subsequently, in the think/no-think phase, participants are presented with some of the cue words again. In this phase, for half the cues (e.g. “ambition”) participants recall the associated target, and for half the cues (e.g. “ordeal”) participants avoid letting the target come into their mind (see Figure 6.3). On a final cued recall test, Anderson and Green (2001) found that participants recalled fewer targets that they suppressed (e.g. “roach”) than baseline targets (items that did not appear at all in the think/no-think phase, e.g. “poodle”). They concluded that this procedure might model Freudian repression, by showing that deliberate attempts to suppress may result in forgetting (Anderson & Levy, 2002; but see Kihlstrom, 2002; see also Erdelyi, 2006; Kihlstrom, 2006). TNT has been argued to impair both memory accessibility and availability.This is supported by evidence that participants show poorer recall for suppressed items even when recall is cued with a novel cue (e.g.,“insect” for “roach”; Anderson & Green, 2001).

While some researchers have replicated the forgetting effect following suppression in this paradigm (for review, see Levy & Anderson, 2008), others have had difficulty. For example, across three attempted replications with increasingly precise adherence to Anderson and Green’s (2001) original procedure, Bulevich, Roediger, Balota, and Butler (2006) failed to find a TNT effect. It is worth noting that, compared to RIF and DF, the magnitude of the TNT effect is quite small

FIGURE 6.3 The think/no-think procedure (Anderson & Green, 2001)

(Anderson & Green, 2001; Levy & Anderson, 2008). Hertel and Calcaterra (2005) argued that the use of particular strategies during suppression may predict successful forgetting in TNT. They replicated the TNT effect only when participants used the strategy of thinking about an alternate word during suppression, either because they were instructed to do so or did so spontaneously (but see Levy & Anderson, 2008).

Like RIF and DF, some researchers have examined motivational influences on TNT; doesTNT dilferentially impact recall of emotional material? Depue, Banich, and Curran (2006) compared TNT for negative and neutral stimuli, and found stronger forgetting effects for negative stimuli. They argued that cognitive control processes may be activated more strongly for emotional information. Although this finding is consistent with a functional view of forgetting, it contrasts with the mixed findings for emotional material in the RIF and DF paradigms. Also, like RIF and DF, other researchers have focused on whether specific populations might show stronger or weaker TNT effects. For example, Joormann, Hertel, LeMoult, and Gotlib (2009) examined TNT of positive and negative words in depressed and non- depressed participants. They found that, while non-depressed participants forgot positive and negative words they had suppressed, depressed participants did not show forgetting of negative words. However, when trained to think of an alternate word during suppression (as in Hertel & Calcaterra, 2005), depressed participants successfully forgot negative words. More recently, Noreen et al. (2020) reported that participants who were low in working memory and higher in depression symptoms showed weaker TNT effects, suggesting that individual differences might impact on effective suppression abilities. Finally, Noreen et al. (2014) gave participants hypothetical scenarios to read and to rate whether they would forgive a transgressor in each one. Following a suppression instruction, Noreen et al. (2014) found that people showed TNT effects for scenarios where they had forgiven the transgressor but not when they had not forgiven.These results suggest that both motivations and strategies may determine the success of suppression in the TNT paradigm.

As with RIF and DF, we have explored whether TNT influences autobiographical memories, using a similar adaptation. In a series of experiments that adapted the TNT procedure to autobiographical memories (similar to our adaptations of RIF and DF), we asked participants to generate autobiographical memories in response to cue words. Then, participants were presented with some of the words, half of which they responded to by recalling the associated memory, and half of which they avoided by suppressing the associated memory. To date, we have conducted five experiments. In the first, participants completed three suppression cycles during the TNT phase. In the second, participants completed 12 suppression cycles. In the third, we instructed participants to think about an alternative memory during suppression (as in Hertel & Calcaterra, 2005). In the fourth, we introduced competition between the memories: participants elicited six memories to each of six cues (as in the RIF paradigm, see Barnier, Hung et al., 2004), so that the respond memories directly competed for recall with the unwanted avoid memories via a shared cue. In our final experiment, we combined 12 suppression trials, a distraction condition, and a cue structure that created competition between the memories, plus a delay between memory elicitation and the TNT phase to reduce overall recall. We also asked participants about their life experiences, particularly about their exposure to trauma and attempts to suppress memories of this trauma in their daily lives (as suggested by Levy & Anderson, 2008). We have had difficulty finding a robust TNT effect. Overall, participants remember their autobiographical events despite repeated attempts to suppress (their memory performance is mostly at ceiling). However, introducing competition between the memories decreased memory overall and may have aided suppression (at least for a subset of participants) and in our most recent experiment, there is some indication that trauma exposure may predict suppression success (Levy & Anderson, 2008). More recently, Noreen and MacLeod (2013) did find TNT effects for positive and negative autobiographical memories, where participants recalled fewer details associated with memories they had intentionally suppressed, as well as slower recall latencies, suggesting that TNT effects can extend to autobiographical material.

Results with TNT are interesting in light of work in the related “thought suppression” paradigm (Wegner, Schneider, Carter, & White, 1987). In our lab, in a thought suppression study comparing repressive copers and nonrepressors, we found that nonrepressors were able to suppress positive memories during a suppression period, but experienced a rebound effect following suppression; they were unable to suppress negative memories at all (Barnier, Levin, & Maher, 2004). In other words, nonrepressors’ initial suppression success, at least for positive memories, did not result in later forgetting, which contrasts with findings from the TNT paradigm. However, repressive copers were particularly successful in suppressing negative events, even when they weren’t instructed to do so (Barnier, Levin, et al., 2004; see also Geraerts, Merckelbach, Jelicic, & Smeets, 2006a), and they showed no rebound effect (but see Geraerts et al., 2006a).This is similar to findings from the TNT paradigm. Thus, it remains unclear when and why suppression (whether in TNT or thought suppression) might result in successful forgetting of autobiographical memories.

More recently, Noreen and MacLeod (2013) did find TNT effects for positive and negative autobiographical memories, where participants recalled fewer details associated with memories they had intentionally suppressed, as well as slower recall latencies.

Conclusion

Based on this review, it is clear that the effects of RIF, DF, and TNT paradigms extend from the simple materials used to develop the original methodologies, to emotional words and sometimes to autobiographical memories. However, as the material increases in complexity (emotionality and personal meaningfulness) so do the effects.These paradigms can be argued to model different mechanisms of goal- directed forgetting and provide good laboratory analogues for everyday, real-world forgetting. As noted above, one assumption of a functional view of memory is that people might try to forget negatively valenced or upsetting memories. In general, results across these paradigms suggest that sometimes people remember more emotional than unemotional material, sometimes they remember as much, and sometimes they forget more emotional material than unemotional. This implies that in remembering and forgetting the past, people are not just influenced by the simple valence of a piece of information or of an event. Likely there are other dimensions predicting its self-relevance, and thus, whether it is prioritized for remembering or forgetting.