Negative memories serve functions in both adaptive and maladaptive ways

ABSTRACT

Autobiographical memories are said to serve at least three functions: they direct people’s behaviour, inform their identity, and facilitate social bonding and communication. But much of the research on these three functions has not distinguished between memories that serve functions in adaptive ways from those that serve functions in maladaptive ways. Across two experiments, we asked subjects to provide either positive or negative memories. Then, to operationalise adaptive and maladaptive functions, we asked subjects to rate the extent to which those memories serve directive, self, and social functions in ways that “help” and in ways that “hurt”. To investigate whether people believe the adaptive benefits of their memories outweigh any maladaptive effects, we also asked subjects how willing they would be to erase the memories if given the opportunity. We found that negative memories served functions in both helpful and hurtful ways, whereas positive memories were primarily helpful. Furthermore, the more helpful a memory was, the more reluctant subjects were to erase it. Conversely, the more hurtful a memory was, the more willing subjects were to erase it. These results suggest it is important to distinguish between adaptive and maladaptive functions when investigating the functions of autobiographical memory.

KEYWORDS:

• Autobiographical memory
• function
• maladaptive
• negative memory
• posttraumatic growth

Acknowledgements

We thank Cassandra Burton-Wood for her invaluable comments and advice; and Matti Vuorre for his help with R code. R. Burnell gratefully acknowledges support from the University of Waikato. A. S. Rasmussen gratefully acknowledges support from the Danish National Research Foundation (Grant DNRF89). M. Garry gratefully acknowledges support of the New Zealand Government through the Marsden Fund, administered by the Royal Society of New Zealand on behalf of the Marsden Fund Council.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 This exclusion rate is lower than that of many other published studies run on Mechanical Turk, including those from our own laboratory. This low rate could be due to the stringent criteria we used for this study, which required at least a 99% approval rating.

2 The reason we employed a logistic regression with subjects’ dichotomous choice as the dependent measure is that the bimodal distribution of subjects’ ratings of their likelihood to erase the memory made those data inappropriate for linear regression. In addition, because there was very little variance in the maladaptive functions of subjects’ positive memories (which were at floor) we restricted these analyses to negative memories only.

3 The data for Experiment 2 were collected before the data for Experiment 1.

4 We did not exclude subjects on the basis of the length of their description because some subjects provided short descriptions (e.g. “getting married”) that appeared to be genuine autobiographical memories.

Additional information

Funding

A. S. Rasmussen gratefully acknowledges support from the Danish National Research Foundation (Grant DNRF89). M. Garry gratefully acknowledges support of the New Zealand Government through the Marsden Fund, administered by the Royal Society of New Zealand on behalf of the Marsden Fund Council.