Autobiographical memory functions in the recall of authentic moments

Authenticity, or being true to oneself, relies on integrating autobiographical memories into a coherent life-story. These autobiographical memories serve three adaptive functions: self-identity, social connection and directing behaviour in the future. But little is known about how memory and the subjective sense of authenticity interact. Using a between-subjects experimental design, we compared memories of authentic and inauthentic moments with controls. Memories of in/authentic events were more likely to use the self function and less likely to use the social function than controls. In addition, we noted a trend for memories of inauthentic events to be more likely than controls to use the directive function. These results highlight the importance of an ecological approach to memory as well as the potential adaptive value of recalling inauthentic experiences in developing a coherent sense of self.


Introduction
Authenticity -a subjective sense of being true to oneself -contributes to positive life outcomes such as well-being and engagement (Sutton, 2020) and constructing this authentic sense of self is thought to rely on the recall of autobiographical memories that can be integrated into a coherent story (Boucher, 2011;Fivush & Graci, 2017).But we know surprisingly little about what functions autobiographical memories might serve in supporting authenticity.This lack of data thwarts our understanding of mechanism and hinders progress in the development of theory.To address this gap, we asked people to recall times they felt authentic or inauthentic, and identified the functions that those memories served.

Functions of autobiographical memories
The functional approach to autobiographical memories (AM) aims to understand how memories of our experiences are linked to subsequent psychological outcomes (Waters et al., 2014) and has identified three distinct adaptive functions served by autobiographical memories: self-continuity or definition, building social connection and directing behaviour in the future (Bluck et al., 2005).Greater use of these three functions is associated with higher psychological well-being (Waters, 2014).
The self function is concerned with using memories of past experiences to build self-knowledge and develop a sense of continuity in the self-concept.This function can help someone to consider how much they may have developed over time (Bluck & Alea, 2011).The social function is fulfilled when memories are used to foster and build social bonds, for example by sharing memories with others or learning about a person through using those memories.This function helps to develop intimacy in relationships (Bluck & Alea, 2011).Finally, AM may be used to direct future behaviour: the directive function.This includes using memories to solve problems or develop plans and guidance for the future.The directive function includes using those memories of past events to understand an experience or create a sense of meaning (Bluck et al., 2005).
But these functions do not operate in a vacuum.The ecological approach to autobiographical memory emphasises that remembering occurs within a context and is therefore subject to influence by a range of factors (Lind et al., 2019).These factors include individual differences, such as the influence of self-esteem and ruminative focus on the specificity and detail displayed in recall of autobiographical memories (Roberts et al., 2021), and types of memories.For example, single events tends to serve more self and directive functions, while recall of repeated or recurring events serves a social function, and extended events serve all three (Waters et al., 2014).Similarly, the emotional valence of a memory influences the functions that memory serves, with the directive function being more likely than other functions to be used with negative memories (Lind et al., 2019).
The emotional valence of a memory is, however, distinct from how adaptive or helpful a person finds it.Pillemer (2001) notes that negative memories can be reinterpreted in the context of an overall life story as motivating rather than demoralizing.Negative memories can be adaptive and we are less likely to wish to 'delete' a helpful memory even if it is negatively valenced (Burnell et al., 2020).Similarly, inauthenticity is commonly perceived as a negative state or experience that we are strongly motivated to avoid (Lenton et al., 2013), yet initial indications are that memories of being inauthentic can serve adaptive functions such as directing future behaviour (Sutton & Render, 2021).
While it is useful to know how functions used in diverse types of memories may differ, there have been calls for research to move into identifying patterns or interactions in how these functions are used (Pillemer, 2003).A recent example of work answering this call is the investigation of memories of trust relationships in the workplace which explored interactions between memory valence and cultural context (Wasti et al., 2021).This study found complex interactions between culture, valence and memory function.For example, positive memories served a social function more than negative memories across the three cultures surveyed, but the self function was used more in a Turkish sample than Singaporean or American, especially for positive memories.Studies like these continue to point to the need for an ecologically valid expansion of AM function research and a nuanced understanding of how contextual factors influence AM.

Autobiographical memory and authenticity
One of the most important contextual factors is the individual's conceptual self, that is, their current knowledge of personal characteristics such as traits and attitudes (Conway et al., 2004).This conceptual self both guides the recall of AM and is influenced by the memories in the construction of a coherent self-narrative or 'working self' (Demiray & Bluck, 2011).Similarly, authenticity, our sense of being true to ourselves, is only possible to the extent that we can evaluate a current experience or recalled memory against a conceptual understanding of who we are.While there are several psychological models of authenticity, they all involve elements of both self-knowledge and self-expression (Harter, 2002).Maintaining this ongoing life-story of who we are (McAdams, 2001) or sense of self-continuity through the recall of our personal past, is a key function of autobiographical memory (Bluck & Liao, 2013).
Although the research on authenticity and autobiographical memory is nascent, there are some emerging findings indicating a complex interaction between the two.Recalling a nostalgic memory can increase a sense of authenticity in the past self and help those who find themselves in contexts that restrict authenticity (Baldwin et al., 2015).In addition, higher authenticity is associated with recalling a memory of achieving a goal or helping others (Smallenbroek et al., 2017).Given that a range of research has demonstrated differential function use in different types of memories, it is likely that memories of authentic or inauthentic moments too, will serve different functions.A small study, the first of its kind to investigate differential AM function use in memories of authentic and inauthentic episodes, indicated that recall of inauthentic events served primarily to direct future behaviour while recall of authentic events served the self function (Sutton & Render, 2021).

The present study
This study aims to investigate the interaction between authenticity and autobiographical memory by identifying the AM functions used in memories of times people felt authentic or inauthentic.To do this, we partially replicate and extend on an initial study that used archival data to identify withinperson differences between functions used in recall of these experiences (Sutton & Render, 2021).Instead of within-person differences, we focus here on between-person differences and include a control condition of 'free recall' as a comparison to memories of authentic or inauthentic moments.
Self-knowledge is a basic component of authenticity (Kernis & Goldman, 2006;Wood et al., 2008): we cannot feel true or untrue to ourselves if we do not know anything about ourselves.We therefore expect that the self function will be used significantly more in memories of authentic or inauthentic experiences than in controls.The social function, with its focus on developing social relationships, is expected to be used less in memories involving a sense of in/authenticity, which is self-focused, than in control memories.Finally, previous work has shown that reflections on inauthentic behaviour include elements of problem-solving and considering future behaviour (Sutton, 2018) and we therefore expect that recall of inauthentic moments are most likely to serve the directive function than controls.

Participants
A total of 282 psychology students from New Zealand participated in the study in exchange for partial course credit.Incomplete responses (N = 19) and speeders (N = 22), those who completed the questionnaire 50% faster than the median (Greszki et al., 2014), were removed.In addition, on closer inspection of the memory narratives, two participants were excluded for low quality responses.A priori power analysis (using G*Power) indicated that a total sample of 245 had 95% power to detect a medium effect size.After data cleaning, sample size was 239, giving a power of 94%.Of the final 239 respondents, 81.6% were female, 15.5% male and 2.9% other/unspecified.Most respondents (66.1%) were less than 25 years old, followed by 21% aged 25-34 and 12.6% older than 35 years.56.1% reported NZ European ethnicity, 6.7% Māori, 20.1% other ethnicities and 17.2% mixed ethnicity.

Procedure
Data were collected through an online questionnaire.Participants read an information sheet and provided informed consent before proceeding to the questionnaire.The study was approved by the authors' institutional ethics committee.Participants first reported basic demographic data (sex, age and ethnicity) and then were randomly assigned to one of three groups (authentic, inauthentic or control condition) and asked to recall a memory relevant to that condition: Please take a moment to think back over the last year and recall a time in your personal or professional life when you behaved in a way that made you feel (un)true to yourself, that made you feel (in)authentic.(Or, for the control condition: Please take a moment to think back over the last year and recall a memorable time in your personal or professional life.)Describe what happened, where and when the memory took place, who was involved, and thoughts and feelings during the event.

Measures and coding of event narratives
To determine whether respondents in the experimental conditions had followed instructions, all narratives were coded for the extent to which they represented authentic or inauthentic moments.Definitions of authenticity and inauthenticity were provided to two coders with expertise in authenticity but blind to the aim of the study.After training, coders read each narrative and rated it on the questions To what extent does this narrative represent a memory of a time the person was (in) authentic?on a five point scale from 1 = not at all to 5 = a great deal.Each narrative therefore received scores from two coders and the mean score was used in the subsequent analysis.
An ANOVA of these ratings by condition was conducted and demonstrated that the authenticity rating was significantly different across conditions (F (2, 236) = 114.45,p < 0.001,η 2 = 0.49), as was the inuthenticity rating (F (2, 236) = 114.45,p < 0.001, η 2 = 0.70).Post-hoc comparisons using Bonferroni corrections showed that memories in the authentic condition scored significantly higher on authenticity than memories in either the control or inauthentic condition.Similarly, memories in the inauthentic condition scored significantly higher on inauthenticity than memories in the authentic or control condition (see Fig. 1).The control or free recall condition is therefore demonstrably less concerned with memories of authentic moments than the authentic condition and less concerned with memories of inauthentic moments than the inauthentic condition.
The author, plus two trained coders1 who were blind to the condition and research question, used Waters et al.'s (2014) autobiographical memory function coding scheme to score the recall narratives.Each narrative was coded for the expression of self, directive and social functions on a 4-point scale from 0 (no content expressing this function) to 3 (extensive elaboration of this function).A 15% sample of the narratives were scored by all coders, after which coders met to discuss and resolve discrepancies, before independently scoring the remaining narratives.These scores were combined into a mean score for each function.
The majority (96.2%) of the recall narratives received a score on at least one AM function.Table 1 provides examples of the coding for authentic and inauthentic events.

Results
Table 2 provides the descriptive statistics for variables in this study.There were no significant differences between conditions on ethnicity or sex.Bivariate correlations revealed significant positive correlations between length of narrative and scores on all three AM functions.As the system gives higher scores for greater elaboration and detail, this is to be expected.A one-way ANOVA of the length of recall narrative (word count) by experimental condition showed a significant difference, though the effect size was small and 95% CI crossed zero (F (2,239) = 3.51, p = 0.03, η 2 = 0.03 [95% CI 0, 0.08]).Bonferroni post-hoc tests revealed that participants in the control condition wrote more than those in the inauthentic condition (mean difference = 28 words, p = 0.04) but there was no significant difference between the number of words in the two experimental conditions.To avoid confounds, therefore, word count is therefore included in the following analyses as a covariate.

Discussion
Using a between-subjects experimental design, this study partially replicates and significantly extends previous work on memories of authentic and inauthentic moments (Sutton & Render, 2021) to demonstrate distinct patterns in AM function use.The findings contribute to the ecological approach to autobiographical memory (Lind et al., 2019) by identifying further contextual factors that influence how memory functions, as well as answering the call for a better understanding of the patterns and interactions between these functions (Pillemer, 2003).They also provide further support for the self-memory system model (Conway et al., 2004) by identifying the functions of AM in served by memories involved in constructing an authentic sense of self.
A previous study using within-person analysis demonstrated that people were least likely to use the social function in in/authentic memories, with the self function used most in authentic memories and the directive in inauthentic memories (Sutton & Render, 2021).Using a between-subjects design, we have found support for some of these distinctions.Memories of a time an individual felt authentic are more likely to serve a self-identity and less likely to serve a social function than controls.In comparison, memories of inauthentic moments are similarly more likely to be used for self-identity and less for social bonding than controls, but also tended to serve the directive function more than controls.This trend for the directive function to be used more than control was also seen for authentic memories, though it did not reach significance, and we also note that there were no significant differences in function use between the authentic or inauthentic conditions.This indicates that, whether the memory is of a time one was authentic or inauthentic, it is used in a very similar way: primarily for self-identity.Authenticity necessarily involves an evaluation of experience with the current self-concept and the importance of the self-identity function in memories related to authenticity provides further support for the selfmemory system model in which beliefs about and knowledge of the self are supported by specific autobiographical memories (Conway, 2005).Similarly, authenticity has been shown to be distinct from simple consistency in personality traits and instead be more appropriately defined in terms self-coherence (Sutton, 2018).The development of this integrating narrative is central to the self-memory system and memories of moments of being true or untrue to oneself are used to build a narrative of the current working self.
While previous qualitative work has shown that the content of reflections on authentic or inauthentic moments often involves reference to others, such as work colleagues or friends (Sutton, 2018), it is of note that the findings here demonstrate the social function is used less in these memories than in control memories.
Inauthenticity is perceived as a negative experience that people are strongly motivated to avoid (Lenton et al., 2013), yet mounting evidence indicates that memories of negative events may be valued for their adaptive or positive effect (Burnell et al., 2020;Pillemer, 2001).We demonstrate that memories of both authentic and inauthentic experiences Table 1 Coding examples of self, social and directive functions in autobiographical memories of authentic and inauthentic experiences.These examples are from the highest scores in each condition

Self
Code Mention of self-enhancement or self-deprecation due to reflection or remembering the experience, mention of similarity or difference of self and other, identification with an individual or group, identification of personal goals, explicit mention of personal traits Authentic example …During my final year of high school, I organised a clothing drive for the under privileged community that lives close to my school.They had just experienced a massive fire that had burnt … all their belongings.We organised a day for people to bring in anything they could spare.We made sure to hand everything over to the right people so that it would be sorted through and distributed to those affected.The amount of gratitude that I received was so overwhelming I struggled not to break down.I knew how much this was going to help them and my philosophy in life is that if you are born with more it is your duty to give what you can to those who have less Inauthentic example [describing the breakdown of a relationship] Over the course of the relationship I learned to stay quiet when in public, as to not offend them with comments I made, I learned that sharing my past experiences and memories would result in arguments.I felt a sense of loss because I didn't know who I was, and whether the person I had become was someone I was proud of… I am angry that I put so much of my self-worth into his hands, as beforehand I believed I knew myself, my values, and my worth very well Control example At the beginning of the year I decided to enrol in university.I never thought I would be here at twenty nine, my family are not the "go to university" type of people.… I never felt like I was good enough or smart enough to be here but I seem to have proven my past self wrong.When I got the message that I was accepted in I was ecstatic.… I knew I was making a decision that would have an impact on my future and that made me feel really good about myself Social Mention of a relationship or tradition as meaningful or valuable; description of a relationship as helpful; missing an individual or period in a relationship Authentic example Last year I hosted Christmas at my house.My whole family came over and we played games ate food and drank.I was on a high all day, being surrounded by family, food and love.I was … honored to have a family that could make it all possible Inauthentic example I had neglected a lot of my friendships which in turn are still suffering even though I walked away from the person causing those changes.I am still hurting at the breakdown of those friendships and questioning as to how I can fix them Control example I have a vivid memory of myself, my husband and my three kids all being together for the first time since the beginning of the year due to lockdown.It was a clear, chilly winters night … We had built a couch out of wooden pallets to sit on around the fire.The kids all had skewers and were roasting marshmallows over the fire … I remember saying to my kids that this moment right here, doing this sort of thing is what I pictured being a mum would be like and I would always remember this moment.I felt immense pride and overwhelming love for these three children I had brought into this world Directive Mention of a change in a specific behavior, attitude or principle change as a result of the experience Authentic example During my final year of high school … everyone around me had hopes that I would do a particular degree and pursue a certain career path, however this was not what I actually wanted to do.I ended up enrolling for the degree that I actually wanted to study and began working towards the career I actually see myself in long term.I began working and volunteering in areas that focused on my passions, and now I actually see myself in the vocational area I'm deeply passionate about Inauthentic example I had to roll out a program at work to help the people I work with but the times [given to me] were not suitable [for my colleagues].I could have been more proactive in challenging those times but I was afraid of the backlash from management.I will be more assertive next time as it will benefit the people I am working with Control example After a fall from my horse and two months later, I received a double kick from another horse, I had lost all my confidence around horses and doubted if I should still own mine.My fear for horses stayed with me up until 6 months ago when I pushed myself to take my horse for a training weekend away, and I achieved so much … I also proved to myself that I could do it.I am now back hands-on with [horse's name], floating him to western events and feeling like I have my love again were found to serve the adaptive function of contributing towards self-identity.In addition, memories of inauthentic moments were somewhat more likely to be used for directive purposes than were control memories, indicating that participants may use the memory of this potentially negative experience to change their behaviour and avoid similar experiences in future.

Limitations and future research
Scoring of the narratives in this study used a system that gives greater weight to recall with greater elaboration, hence the necessity of controlling for word count.Future research using a scoring system that evaluates the relative importance of these three functions, without reference to extent of elaboration, could provide further confidence in these findings.
The control condition we utilised here was a 'free recall' condition rather than designed to specifically exclude moments of authenticity or inauthenticity.If future research is able to exclude all (in)authentic content from control memories, the relationships found here may prove even stronger.
Memory content may influence feelings of authenticity: a memory that is appraised as less self-congruent (equivalent to an 'inauthentic' moment in this study) is expected to reduce feelings of authenticity, and hence have an influence on whether the individual uses 1 st or 3 rd person perspective in retrieving the memory (Sutin & Robins, 2008).According to this model, adopting a 3 rd person perspective on inauthentic moments may help to provide distance and objectivity and thereby lead to improved feelings of authenticity.Future research could test this model by investigating the interactive effects of memory perspective and function on subsequent feelings of authenticity.
Greater use of AM functions is related to higher wellbeing (Waters, 2014), as is being authentic (Sutton, 2020).It is possible that recall of authentic moments could function to improve well-being in a similar way that writing about extended life events (life story 'chapters') improves selfesteem (Steiner et al., 2019) and the findings reported here serve as a foundation for future work to clarify the interactions between memory and authenticity.This increased understanding of how authenticity relates to autobiographical memory is essential if we are to develop a scientificallyinformed and ethical response to emerging technologies in memory modification.
In conclusion, this study demonstrates that memories of times we are authentic and inauthentic serve distinct functions.While both contribute towards building self-identity, recall of inauthentic moments is also useful in providing future direction and changing behaviour.

Fig. 1
Fig. 1 Authenticity and Inauthenticity ratings for memory in each condition.Note.Error bars represent 95% CI

Table 2
Descriptive statistics and bivariate correlations ** p < 0.01 Estimated marginal means for memory function by recall condition.Note.Error bars are 95% CI and significant differences indicated by *; covariate (narrative word count) is evaluated at the value of 105.91