That memory is malleable is a well-researched and widely endorsed view in psychology (Schacter Reference Schacter2001). A key issue emerges from Mahr & Csibra's (M&C's) conceptualization: Why would a memory system be designed in such a way? What are the “forces” that could have led to such malleability? These “forces,” M&C contend, have mainly to do with the intrinsic communicative function of memory. Because memories are used to argue about a particular version of an experienced past, episodic memory should be subjected to fallacies that typically apply to argumentation: confirmation bias, updating, and selectivity.

This claim has solid grounding in the extensive literature on social influences on memory. However, the approach proposed by M&C does not go far enough, I argue, in exploring the communicative function of episodic memory. This is because it focuses almost exclusively on the individual expressing the memory in a communicative setting. According to this view, the cognitive operations involved in these communicative acts are isolated to the individual's mind, with little consideration for the social space in which communication takes place. Understanding the listener's experience during the communicative event, I contend, highlights the role of an unexplored facet of communication – its relational motivation – in shaping memory. I want to argue that our memories are as adaptable as they are – not necessarily because they involve a communicative dimension, but because communication involves relating to another individual. In other words, adaptive memory systems, able to integrate social influences, will allow for the creation of shared memories among the communicators. This, in turn, facilitates joint behavior and coordinated action and, due to these benefits, might offer adaptive advantages to the individual.

Evidence for the role that relational motivations play in shaping memory comes from the processes involved in the selective retrieval of previously encoded memories. Several studies have shown that listening to another person retrieving previously encoded information leads to the strengthening of the retrieved information, but at the same time, it results in the induced forgetting of related, unretrieved information (Cuc et al. Reference Cuc, Koppel and Hirst2007). Importantly, these effects occur only when the listener is concurrently retrieving the information along with the speaker. But not all listening experiences result in the concurrent, and covert, retrieval of information along with the speaker. On the one hand, concurrent retrieval is influenced by the listeners' motivations to be accurate about the past (i.e., epistemic motivations), as shown by Cuc et al. (Reference Cuc, Koppel and Hirst2007) and Koppel et al. (Reference Koppel, Wohl, Meksin and Hirst2014). These motivations were front and center in M&C's approach and for good reason. More importantly for the claims made herein, the motivation to relate to the speaker (i.e., relational motivation) impacts the degree to which listeners experience retrieval-induced forgetting following the listening task. When participants (Princeton students) listened to another Princeton student selectively practicing previously encoded information, they were more likely to forget information that was related to what the speaker mentioned than when they listened to a Yale student (Coman & Hirst Reference Coman and Hirst2015). Thus, the listeners' motivation to relate to an in-group member impacted the degree to which they concurrently retrieved the information with their fellow in-group member.

These phenomena have consequences for both conversational partners. If the speaker and the listener strengthen similar memories and they both experience suppression of the undiscussed related information, then their memories will become more aligned following collaborative remembering, as found by Stone et al. (Reference Stone, Barnier, Sutton and Hirst2010) and Coman and Hirst (Reference Coman and Hirst2012). And having similar memories could impact people's sense of shared identity (Fivush Reference Fivush2010a), their ability to make collective decisions (Kameda et al. Reference Kameda, Ohtsubo and Takezawa1997), and their collective behavior (Harris Reference Harris2006). So adaptive memory systems constitute the building blocks of social formations (Coman Reference Coman and Wright2015). This view goes beyond simply acknowledging that we communicate our memories and emphasizes the why of communication, which might constitute a more proximal explanation for the malleability of human memory.

Expanding the communicative dimension that characterizes our episodic memories from its epistemic function to its relational function will result in a more complex view of human memory. The reason why we have the flexible memory systems that we do is because this allows communicating individuals to create a shared reality or worldview (Hardin & Higgins Reference Hardin, Higgins, Sorrentino and Higgins1996). Through its impact on group identity and collective coordination, the ability to create a shared reality with others might have provided selective advantages in the evolution of the human species.