Suddendorf & Corballis's (S&C's) claim that memory systems are adaptive for their contribution to future survival is consistent with Tulving's (Reference Tulving1983; Reference Tulving, Terrace and Metcalfe2005) arguments, and with my proposal for the evolution and development of memory (Nelson Reference Nelson, Collins, Conway, Gathercole and Morris1993a; Reference Nelson1993b; Reference Nelson, Ellis and Bjorklund2005). In considering the emergence of autobiographical memory in childhood, Nelson and Fivush (Reference Nelson and Fivush2004) proposed a constellation of contributions to this manifestation of time travel, similar to S&C's proposal of a constellation of mechanisms responsible for foresight. We emphasized the development of representational language, conversational exchanges about past and future, and cultural practices, thereby placing more weight on the co-development of culture and biology in the emergence of episodic memory – and foresight – in both phylogeny and ontogeny than S&C do. The neglect of culture in mind in S&C's account is in my view a serious drawback to their account of the uniqueness of the human ability to remember the past and foresee the future.

Prominent among the achievements Nelson and Fivush (Reference Nelson and Fivush2004) identified as necessary to autobiographical memory was skill in narrative construction and understanding and its use in personal memory recounts. Narrative is a unique cultural production, as universal in human societies as language itself. It provides the structural glue that ties together the who, what, where, when, and why that S&C recognize as necessary to complex foresight. But their theater metaphor strangely neglects the essential structure of narrative, the plot or drama that their “playwright” must produce. Instead, they imagine the playwright picking and choosing among pieces of prior specific episodic memories to make up a new scenario. But without the structure of a narrative, situated in a specific cultural setting, the play – the memory or foretelling – is untethered. This is well observed in the many extensive records of children's contributions to memory talk now in the literature, which begin with bits and pieces of scenes, and only with practice and the help of older conversational partners eventually come to resemble narratives of a “self” experience with a beginning, identified characters, setting, some highpoint, and ending, perhaps with an evaluation. Autobiographical memory consisting of narratives from the personal past emerges gradually over the preschool years as children gain practice in reminiscing with others about things they have experienced together or separately (Nelson & Fivush Reference Nelson and Fivush2004). There is much less research on children's experience in talking about the future, but there is evidence that future talk is less frequent than talk about the past and occurs more with older preschoolers (Eisenberg Reference Eisenberg2006), suggesting that personal memory is the key to the child's engagement in time travel.

From the perspective of these developments it seems clear that, as Tulving claimed, the unique time travel addition in human evolution is not in foreseeing the future, but rather, the ability to recall the self in specific past events, that is, episodic memory (EM). EM may serve the future, as does all memory, but this is not its unique function. General scripts may be more useful for predicting what will happen next, allowing for flexibility through open slots and different paths. In fact, very young children tend to focus on the general “way things are” rather than on the particularities of a specific past. This makes functional sense when the goal is to understand, anticipate, and participate in the social world. Declarative memory (facts about the world, including routines and scripts) is a collection of not-now know-how and know-about, and may also include episodic fragments and even singular episodes, without specifying that they “happened to me at a particular time in the past.” In contrast, EM does preserve specific complex events of personal significance. EM is the only form of memory that is about the past; it is also the only form that is about the self (autonoetic) (Tulving Reference Tulving1993).

Self-memory is late in developing, and is followed by developing ideas of a self-future. The critical questions about these emerging senses of self in time – in development as in evolution – are why and how? Children begin reporting on specific past episodes (around 3 years of age), but it is several years before most children fluently compose narrative accounts of their own experience, or specific plans for the future. Traditional developmental accounts assume that egocentrism is characteristic of early childhood. Therefore, the slow development of understanding of self in time is not expected, and relatively little research has been devoted to it (Moore & Lemmon Reference Moore and Lemmon2001; Nelson Reference Nelson1996; Reference Nelson, Moore and Lemmon2001).

Early in development self-experience is the source of all memory, because one person's experience is only accessible to another through language, which infants and young children are not privy to. Thus, the only experiencer in a child's early memory is the I. It is not until the child has acquired sufficient skill with language to engage in conversations with others about their past and future experiences that identifying different experiencers other than the self becomes critical. Hence, only when the child is exposed (through language) to contrasting pasts and others' futures that the child's own self-in-the-past becomes salient, distinguished from a generalized “not now” held in common with others.

Relatedly, younger children remain ignorant about or indifferent to the source of their memory – or of information in general – up to the age of about 6 years. Younger children typically do not monitor whose memory is reported or where a particular bit of knowledge arrived from (Roberts & Blades Reference Roberts and Blades2000). This is understandable, given that the basic memory system conserves information from the child's (or more generally, an animal's) own experience in the social and physical world. It is only humans who share information about happenings that they have individual knowledge of. Therfore, only humans must focus attention on sources, specifically between self-experience or others' report. Then the self-narratives of autobiographical memory emerge from a murky past in culturally supported forms learned through everyday social interactions during the preschool years. These complexities of the developing self in the social world strongly imply that the emergence of both personal memory and personal possible futures is attributable to the social and cultural conditions and narrative framing of human lives and experiences.