[W]e are all virtuoso novelists, who find ourselves engaged in all sorts of behavior, more or less unified, but sometimes disunified, and we always put the best ‘faces’ on it we can. We try to make all of our material cohere into a single good story. And that story is our autobiography. (Dennett 1992, p. 114)
As persons, our beliefs and desires are structured in a more or less coherent fashion, such that a mental autobiography—an autobiographical framework (Gerrans 2013) or narrative (Schechtman 1996)—can be attributed, which can explain our cognitive structure. Many people have proposed theoretical entities such as the “autobiographical self” (Damasio 1999, 2010), the “conceptual self” (Conway 2005; Conway et al. 2004), and the “narrative self” (Feinberg 2009), etc. to account for how one comprehends and navigates through the world and over time—that is, how one is able to make sense of external or internal signals, to have preferences, to have goals and to values, to know who oneself is, to be a diachronically persisting agent, to recall the past, and to imagine the future. In general, these different versions of the “extended self” (Gallagher 2000) are characterized by the following phenomenal and epistemic properties.
First, phenomenally, we experience ourselves as thinkers of thoughts (e.g., “I think…” or “I believe…”) and as beings who recollect the past and plan for the future; while at the same time we have a sense of ownership of relevant beliefs (e.g., “this thought is mine”). Second, subjectively, events and objects are presented in a way that manifests their relevance to the subject. In addition, epistemically, we tend to treat the self-told story as if it were highly reliable: The content is treated as objectively real, and its truthfulness is seldom questioned. This is the way we consciously comprehend the world and our place within it, and it is thought to be reliable. Accompanied by a certain degree of the “feeling of familiarity” and the “sense of pastness” (Russell 2009, p. 208), there is a degree of certainty about the veridicality of a mental autobiography. When inconsistency or non-veridicality is detected and such certainty is lost (e.g., due to introspection or contradiction to external information), the mental autobiography will be modified to re-create a new subjective reality—a new story about ourselves with more or less difference (e.g., self-deception).
Delusional patients have anomalous forms of mental autobiography: Their mental autobiographies are radically distorted, for different reasons. For instance, RZ, a 40-year-old female patient with reverse intermetamorphosis, believed that she was her father (and sometimes believed that she was her grandfather) during her assessment by Breen et al. (2000). When asked to sign her name and answer questions about her life, she signed her father’s name and provided her father’s personal history. She acted according to her delusional beliefs. Here we see that her mental autobiography constructs her subjective reality. Semantic dementia patients who suffer from an incapability of constructing personal futures (Irish & Piguet 2013) provide examples of the loss of partial subjective reality.[2] It is speculated that this form of futureless mental autobiography accounts for the higher suicide rate in semantic dementia (Hsiao et al. 2013). As we will see, patients suffering from CD also maintain a mental autobiography.[3] They believe that they are dead or no longer exist: They may refuse to eat or visit the graveyard—the place in which they believe they belong. But how are mental autobiographies constructed? The rest of this section considers how memory systems and simulation models lead to the construction of a mental autobiography.
Studies on misrepresentations in memory have suggested that—against the traditional and folk-psychological idea of a “store-house” (Locke 2008), in which memory as a copy of past experience is stored for future use—memory is constructive in nature. It represents different facets of experience, which are distributed across different regions of the brain, where retrieval is realized in a process of pattern completion, which allows a subset of features to comprise a past experience (Schacter et al. 1998). The prevalence of misremembering (episodic memory in particular) and the view of constructive memory have led to the debate over the function of memory: If the proper function of memory is to veridically represent past experiences or events, is our memory system fundamentally defective? Or, does it serve other functions? If there is any adaptive advantage of memory systems, they must serve a function that concerns the current and/or future states of the organism (Westbury & Dennett 2000). New findings regarding a default-mode network suggest a “constructive episodic simulation hypothesis” (Schacter & Addis 2007a, 2007b), according to which the constructive nature of episodic memory is partially attributable to its proposed role in mentally simulating our personal futures (e.g., planning a future event). This hypothesis is supported by fMRI evidence showing that remembering the past as well as imagining the possible past and future share correlates with the activities of the default mode network (Addis et al. 2007; De Brigard et al. 2013; Szpunar et al. 2007). Therefore, it is suggested that episodic memory is adaptive in that it allows us to employ past experiences in such a way as to enable simulations of possible future episodes.
However, simulation is not realized by episodic memory alone. Though memory systems (i.e., procedural, semantic, and episodic memory) can be conceptually distinguished, they are considered parts of a “monohierarchical multimemory systems model” (Tulving 1985): Semantic memory is a specialized subsystem of procedural memory that lies at the lowest level of the hierarchy, and semantic memory in turn contains episodic memory as its specialized subsystem. The subsystems at higher levels are dependent on and supported by those at lower levels. That is, our everyday autobiographical memory is realized by multiple memory systems. For instance, a recent study has shown the importance of semantic memory in the construction of autobiographical memory: While episodic memory provides episodic details, semantic memory acts as a schema for integrating them (Irish & Piguet 2013). That is, our mental autobiographies are constructed by the interplay of multiple memory systems (e.g., Tulving’s SPI model, see Tulving 1995).
This applies to prospection as well. Different categorizations of prospection are proposed (e.g., Atance & O'Neill 2001; Szpunar et al. 2014). In this commentary, I adopt a distinction offered by Suddendorf & Corballis (2007), who distinguish procedural, semantic, and episodic prospection (p. 301, Figure 1). Suddendorf & Corballis (2007) suggest that the function of the memory and anticipatory systems is to provide behavioral flexibility; and they also examine the phylogenetic development of different memory systems. According to their model, the flexibility of anticipatory behavior supported by different memory systems can offer varies in degree. From the primitive form, procedural memory enables stimulus-driven predictions of regularities and allows behavior to be modulated by experience, such that the resulting behavior is stimulus-bound. Declarative memory provides more flexibility because it can not only be retrieved involuntarily, but can also be voluntarily triggered top-down from the frontal lobe, which enables decoupled representations that are not directly tied to the perceptual system. That is, even though we are still tied to the present in that we recall and imagine the future at the present moment, the content of representation can extend beyond the current immediate environment. Specifically, semantic memory is considered more primitive than episodic memory as it has less scope for flexibility (Suddendorf & Corballis 2007).[4] The former, in allowing learning in one context to be voluntarily transferred to another, provides the basis for reasoning. However, this is about regularities and not particularities. Episodic memory supplements this weakness: A scenario can be simulated and pre-experienced. Through mental reconstruction or memory construction, episodic memory not only recreates past events, it also allows the learned elements to be incorporated and arranged in a particular way in order to simulate possible futures. It thereby provides greater flexibility in novel situations and provides for the possibility of making long-term plans, extending even beyond the life-span of the individual.
To sum up, our mental autobiography is constructed through the interaction of multiple memory systems at different levels. The simulation model should not only be associated with the episodic memory system; rather, it should be understood as a hierarchical model of multiple memory systems—i.e., procedural, semantic, and episodic memory as well as procedural, semantic, and episodic prospection. In the next section I will consider the role of memory systems within the predictive coding framework.