Remembering the Future: Episodic Simulation, Mental Time Travel and the Affective Basis of Prospection

Abstract

Episodic memory has traditionally been characterised as a system for retaining the past. Converging neuropsychological, neuroimaging and behavioural evidence has recast it as a system whose primary function may be prospective: supplying the materials from which possible future events are simulated. The present review examines this claim through Tulving’s account of autonoetic consciousness and chronesthesia, Suddendorf and Corballis’s mental time travel framework, and the constructive episodic simulation hypothesis of Schacter and Addis. We consider evidence from amnesic patients, from fMRI studies showing a shared core network for remembering and imagining, from overgeneral memory in depression, and from phenomenological and affective-forecasting work indicating that future simulations are emotionally weighted and that the ventromedial prefrontal cortex supports their affective content. We conclude that the loss of emotionally detailed episodic memory constrains not only recollection but the capacity to prospect, and derive requirements for any programme aiming to preserve human emotional memory: what must be preserved is not a record but the reconstructive, self-referential and affective components that permit recombination.

Keywords: episodic future thinking; mental time travel; autonoetic consciousness; constructive episodic simulation; default network; affective forecasting

1. Introduction

For most of the twentieth century, the study of memory was the study of retention; imagining the future was assigned to a separate faculty. This division was challenged from two sides. Clinically, Tulving (1985) reported that a densely amnesic patient was equally unable to recollect his past and to describe his following day. Theoretically, Suddendorf and Corballis (2007) proposed that mental travel into one’s own past and future is a single faculty whose adaptive value lies in the future. Schacter and Addis (2007) then supplied a mechanism: episodic memory is constructive precisely because its function is to permit flexible recombination of past elements into novel simulations.

The literature that followed, grouped under the label of episodic future thinking (Schacter, Benoit, & Szpunar, 2017), converges on three claims: that remembering and imagining depend on a largely shared neural core; that damage to that core impairs both; and that the simulations it generates are emotionally weighted in ways that guide decision. The present review organises the evidence around these claims and draws out a consequence often left implicit: if the affective quality of episodic memory gives simulated futures their motivational force, then the loss of emotionally detailed memory does not merely impoverish the past; it constrains the futures a person can represent. This changes what an effort to preserve emotional memory must preserve.

2. Theoretical Foundations

2.1 Autonoetic consciousness and chronesthesia

Tulving (1985) distinguished three forms of consciousness aligned with three memory systems: anoetic (procedural memory), noetic (semantic memory, knowledge of facts without any sense of when they were acquired) and autonoetic, which accompanies episodic memory and consists in the awareness of oneself re-experiencing an event in one’s personal past. He later introduced the term chronesthesia (Tulving, 2002) for the capacity to be aware of subjective time and to locate oneself within it, and framed it as bidirectional: what permits re-experiencing the past permits pre-experiencing the future.

2.2 Mental time travel

Suddendorf and Corballis (2007), extending their earlier formulation, proposed that mental time travel is a unitary ability built on a representation of the self, the ability to disengage from current perception, and a generative system for constructing scenarios. On their account, memory is episodic because a store of specific, contextually bound experiences is the raw material for constructing specific futures; their further claim that the ability is uniquely human remains contested.

2.3 The constructive episodic simulation hypothesis

Schacter and Addis (2007) turned an apparent flaw of memory into an explanation of its design. Recollections are assembled from fragments and are vulnerable to source confusions; on an archival view, these are pure costs. Schacter and Addis proposed instead that a system built to simulate future events must extract and recombine elements of past experience, and that this flexibility inevitably produces reconstruction errors. The hypothesis predicts overlapping neural systems for remembering and imagining, and impaired future thinking wherever episodic memory is impaired. On this view the brain is fundamentally prospective: stored information serves simulation rather than retention for its own sake. Addis (2020) has since reframed remembering itself as one instance of event simulation, differing from imagining chiefly in the constraints placed on the construction process.

3. Neuropsychological Evidence

3.1 Patient K.C.

K.C. sustained bilateral medial temporal lobe damage in a motorcycle accident in 1981. Intelligence, language and semantic knowledge of his own life were preserved, but he could not recollect a single personal episode. Tulving (1985) observed that when K.C. was asked what he would do tomorrow, he reported the same subjective blankness as when asked what he had done yesterday. The observation was informal, but it is the first documented dissociation between intact semantic knowledge and abolished autonoetic access in both temporal directions.

3.2 Patient D.B.

Klein, Loftus and Kihlstrom (2002) provided a systematic replication. D.B., a 78-year-old man rendered amnesic by anoxia following cardiac arrest, was severely impaired both on recollecting his own past and on imagining his own future. When asked about the future of public affairs, however, he answered coherently and comparably to controls. D.B. retained the concept of the future and could reason about it, but could not place himself within it. As with K.C., the evidence is a single case with non-standardised tasks, limiting generalisation but not the logic of the dissociation.

3.3 Hippocampal amnesia and scene construction

Hassabis, Kumaran, Vann and Maguire (2007) tested five patients with bilateral hippocampal damage using a paradigm that removed personal reference: participants imagined fictitious scenes cued by short prompts, for example lying on a white sandy beach in a tropical bay. Descriptions were scored for content and spatial coherence. Four of the five patients produced descriptions that were reduced in content and, most strikingly, fragmented: they named sand, sun and water but did not bind them into a scene one could occupy. The authors interpreted the hippocampus as providing the spatial scaffold within which details are integrated, whether the representation is a memory, a future event or a fiction. The deficit lies at the level of scene construction, on which both recollection and prospection depend. One patient performed normally, a reminder that lesion extent matters.

4. A Shared Core Network

4.1 Neuroimaging of remembering and imagining

Addis, Wong and Schacter (2007) provided the first direct fMRI comparison of past and future event generation, using matched cue words and separating construction from elaboration. Both conditions engaged a common set of regions: hippocampus and parahippocampal cortex, medial prefrontal cortex, posterior cingulate and retrosplenial cortex, and lateral temporal and parietal cortex. During construction, future events produced greater activity in right frontopolar cortex and anterior hippocampus, consistent with the added demand of recombining elements into a novel configuration. The overlap has been widely replicated; divergences concern the interpretation of the future-greater-than-past differences rather than the overlap itself (Schacter et al., 2017).

4.2 Self-projection and the default network

Buckner and Carroll (2007) noted that these regions coincide closely with those engaged by theory of mind and spatial navigation, and with the default network, active during rest and suppressed during externally directed tasks. They proposed that these tasks share a common process, self-projection: the shifting of perspective from the immediate environment to an alternative situation, whether past, future, another person’s or another place. On this view the default network is not idle but simulates situations that are not present.

5. The Affective Dimension of Prospection

5.1 Overgeneral memory and the impoverished future

Depressed and suicidal patients, given a cue word such as happy and asked for a specific memory, tend to respond with categorical summaries rather than single dated episodes. Williams et al. (1996) extended this overgeneral memory to the future. Parasuicidal patients who produced overgeneral memories also produced less specific and less imageable descriptions of future events, with the two measures correlated. In a second study, healthy participants trained to retrieve memories categorically then generated less specific future events than those trained to retrieve specifically. The induction result moves the relation towards causation and offers a cognitive account of the foreshortened future in depression. If specific, affectively textured memories are unavailable, the future cannot be populated with such scenes, and hopelessness becomes less a judgement than a representational vacancy.

5.2 Phenomenology of future thinking

D’Argembeau and Van der Linden (2004) asked healthy participants to generate past and future personal events varying in valence and temporal distance and to rate their sensory, contextual and emotional qualities. Future events contained fewer sensory and contextual details than past events, consistent with their being constructed rather than retrieved; positive events were richer in detail than negative events, and near events richer than distant ones. In later work the same authors found that individual differences in visual imagery and habitual emotion regulation predicted the phenomenological quality of both remembered and imagined events, with expressive suppression associated with less vivid representations in both directions. Emotion is thus a determinant of the vividness of future simulations, not an optional colouring.

5.3 Affective forecasting

Gilbert and Wilson (2007) reviewed evidence that people use prospection to preview the hedonic consequences of possible events, and that these affective forecasts are systematically biased because simulations are unrepresentative, essentialised and abbreviated. The relevant point here is the mechanism: the affective response to a simulated event is the signal by which the organism evaluates it, obtained by running the simulation and reading off the feeling it produces. Prospection is thus affective by design, and guides choice only if the simulation is detailed enough to elicit a discriminating emotional response.

5.4 The ventromedial prefrontal cortex

Benoit, Szpunar and Schacter (2014) examined the neural basis of this affective component. Participants imagined future events involving familiar places and people varying in how much they were liked. Ventromedial prefrontal cortex (vmPFC) activity tracked the imagined event’s affective value, and vmPFC connectivity with regions representing the constituent elements increased during integration. The authors interpreted vmPFC as integrating distributed knowledge about the elements of a simulation into a representation of its affective value.

6. Synthesis: Emotional Episodic Memory as a Constraint on Prospection

The evidence supports a chain of dependencies. Scene-like personal prospection depends on a construction process in the hippocampal-cortical core network; that process draws on stored episodic elements; and the resulting simulations acquire motivational force from an affective evaluation in which vmPFC is centrally involved. Hippocampal amnesia, overgeneral memory and habitual affective suppression show that degradation at different points in this chain produces a correspondingly degraded future. This is the thesis of the present review: the loss of emotional episodic memory constrains prospection in two ways, by removing the specific elements from which scenes are built and by removing the affective signal by which simulated scenes are evaluated. Such a person may retain the concept of the future, as D.B. did, but cannot populate it with self-involving, emotionally weighted episodes.

The claim requires caution. Most causal evidence comes from small patient samples or short-term inductions, the fMRI evidence is correlational, and the considerable role of semantic memory in future thought has not been reviewed here. Nor is the relation monotonic, since excessively vivid negative simulations are implicated in anxiety; the evidence establishes a dependency, not an optimum.

7. Implications for Emotion Preservation

A programme aiming to preserve human emotional memory must be assessed against a criterion that is not archival: not whether an experience has been recorded but whether what is retained can support the reconstructive, self-referential and affective operations by which episodic material becomes usable for prospection. Several requirements follow, stated as testable propositions.

First, preservation of semantic facts about emotional events is insufficient: K.C. retained the facts of his life and could not use them to imagine a future. A preserved representation should be evaluated on whether it supports scene construction, measurable by the scoring of Hassabis et al. (2007), rather than on factual completeness.

Second, specificity is a preservation target in its own right, since categorical summaries do not support future imagery and specific episodes do. Interventions restoring access to specific episodes should therefore improve the specificity of future thinking; strategies which abstract or summarise experience should not.

Third, affective detail must be preserved together with perceptual and contextual detail, since emotional intensity and sensory richness covary and affective value is computed by integrating the elements of a scene. A representation stripped of its emotional component should yield simulations that are constructed but not evaluated, and hence inert for decision.

Fourth, the autonoetic component, the sense of oneself within the scene, is the least understood and most fragile; D.B.’s dissociation shows it can be lost while temporal reasoning survives. Whether any external trace can re-instate rather than merely cue it is an open question that a preservation programme should treat as central.

Finally, the goal is not fidelity. A perfectly reproduced episode has limited prospective value; what matters is whether its elements remain available for recombination. Preservation should be judged by the futures it makes constructible.

8. Conclusion

The reorientation of memory research towards prospection has replaced the picture of episodic memory as an archive with one of a generative resource. On this picture, the observation that amnesic patients cannot imagine tomorrow, and that depressed patients imagine it only in outline, is a direct expression of the system’s design. The affective findings sharpen the point: emotion is how a simulated event is evaluated, and its vividness depends on the vividness of the memory from which the simulation was built. For a foundation concerned with preserving emotion, the consequence is that emotional memory is worth preserving not only for what it holds of the past but for what it makes possible in the future; the two cannot be separated without losing both.

References

Addis, D. R. (2020). Mental time travel? A neurocognitive model of event simulation. Review of Philosophy and Psychology, 11(2), 233–259.

Addis, D. R., Wong, A. T., & Schacter, D. L. (2007). Remembering the past and imagining the future: Common and distinct neural substrates during event construction and elaboration. Neuropsychologia, 45(7), 1363–1377.

Benoit, R. G., Szpunar, K. K., & Schacter, D. L. (2014). Ventromedial prefrontal cortex supports affective future simulation by integrating distributed knowledge. Proceedings of the National Academy of Sciences, 111(46), 16550–16555.

Buckner, R. L., & Carroll, D. C. (2007). Self-projection and the brain. Trends in Cognitive Sciences, 11(2), 49–57.

D’Argembeau, A., & Van der Linden, M. (2004). Phenomenal characteristics associated with projecting oneself back into the past and forward into the future: Influence of valence and temporal distance. Consciousness and Cognition, 13(4), 844–858.

Gilbert, D. T., & Wilson, T. D. (2007). Prospection: Experiencing the future. Science, 317(5843), 1351–1354.

Hassabis, D., Kumaran, D., Vann, S. D., & Maguire, E. A. (2007). Patients with hippocampal amnesia cannot imagine new experiences. Proceedings of the National Academy of Sciences, 104(5), 1726–1731.

Klein, S. B., Loftus, J., & Kihlstrom, J. F. (2002). Memory and temporal experience: The effects of episodic memory loss on an amnesic patient’s ability to remember the past and imagine the future. Social Cognition, 20(5), 353–379.

Schacter, D. L., & Addis, D. R. (2007). The cognitive neuroscience of constructive memory: Remembering the past and imagining the future. Philosophical Transactions of the Royal Society B, 362(1481), 773–786.

Schacter, D. L., Benoit, R. G., & Szpunar, K. K. (2017). Episodic future thinking: Mechanisms and functions. Current Opinion in Behavioral Sciences, 17, 41–50.

Suddendorf, T., & Corballis, M. C. (2007). The evolution of foresight: What is mental time travel, and is it unique to humans? Behavioral and Brain Sciences, 30(3), 299–313.

Tulving, E. (1985). Memory and consciousness. Canadian Psychology, 26(1), 1–12.

Tulving, E. (2002). Episodic memory: From mind to brain. Annual Review of Psychology, 53, 1–25.

Williams, J. M. G., Ellis, N. C., Tyers, C., Healy, H., Rose, G., & MacLeod, A. K. (1996). The specificity of autobiographical memory and imageability of the future. Memory & Cognition, 24(1), 116–125.