Mental Time Travel? A Neurocognitive Model of Event Simulation.

In the paper

“According to this view, because past and future representations themselves do not differ in structure or content, it is hypothesized that it is the fluency of the constructive process -an attribute of both memory and imagination -that indicates whether or not a given event was previously experienced (Perrin & Michaelian, 2017).”

What the sources say

• Partly supported

  The Conclusion (section 7; page unavailable) says that, on an attributionalist account, more fluent simulations tend to be attributed to past experience, while cautioning that fluency does not guarantee prior experience. The paper does not establish that past and future representations have identical structure or content; its discussion of discontinuism (section 4; page unavailable) reports differences in sensory detail and emotional valence. (LLM, conf=0.86)

  cites Perrin & Michaelian, 2017 sourceExperimental — accuracy not yet measured

In the paper

“By the continuist view, failures of source/reality monitoring occur because remembered and imagined events are often very similar in content and phenomenology, for instance, when imagined events are elaborate, vivid, and plausible (Devitt, Monk-Fromont, Schacter, & Addis, 2016;Garry, Manning, Loftus, & Sherman, 1996;Johnson et al., 1993).”

What the sources say

• Supported

  Experiment 2 explains that source-monitoring errors can occur when the phenomenological characteristics of imagined and real events overlap. The General Discussion links vivid simulation and memory-like perceptual detail to greater source confusion, while Experiment 1 reports that more plausible imagined conjunction events were more likely to be accepted as real. (LLM, conf=0.88)

  cites Devitt, Monk-Fromont, Schacter, & Addis, 2016 source
• Partly supported

  The passage labeled Abstract says imagined and experienced childhood memories can have few qualitative differences, making their sources hard to distinguish. The Procedure asks participants to imagine events clearly and completely, and another passage says imagined events may seem more plausible. The paper does not establish that elaborate, vivid imagined events are what cause source-monitoring failures; its childhood-memory explanation instead suggests that even imagined and genuine memories containing diffuse information may be confused. Page numbers are unavailable in our copy. (LLM, conf=0.85)

  cites Garry, Manning, Loftus, & Sherman, 1996 sourceExperimental — accuracy not yet measured
• Not checked yetcites Johnson et al., 1993 source

In the paper

“The overall proportional resolution of non-personal simulations is also likely lower, with a greater proportion of lower-resolution elements such as general semantic knowledge (Szpunar et al., 2007).”

What the sources say

• Partly supported

  The Discussion (page not provided) says participants relied more on semantic knowledge when imagining Bill Clinton than when envisioning their personal futures. It does not assess the overall proportional resolution of simulations, and its nonpersonal comparison is limited to imagining that familiar individual. (LLM, conf=0.84)

  cites Szpunar et al., 2007 sourceExperimental — accuracy not yet measured

In the paper

“Our neuroimaging studies suggest that the threshold sits somewhere between the simulation of plausible future events and high novel or implausible events, and once reached, DMN activity decreases, while activity in anterior DMN and frontoparietal network increase.”

In the paper

“We observed this phenomenon during simulation of future events involving incongruent (relative to congruent) details (Roberts, Wiebels, et al., 2017), as have others during tasks involving simulation of events in unfamiliar (versus familiar) contexts (Szpunar, Chan, & McDermott, 2009).”

What the sources say

• Not checked yetcites Roberts, Wiebels, et al., 2017 source
• Not checked yetcites Szpunar, Chan, & McDermott, 2009 source

In the paper

“Finally, I suggest that simulation is a general capacity that underpins other domains of cognition, such as the perception of ongoing experience.”

• Also stated in: § 1. Introduction · hedged

  “Finally, I will suggest that simulation may be a general cognitive capacity underpinning other forms of cognition, including perception (Section 6).”

In the paper

“Psychological research has demonstrated that episodic memory and imagination emerge at similar times during childhood (Atance & O'Neill, 2001), and that in adulthood, individual differences in emotion regulation and mental imagery abilities are correlated similarly with past and future events (D' Argembeau & van der Linden, 2006).”

What the sources say

• Not checked yetcites Atance & O'Neill, 2001 source
• Not checked yetcites D' Argembeau & van der Linden, 2006 source

In the paper

“Rather, it is the 'mental' rendering of experience is the most fundamental function of this simulation system enabling humans to re-experience the past, pre-experience the future, and also comprehend the complexities of the present.”

• Also stated in: § 1. Introduction · plain

  “Although this simulation system can be co-opted to mentally traverse time and space, it is the 'mental' rendering of experience that is the key function of this system and crucial to many aspects of higher cognition.”

In the paper

“I also propose that by forming associations between schemas, the simulation system constructs multi-dimensional cognitive spaces, within which any given simulation is mapped by the hippocampus.”

• Also stated in: § 1. Introduction · plain

  “I also propose that this 'simulation system' supports the construction of multi-dimensional cognitive spaces, within which any given simulation can be mapped by the hippocampus (Section 5).”

In the paper

“This proposal has some important implications for the construct of 'MTT', suggesting that 'time' and 'travel' may not be defining, or even essential,”

• Also stated in: § 1. Introduction · hedged

  “This view has some important implications for how we understand MTT, suggesting that 'time' and 'travel' may not be defining characteristics.”

In the paper

“Since the seminal work of Bartlett (1932), numerous cognitive studies have demonstrated that a memory may be missing elements and/or may contain new information (information that is schema-relevant and/or was newly encountered since the original experience; Schacter, Norman, & Koutstaal, 1998).”

What the sources say

• Not checked yetcites Bartlett (1932)
• Not checked yetcites Schacter, Norman, & Koutstaal, 1998 source

In the paper

“However, it is becoming increasingly apparent that the brain's default network is not only involved in MTT into the past and future but also engaged other forms of MTT such as counterfactual thinking (De Brigard, Addis, Ford, Schacter, & Giovanello, 2013), and other forms of cognition, including creativity (Roberts & Addis, 2018), theory of mind (Frith & Frith, 2003), narrative comprehension (Mar,”

What the sources say

• Not checked yetcites De Brigard, Addis, Ford, Schacter, & Giovanello, 2013
• Not checked yetcites Roberts & Addis, 2018 source
• Not checked yetcites Frith & Frith, 2003 source

In the paper

“Because episodic retrieval involves reactivating the elements of a past experience, the exact brain regions involved in a given memory will vary depending on the representational content.”

In the paper

“The episodic memory system supports the cognitive processes that are utilized during future simulation: episodic retrieval to reactivate details and relational processing to integrate details into a novel event representation.”

In the paper

“If people actually retain as few details about their experiences as suggested by the results of a recent study (Misra, Marconi, Peterson, & Kreiman, 2018), then it is not surprising that remembered past events involve some degee of imagination, such as filling in missing details with internallygenerated content including details from related experiences, conceptual information fleshing out aspects”

What the sources say

• Not checked yetcites Misra, Marconi, Peterson, & Kreiman, 2018 source
• Not checked yetcites Bartlett, 1932
• Not checked yetcites Rubin & Umanath, 2015 source

In the paper

“As discussed in Section 2.1, most contemporary researchers appear to consider memory as a cognitive faculty distinct from other cognitive domains.”

In the paper

“Episodic memory provides the raw materials for future imagination, and the distributed storage of episodic memories as elements heightens the accessibility of episodic details.”

In the paper

“For instance, conjunction errors (i.e., the integration of erroneous details into a memory) involve recombinations of details, and there is no reason to suggest this would be any different from the recombination process argued to underlie future imagination (Addis & Schacter, 2012;Schacter & Addis, 2007).”

What the sources say

• Not checked yetcites Addis & Schacter, 2012 source
• Not checked yetcites Schacter & Addis, 2007

In the paper

“(4) Continuums of associative strength and schema-reliance provide a continuist framework that clarifies the involvement of the DMN and related networks, and accommodates both similarities and differences between past and future events;”

In the paper

“The central tenet of the constructive episodic simulation hypothesis is that although the constructive nature of episodic memory results in relatively frequently distortions during retrieval, thereby reducing the veracity of memory, it is nevertheless an adaptive 'design feature' that facilitates the flexible simulation of future events (Schacter & Addis, 2007).”

What the sources say

• Not checked yetcites Schacter & Addis, 2007

In the paper

“Past and future events tend to centre on similar themes, such as social/leisure activities, relationship episodes, successes/failures, illnesses/accidents of the self/others (D' Argembeau & van der Linden, 2004), and feature events from the individual's cultural life script (e.g., in Western societies, typical events include graduating, getting a job, getting married, etc.; Rasmussen & Berntsen,”

What the sources say

• Not checked yetcites D' Argembeau & van der Linden, 2004 source
• Not checked yetcites Rasmussen & Berntsen, 2013 source

In the paper

“Interestingly, the prevailing models of episodic memory as a distinct faculty typically characterize memory as constructive, thus also implying memory is modifiable in ways akin to 'imagination'.”

In the paper

“Much of the literature to date has focused on whether constituent details are episodic or semantic in nature, finding that across past and future events the numbers of episodic details is strongly correlated, as are the numbers of non-episodic details (e.g., Addis et al., 2008).”

What the sources say

• Not checked yetcites Addis et al., 2008 source

In the paper

“(6) I hypothesize that, as a way of interpreting, predicting, and interacting with the world, simulations are situated within complex cognitive 'spaces' that code for multiple dimensions of experience.”

In the paper

“If memory and imagination are manifestations of the same neurocognitive system, then the operational rules of simulation should be the same irrespective of whether the simulation being constructed is a remembered or imagined event”

In the paper

“However, it is difficult (if not impossible) to determine whether any one detail is represented perceptually or conceptually on the basis of verbal descriptions alone, and neuroimaging profiles have not (as of yet) resolved the issue given that episodic and semantic memory engage overlapping DMN regions (Binder, Desai, Graves, & Conant, 2009;Burianova & Grady, 2007).”

What the sources say

• Not checked yetcites Binder, Desai, Graves, & Conant, 2009 source
• Not checked yetcites Burianova & Grady, 2007

In the paper

“Furthermore, just as schemas facilitate the reconstruction of memories, schemas also modify memories to fit with our general understanding of the world (e.g., cultural schemas; Bartlett, 1932) and ourselves (e.g., self schema; Conway & Pleydell-Pearce, 2000), again bringing memories into the domain of imagination.”

What the sources say

• Not checked yetcites Bartlett, 1932
• Not checked yetcites Conway & Pleydell-Pearce, 2000

In the paper

“Thus it may be that one of the only consistent differences between a memory and a vivid imagining is a relative difference in the fluency of construction (Michaelian, 2011), although even fluency differences can dissipate with repeated imaginings (Wiebels et al., 2019).”

What the sources say

• Not checked yetcites Michaelian, 2011 source
• Not checked yetcites Wiebels et al., 2019

In the paper

“Memory scientists have, by-and-large, focused on the ability to remember the ground truth of experience, thus emphasizing a direct causal link to the past that imagined events cannot possess.”

In the paper

“Most cognitive neuroscience theories sit somewhere between moderate discontinuism to moderate continuism.”

In the paper

“As a template, the schema provides predictions about the elements that are/were/will be likely experienced in the past/present/future event, and if those details not readily available (i.e., not currently perceived or activated), the schema guides the addition of relevant conceptual content to the simulation.”

In the paper

“the schema provides a structure in the form of predicted relations between elements; with respect to simulations, it is likely that temporal, spatial and causal associations predominate given that events that unfold over time and in space.”

In the paper

“the hippocampus is critical for mediating the (re)creation of associations at encoding and retrieval.”

What the sources say

• Not checked yetcites Schacter et al., 1998 source

In the paper

“As internallygenerated and/or externally-perceived multimodal information is accumulated and integrated with the event schema, the emerging simulation is held online in working memory (cf.episodic buffer; Baddeley, 2000) enabling it to be successively modified and refined.”

What the sources say

• Not checked yetcites Baddeley, 2000 source

In the paper

“Although concepts and categories bear resemblances to schemas (e.g., being acquired gradually and organised in associative networks), their relative fixedness is a direct result of inter-relationships being based on fixed definitions (i.e., the defining features of a category; Ghosh & Gilboa, 2014) and thus they lack the adaptability necessary for higher-level associative processes such as rel”

What the sources say

• Not checked yetcites Ghosh & Gilboa, 2014 source

In the paper

“Associative strength reflects a change in the weights between elements that occurs simply by virtue of those elements occuring together (cf.Hebb, 1949).”

What the sources say

• Not checked yetcites Hebb, 1949

In the paper

“episodic memories are sets of associated elements that are reconstructed at retrieval (Schacter et al., 1998)”

What the sources say

• Not checked yetcites Schacter et al., 1998 source

In the paper

“associative strength can explain findings that past events are constructed more fluently and quickly than imagined events (Anderson, Dewhurst, & Nash, 2012;Wiebels et al., 2019), given that associative strength is higher for remembered events.”

What the sources say

• Not checked yetcites Anderson, Dewhurst, & Nash, 2012 source
• Not checked yetcites Wiebels et al., 2019

In the paper

“while some regions of the hippocampus are commonly engaged by remembering and imagining, the anterior hippocampus has been found to be more active when imagining versus remembering (Addis & Schacter, 2012).”

What the sources say

• Not checked yetcites Addis & Schacter, 2012 source

In the paper

“Multiple schema will be activated including relevant 'event' schemas, each specifying elements and relations to be incorporated in the simulation, other relevant schemas (e.g., higher-order schemas, such as the self-schema, in which the event schema is nested), as well as 'reasoning schemas' that guide how other activated schemas are (consciously or unconsciously) adjusted, biased and/or integrate”

In the paper

“the cognitive flexibility afforded by schema and relational reasoning may be the key to the adaptive nature of simulation.”

In the paper

“Higher-level forms of relational reasoning, such as analogy and inference, involve mapping correspondences and creating higher-order relations between schemas (i.e., relations between relations; Gentner, 1983;Holyoak, 2012).”

What the sources say

• Not checked yetcites Gentner, 1983 source
• Not checked yetcites Holyoak, 2012 source

In the paper

“schemas not only shape the content of simulations, but also how we reason about the past, present, and future, and how we situate events more broadly with respect to our self and our lives”

In the paper

“This effect is more apparent when the future simulations are specific (Addis, Cheng, Roberts, & Schacter, 2011), are constructed de novo (van Mulukom, Schacter, Corballis, & Addis, 2013) and are especially novel (e.g., implausible future events; Weiler, Suchan, & Daum, 2010).”

What the sources say

• Not checked yetcites Addis, Cheng, Roberts, & Schacter, 2011 source
• Not checked yetcites van Mulukom, Schacter, Corballis, & Addis, 2013 source
• Not checked yetcites Weiler, Suchan, & Daum, 2010 source

In the paper

“the iterative nature of the simulation process enables associative schematic processes to play out: it allows schemas to be modified or created in response to the emerging simulation, and for schemas to fill-in, adjust, reorganize, or completely rework the simulation.”

In the paper

“When imagining non-personal events (i.e., events that do not feature the self), the involvement of self-schemas decreases linearly with psychological closeness to the protagonist (de Vito, Gamboz, & Brandimonte, 2012) as does the engagement of ventromedial prefrontal regions mediating self-schemas (e.g., D' Argembeau et al., 2010;Szpunar et al., 2007).”

What the sources say

• Not checked yetcites de Vito, Gamboz, & Brandimonte, 2012
• Not checked yetcites D' Argembeau et al., 2010 source
• Not checked yetcites Szpunar et al., 2007 source

In the paper

“relational reasoning facilitates the flexible reorganization of schemas and elements to construct new schemas (Holyoak, 2012;Hummel & Holyoak, 2003), and provides structure and coherence (e.g., temporal structure, causal structure, etc.) to our experience.”

What the sources say

• Not checked yetcites Holyoak, 2012 source
• Not checked yetcites Hummel & Holyoak, 2003 source

In the paper

“Rumelhart and Ortony (1977) suggest that elements within schemas are distributions of all possible values of that detail (based on cumulative experience), enabling generation of the most likely or typical 'default' value when necessary”

What the sources say

• Not checked yetcites Rumelhart and Ortony (1977)

In the paper

“differing levels of reliance on schemas may explain some of the apparent differences between different types of simulation”

In the paper

“Although speculative, I hypothesize that the role of the hippocampus is the localization of a given past or future simulation in relevant multi-dimensional cognitive space(s).”

In the paper

“This hypothesis therefore implies that cells that appear to be spatial in nature (e.g., place cells, etc.) may have putative mapping functions that can be utilized within any activated cognitive space(s).”

In the paper

“In their Levels of Processing theory, Craik and Lockhart (1972) proposed that what we think of as memory is actually the processing of information, including the perception of incoming information.”

What the sources say

• Not checked yetcites Craik and Lockhart (1972) source

In the paper

“this nesting is one of the necessary conditions of MTT I suggest in Section 4.1 (along with the simulation having a sufficient proportion of high resolution elements).”

In the paper

“Although the contribution of schemas during remembering is at the lower end of the continuum, schemas nevertheless play a number of roles, including (but not limited to) organizing activated elements, the identification of inaccurate, irrelevant, or missing content, guiding corrections or the filling of gaps, and relational reasoning between the event schema and other relevant schemas such as the”

In the paper

“although the DMN is active, it is not maximally so, as indicated by studies showing activity during future simulation exceeds that during remembering (e.g., Addis et al., 2007).”

What the sources say

• Not checked yetcites Addis et al., 2007 source

In the paper

“it may be that higher-level DMN regions (e.g., angular gyrus, posterior cingulate) accumulate activated elements and hold the emerging simulation online, while iterative interactions with schemas serve to refine the simulation.”

What the sources say

• Not checked yetcites Baldassano et al., 2017 source
• Not checked yetcites Chen et al., 2016 source
• Not checked yetcites Hasson et al., 2015 source

In the paper

“although schema-reliance and associative strength vary across this spectrum of simulation tasks, from past events through to novel or implausible future events, neither process is unique to the simulation of past or future events, and both remembering and imagining is associated with coupling of the DMN and frontoparietal control network (Dixon et al., 2018).”

What the sources say

• Not checked yetcites Dixon et al., 2018 source

In the paper

“By this account, the hippocampus does not construct simulations per se, but rather contextualizes the emergent simulation within broader knowledge structures.”

In the paper

“I have previously suggested that perception is associated with the schema-guided construction of an internal simulation of experience -a process that is fundamentally no different to remembering or imagining events (Addis, 2018).”

• Also stated in: § 4.5 Simulation as a general cognitive capacity · plain

  “In other words, perception is underpinned by the same neurocognitive system as memory and imagination.”

What the sources say

• Not checked yetcites Addis, 2018 source

In the paper

“Neisser (1976) also viewed memory as non-distinct from other aspects of cognition and perception.”

What the sources say

• Not checked yetcites Neisser (1976)

In the paper

“These continuums have been constructed on the basis of findings that reflect the average of multiple simulations, but individual simulations can diverge from this pattern.”

In the paper

“this idea could help to reduce the tension between spatial/relational, and mnemonic/prospective accounts of hippocampal function.”

In the paper

“Such findings fit with the notion that brain networks functionally couple and uncouple as required by task demands, and a number of studies report connectivity between DMN and frontoparietal control network as the need for flexible or creative thinking increases (e.g., Beaty, Benedek, Barry Kaufman, & Silvia, 2015;Roberts & Addis, 2018;Spreng, Stevens, Chamberlain, Gilmore, & Schacter, 2010).”

What the sources say

• Not checked yetcites Beaty, Benedek, Barry Kaufman, & Silvia, 2015 source
• Not checked yetcites Roberts & Addis, 2018 source
• Not checked yetcites Spreng, Stevens, Chamberlain, Gilmore, & Schacter, 2010 source

In the paper

“Their results demonstrate that pre-experiencing vivid future events (i.e., future MTT) is not sufficient to convey the sense that the event will actually occur in one's future without this contextualization with autobiographical knowledge; similar findings are reported for remembered past events (Scoboria et al., 2014).”

What the sources say

• Not checked yetcites Ernst, Scoboria, and D'Argembeau (2019)
• Not checked yetcites Scoboria et al., 2014 source

In the paper

“past events will, on average, contain a higher proportion of strongly-associated elements than most future events and are therefore constructed more fluently.”

In the paper

“Over a lifetime of experience and reflection, we construct and continually refine distributions of various aspects of the self (e.g., one's life story, trait self-knowledge, etc.), meaning that the self is the most elaborate knowledge structure we possess.”

In the paper

“holding a larger collection of singular elements online (as is more likely the case with future events) is a more intensive operation than holding a few chunks of associated elements (as is more likely for past events), thus placing greater demands on the DMN.”

In the paper

“To my knowledge, no neuroimaging study has directly examined recast events, however, this phenomenon has been observed in patients with semantic dementia who can recall past events and recast past events into the future, but are unable to imagine future events (Irish et al., 2012a), exemplifying the similarity of remembering and recasting.”

What the sources say

• Not checked yetcites Irish et al., 2012a

In the paper

“Moreover, as the simulation is iteratively constructed, each version would need to be situated in relevant cognitive spaces by the hippocampus if latent inferences are to be leveraged in refining the simulation.”

In the paper

“Moreover, 'predictive error' signals when a schema fails to provide an adequate model and is then either switched or modified via relational reasoning processes.”

In the paper

“Interestingly, Neisser's (1976) approach, and my own proposal herein, can also be considered variants of predictive coding theory.”

In the paper

“Third, memory and imagination are subserved by the same neurocognitive system: the DMN, with the dynamic involvement of the frontoparietal control network (with the salience network interfacing the two) as required.”

In the paper

“Simulation is the ongoing process of interpreting and understanding our own realities; each individual's unique experiential history modifies their schemas as well as their idiosyncratic corpus of previouslyencountered elements (including the associative strength between said elements), and moulds the multidimensional cognitive spaces in which these simulations are situated.”

In the paper

“My proposal is built on predictive coding principles: an ongoing, dynamic bidirectional interactivity between schemas (models of prior experience) and the emergent simulation whereby schemas shape the simulation of experience, and the simulation modifies schemas.”

In the paper

“Not only is 'time' a non-essential aspect of MTT given that simulations of perception reflect the present, it also obviates the need for 'travel' as simulating current experience does not remove one entirely from their present.”

In the paper

“My reformulation of the constructive episodic simulation hypothesis argues for a more continuist position than we have done previously by arguing that memory and imagination are the same process underpinned by the same neurocognitive system.”

In the paper

“Predictive coding proposes interactions between an internal model of experience and actual experience, with the difference between the two ('prediction error') used to revise the model thus improving its predictive power during subsequent experiences.”

What the sources say

• Not checked yetcites Dayan, Hinton, & Neal, 1995 source

In the paper

“Simulation is an inherently constructive process governed by associative processes operating at different scales: associations (of varying strength) between elements; associative schematic processes that enable construction of a suitable schema (e.g., relational reasoning), as well as the dynamic interplay between schemas and the simulation itself (cf.predictive coding)”

In the paper

“Simulations vary in proportional resolution (i.e., the balance of elements at various resolutions), elaborateness, and complexity.”