In How Authors' Minds Make Stories (Hogan, Reference Hogan2013) and elsewhere (Hogan, Reference Hogan2017; Hogan, Reference Hogan, Hogan, Irish and Hoganforthcoming), I have argued that, for simulation to fulfill its adaptive functions, it requires some specifiable properties, not all of which are consistently recognized. I take simulation to be (roughly) the mental generation of a particular causal sequence that is, so to speak, semi-perceptual and pseudo-motor. By semi-perceptual and pseudo-motor, I mean that the neural processes underlying simulation overlap significantly with full perceptual experience and/or bodily action. This semi-perceptual and pseudo-motor quality is usually spoken of today in terms of embodiment (see Wojciehowski & Gallese, Reference Wojciehowski, Gallese, Hogan, Irish and Hoganforthcoming). Much recent research suggests that large areas of cognitive and affective processing are at least in part embodied. For example, semantic representations (meanings) of many, perhaps all words are semi-perceptual and/or pseudo-motor. Thus, the meaning of “grasp” is not a combination of abstract semantic primitives, but rather a set of cortical activations related to actions of grasping with one's hands (see Pulvermüller & Fadiga, Reference Pulvermüller and Fadiga2010, p. 355). As semi-perceptual and pseudo-motor, simulation too is, in part, an embodied process. Given the continuity between simulation and semantic meaning, it is unsurprising that simulation is often guided by words, particularly in literature (an idea broadly consistent with Scarry's [Reference Scarry1999] observations about literary imagination). On the other hand, simulation is by no means always literary.
The main (adaptive) function of simulation would appear to be the identification of threats and opportunities, removed from the risks associated with real action. For example, Glug imagines trying to wrest a tasty-looking fruit from Mutt, but after briefly envisioning a battle with his behemoth-like antagonist, he refrains. The risk of acting without awareness of threats is obvious. The risk associated with opportunities would be the “opportunity costs” (in the economic sense) incurred by ignoring potentially more advantageous alternatives. The minimization of dangers and of opportunity costs suggests the value of parametric variations – systematically imagined alternatives – in our simulations regarding physical and social conditions and regarding our own behavior in those conditions.
The preceding comments concern the representational component of simulation, defined by our imagination of different possible circumstances and different possible responses to those circumstances. But representation is not sufficient for simulation to have adaptive benefits. Simulation must activate motivation systems. I may imagine some sequence of events, but this does not mean I will act on that imagination. For example, suppose I know about Glug and Mutt. Unless I have some emotional engagement with their situation, I am unlikely to do anything to forestall Glug's (potentially suicidal) decision to challenge Mutt. Most obviously, I might experience empathic fear (fear for Mutt – perhaps fear that I believe Mutt should be feeling, even if he is not actually feeling it). The same, empathic point applies to myself in the future, as I simulate how outcomes of my actions will affect me (a relation between empathy and prudence is indicated by, e.g., work on the temporoparietal junction; see Soutschek et al., Reference Soutschek, Ruff, Strombach, Kalenscher and Tobler2016 [see also Yong Reference Yong2016]). Having self-empathy means that I feel some version of what I would feel were I actually in the simulated circumstances.
There are various complications here, particularly regarding the intensity of the emotional response, which I will leave aside. However, one complication is important in the present context. Aversive emotional responses to simulated dangers could lead us to avoid simulating potentially threatening experiences, which would largely undermine the benefits of simulation. However, there is evidence that compassionate empathic responses involve the reward system, including dopamine pathways (see Kim et al., Reference Kim, Kim, Kim, Jeong, Park, Son and Ki2009, p. 2079), thus what Panksepp (Reference Panksepp1998) calls the SEEKING system. This reward system involvement, then, promotes engagement with the target's experience – thus, continued simulation – even if this is aversive.
The results for our experience of fictional narrative are straightforward. In The Mind and Its Stories (Hogan, Reference Hogan2003), I isolate several cross-culturally recurring story genres. The structure of any genre is, roughly, as follows: An emotion system or set of emotion systems establishes a goal for a main character or characters. This goal may result from a concern for individual or in-group pride, romantic love (a combination of attachment bonding and sexual desire), or other emotion systems. The main characters face the risk of being irrevocably deprived of their goal (e.g., in heroic tragi-comedy, the [in-group] nation appears to be humiliated by the domination of an enemy). However, they manage to reverse the situation and succeed (often establishing even greater glory for the [in-group] nation than formerly). The reader empathizes with the main characters based on his or her own experience of parallel goals (e.g., advancing in-group status) and his or her (embodied) simulation of those characters' (fictional) actions and experiences. Thus, one's engagement with fictional narratives derives principally from one's real-life motivations as these are activated and guided in empathic simulation.
This returns us to the topic of Dubourg and Baumard's (D&B) article. In How Authors' Minds Make Stories (Reference Hogan2013), I contend that authors produce stories in part by integrating story prototypes (e.g., heroic tragi-comedy) with particular, simulated characters and settings. In earlier work, I paid little attention to settings. However, given the cross-cultural recurrence of quest motifs, with their frequent focus on setting, the preceding analyses would lead us to expect that the same sort of process would explain our engagement with exploring fictional worlds (thus settings). In this case, though, the motivational impetus would be provided by seeking, rather than, say, in-group pride. This is precisely what D&B show.
My aim in these comments has been to support D&B's main argument, but to do so by approaching the topic from a different angle. Specifically, I have tried to outline a general account of simulation and fiction that in effect (retrospectively) predicts their main conclusion. Put differently, I hope to have shown that D&B's treatment of imaginary worlds and the work just outlined on simulation and story universals are ultimately parts of a shared research program. Recognition of this commonality should prove mutually enriching, and aid further study and theorization.
You have
Access
In How Authors' Minds Make Stories (Hogan, Reference Hogan2013) and elsewhere (Hogan, Reference Hogan2017; Hogan, Reference Hogan, Hogan, Irish and Hoganforthcoming), I have argued that, for simulation to fulfill its adaptive functions, it requires some specifiable properties, not all of which are consistently recognized. I take simulation to be (roughly) the mental generation of a particular causal sequence that is, so to speak, semi-perceptual and pseudo-motor. By semi-perceptual and pseudo-motor, I mean that the neural processes underlying simulation overlap significantly with full perceptual experience and/or bodily action. This semi-perceptual and pseudo-motor quality is usually spoken of today in terms of embodiment (see Wojciehowski & Gallese, Reference Wojciehowski, Gallese, Hogan, Irish and Hoganforthcoming). Much recent research suggests that large areas of cognitive and affective processing are at least in part embodied. For example, semantic representations (meanings) of many, perhaps all words are semi-perceptual and/or pseudo-motor. Thus, the meaning of “grasp” is not a combination of abstract semantic primitives, but rather a set of cortical activations related to actions of grasping with one's hands (see Pulvermüller & Fadiga, Reference Pulvermüller and Fadiga2010, p. 355). As semi-perceptual and pseudo-motor, simulation too is, in part, an embodied process. Given the continuity between simulation and semantic meaning, it is unsurprising that simulation is often guided by words, particularly in literature (an idea broadly consistent with Scarry's [Reference Scarry1999] observations about literary imagination). On the other hand, simulation is by no means always literary.
The main (adaptive) function of simulation would appear to be the identification of threats and opportunities, removed from the risks associated with real action. For example, Glug imagines trying to wrest a tasty-looking fruit from Mutt, but after briefly envisioning a battle with his behemoth-like antagonist, he refrains. The risk of acting without awareness of threats is obvious. The risk associated with opportunities would be the “opportunity costs” (in the economic sense) incurred by ignoring potentially more advantageous alternatives. The minimization of dangers and of opportunity costs suggests the value of parametric variations – systematically imagined alternatives – in our simulations regarding physical and social conditions and regarding our own behavior in those conditions.
The preceding comments concern the representational component of simulation, defined by our imagination of different possible circumstances and different possible responses to those circumstances. But representation is not sufficient for simulation to have adaptive benefits. Simulation must activate motivation systems. I may imagine some sequence of events, but this does not mean I will act on that imagination. For example, suppose I know about Glug and Mutt. Unless I have some emotional engagement with their situation, I am unlikely to do anything to forestall Glug's (potentially suicidal) decision to challenge Mutt. Most obviously, I might experience empathic fear (fear for Mutt – perhaps fear that I believe Mutt should be feeling, even if he is not actually feeling it). The same, empathic point applies to myself in the future, as I simulate how outcomes of my actions will affect me (a relation between empathy and prudence is indicated by, e.g., work on the temporoparietal junction; see Soutschek et al., Reference Soutschek, Ruff, Strombach, Kalenscher and Tobler2016 [see also Yong Reference Yong2016]). Having self-empathy means that I feel some version of what I would feel were I actually in the simulated circumstances.
There are various complications here, particularly regarding the intensity of the emotional response, which I will leave aside. However, one complication is important in the present context. Aversive emotional responses to simulated dangers could lead us to avoid simulating potentially threatening experiences, which would largely undermine the benefits of simulation. However, there is evidence that compassionate empathic responses involve the reward system, including dopamine pathways (see Kim et al., Reference Kim, Kim, Kim, Jeong, Park, Son and Ki2009, p. 2079), thus what Panksepp (Reference Panksepp1998) calls the SEEKING system. This reward system involvement, then, promotes engagement with the target's experience – thus, continued simulation – even if this is aversive.
The results for our experience of fictional narrative are straightforward. In The Mind and Its Stories (Hogan, Reference Hogan2003), I isolate several cross-culturally recurring story genres. The structure of any genre is, roughly, as follows: An emotion system or set of emotion systems establishes a goal for a main character or characters. This goal may result from a concern for individual or in-group pride, romantic love (a combination of attachment bonding and sexual desire), or other emotion systems. The main characters face the risk of being irrevocably deprived of their goal (e.g., in heroic tragi-comedy, the [in-group] nation appears to be humiliated by the domination of an enemy). However, they manage to reverse the situation and succeed (often establishing even greater glory for the [in-group] nation than formerly). The reader empathizes with the main characters based on his or her own experience of parallel goals (e.g., advancing in-group status) and his or her (embodied) simulation of those characters' (fictional) actions and experiences. Thus, one's engagement with fictional narratives derives principally from one's real-life motivations as these are activated and guided in empathic simulation.
This returns us to the topic of Dubourg and Baumard's (D&B) article. In How Authors' Minds Make Stories (Reference Hogan2013), I contend that authors produce stories in part by integrating story prototypes (e.g., heroic tragi-comedy) with particular, simulated characters and settings. In earlier work, I paid little attention to settings. However, given the cross-cultural recurrence of quest motifs, with their frequent focus on setting, the preceding analyses would lead us to expect that the same sort of process would explain our engagement with exploring fictional worlds (thus settings). In this case, though, the motivational impetus would be provided by seeking, rather than, say, in-group pride. This is precisely what D&B show.
My aim in these comments has been to support D&B's main argument, but to do so by approaching the topic from a different angle. Specifically, I have tried to outline a general account of simulation and fiction that in effect (retrospectively) predicts their main conclusion. Put differently, I hope to have shown that D&B's treatment of imaginary worlds and the work just outlined on simulation and story universals are ultimately parts of a shared research program. Recognition of this commonality should prove mutually enriching, and aid further study and theorization.
Funding
This research received no specific grant from any funding agency, commercial or not-for-profit sectors.
Conflict of interest
None.