What is it (like) to imagine an emotion?

The paper asks: what does it mean to imagine an emotion? While much work examines how emotions arise during imagination (e.g., responses to fiction) or how imagined emotions differ from real ones, the authors step back to distinguish different ways imagination and emotion relate. They argue that conflating distinct kinds of imaginative engagement generates confusion because each likely recruits different cognitive systems and has different psychological and epistemic profiles. They propose that, beyond imagining that an emotion occurs or having emotions in response to imagined scenarios, there is an underexplored mode—emotion-like imagination—where one re-enacts the emotional state so that the phenomenal properties of the emotion are the main content of imagination. They aim to articulate this notion, argue for its cognitive plausibility via emotion regulation mechanisms, and indicate its importance for affective forecasting and empathic understanding.

The authors situate their work within debates on imagination and emotion, especially the paradox of fiction and recreative imagination (Currie and Ravenscroft, 2002; Walton, 1990; Gendler and Kovakovich, 2006; Currie, 2020; Kind, 2016; Liao and Gendler, 2019). Currie and Ravenscroft deny the existence of emotion-like imaginings, claiming emotions are transparent to imagination (we have real emotions in response to imagined content) and highlighting evolutionary advantages of emotional sensitivity to imagined scenarios. The authors argue this supports emotional imagination (emotions as responses to imaginings) but does not preclude emotion-like imagination targeting emotional phenomenology as imagined content. They draw on philosophical work on emotional intentionality and formal objects (Kenny, 1963; Teroni, 2007; Goldie, 2002; Kriegel, 2014) and on features of affect such as valence, bodily sensations and arousal, and motivation (Charland, 2005; Carruthers, 2017; Colombetti and Harrison, 2018; Corns, 2014; Scarantino, 2014). They also note Goldman (2006) as an exception acknowledging experiential imagining of emotions, though lacking a detailed account. Empirically, they frame realizability through the emotion regulation literature (Gross, 1998, 2015; McRae and Gross, 2020), including critiques of compartmentalized models (Barrett et al., 2014), and research on instrumental motives in ER (Tamir et al., 2008; Tamir and Ford, 2012).

This is a conceptual/philosophical analysis supplemented by an integrative theoretical appeal to empirical work on emotion regulation. The approach involves: 1) Proposing a typology of imagination–emotion relations: (i) belief-like imagining emotions (B-Im: imagining that an emotion occurs), (ii) emotional imagination (E-Im: emotional responses to imagined content), and (iii) emotion-like imagination (X-Im: imaginatively re-enacting an emotion's phenomenal properties as the primary content). 2) Providing conceptual criteria distinguishing X-Im from B-Im and E-Im: content focus (emotional phenomenology vs. propositions or objects), mapping relation between target emotion and imagined state, and agency/voluntariness (endogenous, will-involving generation). 3) Analyzing the constitutive features required for X-Im: phenomenal valence (pleasantness/unpleasantness or value), bodily/interoceptive sensations and arousal, and motivational push; and the intentional structure (particular and formal objects), highlighting the need for an intentional base (perception, memory, or other imaginings) to anchor and sustain the state. 4) Arguing for cognitive realizability by reverse-engineering emotion generation via emotion regulation strategies: situation selection/modification, attentional deployment, cognitive change (reappraisal), and response modulation (e.g., breathing, posture, facial expressions). The proposal is that, in X-Im, agents strategically modulate appraisals, attention, bodily responses, and imagery to elicit the target emotional phenomenology. 5) Illustrative case analyses (e.g., imagining being the angry Napoleon at Waterloo) demonstrate iterative, poly-regulated scaffolding (posture, sensory imagery, reappraisal, bodily cues) to enact specific emotion-like states. 6) Exploring applications to affective forecasting and empathic understanding, positing that X-Im can be used strategically to summon motivational forces for decision-making and to access others’ affective perspectives when E-Im fails due to divergent baseline reactions.

- Typology: Three distinct relations between imagination and emotion are articulated: B-Im (imagining that an emotion occurs), E-Im (emotions as responses to imagined content), and X-Im (emotion-like imagination: imaginatively re-enacting emotional phenomenology as the content). - Core features of X-Im: It targets emotional phenomenal properties (valence, bodily sensations/arousal, motivational force) and typically requires an intentional base (perceptual or imaginative) to anchor the emotion’s particular and formal objects. X-Im succeeds only if the evoked emotion matches the targeted one, unlike E-Im where any emotional response suffices. - Agency and generation: X-Im involves endogenous, will-involving processes—agents are causally efficacious in generating the emotional state for representational purposes—contrasting with standard emotions (often exogenously triggered) and with E-Im where agents control the imagined content but are patients of whatever emotion ensues. - Transparency qualified: Emotions can be transparent to imagination (imaginings can yield real emotions), yet this does not negate X-Im; rather, successful X-Im re-instantiates genuine emotional phenomenology while remaining an imaginative act due to its generative, representational intent. - Cognitive realizability via Emotion Regulation: Existing ER research supports that people can regulate towards target emotions for instrumental goals (e.g., eliciting anger before confrontation, enthusiasm before collaboration). X-Im is modeled as reverse-engineering emotion generation by combining ER strategies (situation selection/modification, attentional deployment, cognitive change, response modulation) and imagery to inflate and stabilize the target state. - Constraints: Minimal, objectless valence can sometimes be summoned, but robust, complex X-Im typically requires intentional anchoring and multimodal scaffolding. - Applications: X-Im can enhance affective forecasting by strategically summoning known affective states to motivate goal-consistent behavior, and can support empathic understanding by enabling first-personal re-enactment of others’ emotions when one’s default E-Im response diverges (e.g., understanding fear of dogs despite one’s own positive reactions).

By distinguishing B-Im, E-Im, and X-Im, the paper clarifies what it means to imagine an emotion and prevents conflation across debates on fiction, simulation, and affect. The proposal addresses the research question by arguing that a genuine mode of imagining emotions focuses on experiential properties as content and can be intentionally generated via endogenous control. This reframes the Currie and Ravenscroft transparency thesis: while imaginings can indeed produce real emotions (E-Im), there remains a distinctive, representationally targeted process (X-Im) where the agent aims to instantiate a specific emotional phenomenology. The intentional-base requirement situates X-Im within broader theories of emotional intentionality and explains why anchoring to particular and formal objects is typically necessary for intensity and stability. The significance lies in expanding theoretical resources for understanding how imagination can shape motivation, decision-making, and social cognition. Practically, X-Im offers mechanisms for leveraging emotions instrumentally (e.g., motivating exercise through imagined accomplishment) and for empathically adopting perspectives that one’s spontaneous emotions would not reveal (e.g., seeing a dog as terrifying from another’s viewpoint).

The authors claim to have: (i) provided a useful typology of imagination–emotion relations (B-Im, E-Im, X-Im); (ii) highlighted and characterized the neglected category of emotion-like imagination (X-Im); (iii) offered a conceptual scaffold for X-Im’s constitutive features and a plausible account of its cognitive realizability via emotion regulation strategies; and (iv) sketched theoretical and empirical avenues where X-Im can enrich accounts of affective forecasting and empathic understanding. Future research directions include empirical investigation of X-Im processes and efficacy, the role of different intentional bases and ER strategy combinations, boundary conditions for objectless X-Im, and applications in training, therapy, and social perspective-taking.

- Empirical gap: No existing ER studies directly target or operationalize X-Im; the account extrapolates from adjacent ER findings and offers a theoretical model needing empirical validation. - Intentionality dependence: It remains an open question whether X-Im can be sustained without an intentional object; objectless valence appears weak and short-lived. - Scope and generalizability: The analysis is primarily conceptual and illustrative; it does not specify individual differences (e.g., imagery ability, interoceptive sensitivity) or boundary conditions for complex emotions (e.g., mistrust, nostalgia). - Process model caveat: While the ER process model is used heuristically, emotions may be more dynamically constructed than stage-like; the proposed reverse-engineering route is a simplification. - Volitional control variability: Not all individuals may reliably generate or control target emotions endogenously; factors like context, training, and psychopathology could limit applicability.