Memory of Existing Objects as a Fundamental Idea in the Metaphysics and Ontology of Fictional Entities

Introduction

Kroon and Voltolini (2018) use the term nonexistence datum to suggest that iconic fictional figures such as Hamlet and Holmes do not exist. To substantiate this idea, the authors observe that most people with no specialized knowledge about the matter would assume that fictional entities do not exist as real-world objects, or would feel that "there is at least some sense in which they don't" (p. 3). However, Kroon and Voltolini (2018) identify two different philosophical positions on the existence of fictional characters: fictional antirealists, who believe that among all the things that do exist, there are no fictional entities, and fictional realists, who believe that fictional entities do exist.

This realist versus antirealist debate dates back to the origins of analytic philosophy. Most philosophers accept that the debate likely began with the disagreement between Meinong (1904/1960), a realist, and Russell (1905a, 1905b), an antirealist. According to Meinong, fictional entities exist; moreover, it is possible to study these fictional objects and discover facts about them. In contrast, Russell believed that accepting the existence of fictional entities transgressed fundamental logical principles.

This paper draws on neuroscientific studies that suggest that memory and imagination are closely related. Specifically, there is a scientific consensus that imagination involves the rearrangement of objects or details obtained during previous episodic or perceptual experiences into innovative representations. Following this notion, I propose a fictional realist metaphysical and ontological account of fictional entities. According to this account, fictional entities exist as hylomorphic artifacts composed of the creator's multimodal (e.g., visual, auditory, and tactile) perceptual memory images and recollections of previously encountered real objects. Specifically, fictional entities comprise rearrangements of the memories of different real objects. Thus, while real objects are perceived and imaged mentally, it is the creative reworking of these images that forms the basis of fictional entities, emphasizing their nature as abstract artifacts distinct from the real objects themselves. Fictional entities are individuated by the author's unique memories (e.g., autobiographical), the person-specific imagery that results from these recollections, and the author's intentions. Therefore, I suggest that fictional entities come into being when the creator synthesizes, disassembles, completes, or modifies the mental images resulting from memory retrieval into a new fictional entity. This proposal implies that these fictional entities may cease to exist at a later point in time; for example, if all the copies and records of Homer's Odyssey were destroyed or irrecoverable and Odysseus was completely forgotten. Thus, I propose that fictional characters are ontologically dependent; they could not exist were it not for the existence of objects perceived, imaged, and rearranged by the creator, the works that tell their stories, or a mental state (e.g., thinking about or remembering them).

Philosophical inquiry regarding fictional entities entails two key questions (e.g., Thomasson 1999, Kroon and Voltolini 2018): the first is a metaphysical question about what these entities are or are like, and the second is an ontological question relating to whether there are any such entities. A useful approach to these questions highlights the connections of metaphysics and ontology with science (van Inwagen and Sullivan 2020). Thus, I begin by presenting a general survey of current neuroscientific theories of the human imagination (Section 2). I then detail the proposed metaphysical (Section 3) and ontological (Section 4) accounts outlined above. I conclude by addressing two theories that relate to my theory: van Inwagen's (1990) theory of "anticonventionalism" (p. 3) and Everett's (2013) "pretense theoretic." Like Thomasson's (1999, 2003a, 2003b) account, I propose a fictional realist stance. I also argue that fictional entities are cultural artifacts. Opposing Platonist views of abstract entities, I agree with Thomasson (1999, 2003a, 2003b) that fictional entities exist and are artifactual because they are not found; they are created at a specific moment by a creator. Thus, fictional entities are contingent and not necessary entities that may have never existed had they not been produced.

What Is Imagination?

There is a consensus within neuroscience that the function of memory is not only to remember the past but also to allow us to rearrange components of past experiences and be better prepared to act if we face unexpected circumstances in the future (e.g., Addis et al. 2007, Schacter et al. 2008, Szpunar 2010, Gaesser 2013). According to neuroscientific accounts of the relationship between imagination and memory, imagination involves the rearrangement of details obtained during past episodic experiences into innovative representations of future events (Schacter et al. 2008, Buckner 2010). This interplay between memory and imagination provides an opportunity to "test out" different accounts of what might happen by simulating the outcome of future events based on past lived experiences (Gaesser 2013, p. 1). Imagination thus functions as a survival mechanism, preparing individuals for action when faced with new situations by simulating such situations based on past perceptions and memories. This aligns well with Schacter's (2012) suggestion that the purpose of simulating future events is to allow individuals to plan or predict experience-based actions without the risk of execution and to act appropriately in unexpected circumstances.²

The difference between "imagery" and "imagination" should be clarified at this point. Imagery refers to the ability of neural circuits to generate a representation of an object that is not currently present in the person's sensory environment, but which they have experienced before (Agnati et al. 2013). Imagery is not only visual; it involves representations of other types of perceived sensory information, depending on the sensory nature of the experience, as well as previously viewed objects (Zvyagintsev et al. 2013). Most research has focused on visual, auditory, tactile, and sensorimotor imagery (Kosslyn et al. 2001, Nanay 2018). However, other interesting forms of imagery have been identified, such as emotional mental imagery (Lang 1977, 1979, Ji et al. 2016) and narrative imagery (Sabatinelli et al. 2006). Emotional mental imagery involves semantic or "meaning" representations.³ Narrative imagery comprises mental images of emotional scenarios that can elicit emotional responses similar to those that arose when the scenario was first experienced (Kreiman et al. 2000, Kosslyn et al. 2001). Many studies suggest that image production results from a combination of perceptions (see review by Nanay 2018). Thus, the term multimodal mental imagery is now used to indicate that perceptual processing in one sense modality may be generated or complemented by sensory stimulation in another sense modality (Nanay 2018).

In contrast to imagery, imagination is the ability of neural circuits to rearrange images from perceptual experiences to produce original images (Agnati et al. 2013).

Humans can imagine novel images both while awake and while dreaming. Research has shown that imagination during these two states results in similar novel mental images (Vyshedskiy 2019). Both processes result in the construction of a "novel fantasy," and have been termed "imaginary experience" or "constructive imagination" (Vyshedskiy 2019, p. 89). Although waking and sleeping imaginative processes produce the same end products, they differ in the neurophysiological mechanisms that generate a new mental image. Vyshedskiy (2019, p. 89) refers to the neural mechanisms responsible for imagination during rapid eye movement (REM) dreaming as bottom-up mechanisms, and to the mechanisms of conscious purposeful imagination as top-down mechanisms.

By examining and comparing imaginative processes during sleep and waking states, neuroscientists have identified at least six neurophysiological mechanisms responsible for bottom-up and top-down imagination. Top-down mechanisms include prefrontal synthesis, prefrontal analysis, and the integration of modifiers and mental rotation mechanisms. Prefrontal synthesis is the spatial combination of two or more objects from memory into a novel mental image in the prefrontal cortex. Prefrontal analysis is defined as the disassembly of an object of memory into parts in the lateral prefrontal cortex, and modifier integration refers to activity in the lateral prefrontal cortex that allows the use of imagination to modify a real object (e.g., imagining your car being a different color) (Vyshedskiy 2019). Bottom-up mechanisms include REM sleep dreaming, amodal completion, and categorically primed spontaneous imagination. During REM sleep dreaming, previous memories are activated to mediate perceptual objects. The process of amodal completion involves the completion of incomplete objects appearing in dreams using the memories of complete objects similar to ones previously perceived.

In addition to these bottom-up and top-down mechanisms, there is a hybrid mechanism that serves both bottom-up and top-down imagination. This is called categorically primed spontaneous imagination and is the generation of fantasies about different categories of interest in memory (Vyshedskiy 2019). Some of the best examples of categorically primed spontaneous imagination involve obsessions, such as impulsive fantasizing about food, sex, or drugs, or anxiety caused by compulsive thoughts about possible traumatic events (Vyshedskiy 2019).

This is necessarily a partial review of the neuroscience of imagination literature. Many issues in the neuroscience of imagination remain unclear, and there is much disagreement about the topic. However, the important scientific feature of my proposal is that the mental imagery of retrieved memories of perceived existing objects is used by imaginative structures to generate imaginary objects or representations. As Agnati et al. (2013) express it:

Metaphysical Proposal: What Exactly is a Fictional Entity?

Drawing on the above-mentioned accounts of the memory-imagination relationship and associated mechanisms, I suggest that imagination is an activity by which humans synthesize, disassemble, complete, or modify remembered images of multimodal perceptual experiences of real objects. I propose that fictional entities derive from this activity. In other words, fictional entities are the consequence of the shaping and rearrangement of imagery.

As mentioned above, this is a fictional realist proposal similar to Thomasson's (1999, 2003a, 2003b) accounts. I agree with Thomasson that fictional characters are abstract cultural artifacts, similar to works of literature, music, and law, among others. Thomasson (1999, 2003a, 2003b) is correct in stating that fictional entities are artifactual because they are not discovered. Abstract entities are not simply found or selected from the realm of eternal ideas; they are created at a particular moment by a creator. Fictional entities are contingent entities that depend on their creation.

Although I agree with Thomasson (1999, 2003a, 2003b) that fictional entities exist and are artifacts, there is a fundamental difference between Thomasson's (1999, 2003a, 2003b) proposal and my own. This difference relates to my response to Brock's (2010) objection that fictional characters are not artifacts because there is no good answer to the question of when a fictional entity begins to exist. Thomasson's (1999) view is that fictional entities are created simply by words that construct them in a certain way. For example, Jane Austen brings into existence the fictional character Emma Woodhouse when she writes, "Emma Woodhouse, handsome, clever, and rich, with a comfortable home and happy disposition, seemed to unite some of the best blessings of existence; and had lived nearly twenty-one years in the world with very little to distress or vex her" (Austen 2004, p. 55). In contrast to Thomasson, I suggest that fictional entities are created when the author rearranges the memories of perceived objects into a character and refers to the parts of the rearranged things. In other words (for example), Austen would have to have perceived a woman, handsomeness, cleverness, and wealth and combined these memories into Emma Woodhouse, an attractive, intelligent, and rich young woman. Thomasson (1999) is correct in stating that fictional characters depend on the description of their properties. However, we cannot ignore the fact that this final descriptive process depends on the reworking of memories; thus, fictional entities cannot be created merely by the words that describe them.

In Making Objects and Events: A Hylomorphic Theory of Artifacts, Actions, and Organisms, Simon Evnine (2016) presents a thoughtful and interesting version of fictional entities. He suggests that fictional entities are "hylomorphically complex, artifacts that have properties, or sets of properties, as their matter" (Evnine 2016, p. 141). In other words, creators generate fictional entities out of their properties. Thus, Evnine (2016 p. 70) is undoubtedly correct when he considers artifacts to form "a distinct sui generis category of objects for which what it is to be an object of that kind is to be the impress of a mind on matter." I would like to expand this idea by stating that the imaginary beings that are fictional entities also result from the deliberate deforming or mangling of previous experiences by a creator. However, this does not contradict Evnine's reasonable idea that artifacts derive from the mind (in the case of fictional entities, as reshaped memories) impressing itself on matter.

Therefore, Evnine (2016) is surely correct that fictional entities are hylomorphic. They are made from something. He is also right in assuming that they are artifacts, as minds do seem to "impress" themselves on matter "through the process of working on the matter with certain intentions" (Evnine 2016, pp. 70-71). As stated above, from an evolutionary perspective, we simulate the future according to experience-based imagery to be prepared to act with the intention of surviving. However, I propose a different view to Evnine's (2016) regarding the matter, individualization, and creation of fictional entities. First, I believe that the fundamentals of these entities are not properties or sets of properties. Rather, the matter of fictional entities consists of the implicit mental imagery or images that arise from memories of perceptual information. The matter of a fictional entity is the mental imagery of remembered experiences. This account of how fictional entities are created assumes that the matter or raw materials used by the creator are the objects or episodic memories that she transforms into novel entities or characters. Thus, I suggest that it is the memories of real-world objects and experiences that are transformed and abstracted, and that constitute the matter of fictional entities, which transcend the original tangible objects and experiences.

At this point, a philosophical opponent may argue that a fictional character is conceived rather than imagined, similar to how time and mathematical concepts are articulated. This objection can be addressed by showing that both processes depend on perception and the memory of these perceptions. In an excellent review of the neuroscience of this topic, Allman et al. (2012) explain how the mode-control model of counting and timing draws together scientific findings under a unified theory of numerosity and timing. A full explanation of the model is beyond the scope of this paper. Briefly, the model is based on substantial scientific evidence suggesting that time and number share a common magnitude neurological representation system. Importantly, for my purposes here, non-humans and infants appear to be capable of making magnitude estimates of duration, numerosity, and area through perception and the memory of perceptual events; for example, understanding that a stimulus may last for more or less time than another, or may contain more or fewer elements than another. This perceptual and perceptual memory-based sense of magnitude functions as the foundation of advanced numerical and temporal competence in adulthood. I interpret these scientific results as showing that both created fictional entities and conceived timing and mathematical concepts depend on perception, the memory of these perceptions, and some form of remembered image rearrangement. From a purely philosophical standpoint based on these scientific ideas, my response to this philosophical opponent is that our creation of time, numbers, and indeed fictional entities requires perceptual memories. They are both forms of imagination as I describe it here. The difference is that time and numbers seem to require a sense of magnitude when reworking image memories.

Regarding the individualization of fictional entities, Evnine (2016) suggests that fictional characters are not individuated by the properties he considers to be their matter but by the acts of their creation and, consequently, the intentions of their authors. I suggest that fictional entities are not individualized by the act of creation, as all these imaginative undertakings seem to involve synthesizing, disassembling, completing, or modifying mental images. In other words, the creative act is the same; all humans share the neurophysiological mechanisms of imaginary creation described above. I propose that fictional entities are individualized by the creator's unique memories, particularly their inherently individualized autobiographical memories. However, as no one else has access to the memories of the creator of a fictional character, I do not imply that each person has a different fictional character in mind. I believe that individualization must be completed by what Thomasson (1999) considers the act of creation: the act by which the creator makes the fictional entity a certain way. In other words, I do not agree with Thomasson's (1999, p. 12) view that Jane Austen creates a fictional character, Emma Woodhouse, when she describes her as "handsome, clever, and rich...." On the contrary, I argue that this creation occurs in the reworking of Austen's memories. The individualization of Emma Woodhouse begins with Austen's unique memories, which become the final entity when rearranged. The individualization process culminates when Austen conceives the words that make Emma Woodhouse out to be a certain way.

This finalizing of the individualization process allows us to have an agreed object in mind when we think of Emma Woodhouse. This object in mind is "agreed" because, according to the view presented here, it is not realistic to assume that two readers of Austen's novel have precisely the same object in mind when they think of Emma Woodhouse. Because this memory-imagination relationship depends on memories unique to different readers, it is impossible that these individuals experience the same imagined object. Nevertheless, shared cultural experienced episodes likely ensure that our two readers experience quite similar imagined objects. For example, if a group of people were asked to draw or describe a unicorn or a centaur, it is likely that their accounts would be very similar.

In other words, all fictional entities are ultimately created by the rearrangement of stored perceptual information in new ways. What are highly subjective and individualized are the perceptual experiences and the content of the imagery ultimately rearranged in the creation of the entity. The imagery content, which is highly subjective and environment- or context-based, individualizes the fictional entity. The stored memories that result from the subjective perception of the world and the creator's own self-awareness individualize the fictional entity. Therefore, the compositional process of imagining a fictional entity uses a selection of images or concepts that are individualized mainly by the creator's specific autobiographical memories.

Concerning the creation of the artifact, Evnine (2016, p. 144) is again correct in rejecting the Platonist stance that fictional entities are universal, eternal, necessary beings when he states that the fictional entity "comes into being" through human creation. More specifically, Evnine (2016) suggests that creation occurs when the creator selects the properties that will constitute the matter of fictional entities. It is true that these fictional artifacts are contingent and intentionally created. However, the entity comes into being when the creator synthesizes, disassembles, completes, or modifies multimodal perceptual experiences. As mentioned earlier, in line with Thomasson (1999, 2003a, 2003b), I argue against Platonist views by proposing that these fictional entities come into being at one time, the time of the rearrangement of stored imagery, and cease to exist at another time, when there are no records of them and they are forgotten.

I propose that fictional entities arise from the cognitive reworking of stored memories, presenting a scientifically grounded alternative to theories based on pretence. Although Thomasson (1999, 2003a, 2003b) and Braun acknowledge the importance of memories in creating fiction, I extend this position by emphasizing that the reworking of memories can spontaneously create fictional contexts without the need for explicit creative intentions.

It is important to differentiate between traditional, intentional fiction creation and the spontaneous emergence of fictional contexts through cognitive processes. While traditional fiction typically involves deliberate intent, fictional characters can also emerge through spontaneous creation, where the brain's ability to synthesize and reorganize information results in narrative-like scenarios without explicit intent. This process can produce rich, detailed, coherent scenarios resembling fictional stories, even if they are not intentionally crafted as fiction. For instance, a person might spontaneously imagine a detailed scenario of future events, complete with characters and plotlines, without the conscious intention to create a fictional story. These spontaneously generated narratives exhibit the same underlying cognitive processes of imagination, memory reorganization, and contextual embedding as intentional fiction.

Recent studies confirm the existence of spontaneous (involuntary) future thinking, in which mental images of the future arise unintentionally, often triggered by incidental environmental stimuli, while a person is engaged in routine activities (Kvavilashvili and Rummel 2020). Such spontaneous thoughts frequently occur in response to unrelated triggers, such as seeing a train station on TV and thinking about needing to buy a train ticket tomorrow. Purchasing the ticket tomorrow is a fictional entity, much like Sherlock Holmes, which occurs without intention and is only possible by rearranging episodic memories of ticket purchasing. Spontaneously imagined fictional context is also unintentional and based on the rearrangement of previous episodic contexts.

This research indicates that the spontaneous rearrangement of memories can generate complex, imaginative scenarios without deliberate intent. This process mirrors how traditional fiction is created but operates at a different cognitive level, emphasizing the brain's inherent capacity for narrative construction.

Therefore, although the author's creative intentions are important in traditional fiction, they are not the sole contributors to the creation of fictional entities. The primary mechanism is the spontaneous reworking of episodic memories. The subjective nature of these memories and their rearrangement based on the individual's perceptual experiences and self-awareness lead to the creation of highly individualized fictional entities independently of deliberate fictional intentions.

Waller and Brager (2022) show that various factors, such as the availability of choices, the emotional valence of outcomes, and social and environmental contexts, can influence people's sense of agency. They explore two neuroscientific paradigms to discuss intentions or aspects of agency: the intentional binding paradigm and the Libet paradigm. The intentional binding paradigm, a cognitive neuroscience concept, refers to the subjective shortening of the time between a voluntary action and its external consequence. The Libet paradigm, which is based on a seminal study in the field of neuroscience, suggests that conscious intentions do not generate our actions. Libet and his team (1983) demonstrated that the brain's readiness potential, which indicates unconscious preparation for movement, occurs approximately one-third of a second before participants consciously intend to move their wrists. They interpreted these results as suggesting that the brain's unconscious processes determine the wrist flexion before the conscious intention occurs. This has important implications for the creation of fictional entities, as it implies that our conscious intentions may not be the sole drivers of our creative processes.

It is important to note that the underlying mechanism of how a brain state produces abstract artifacts, such as fictional entities, need not be explicitly detailed to validate the theory of ontological dependence on mental states. The intentional binding research described above highlights the complexity of our sense of agency and creative intentions, which can be both explicit and implicit, and influenced by various factors. This complexity suggests that the creation of fictional entities may be underpinned by intricate, multifaceted mental processes, some of which may not be intentional.

Moreover, the intentional binding paradigm demonstrates that our perception of agency can be influenced in subtle ways, and that we may only sometimes be consciously aware of the origins or motivations behind our actions, including creating abstract artifacts like fictional characters.

Although the precise causal mechanisms linking reworked memories to abstract artifacts remain to be fully elucidated, these processes suggest a metaphysical bridging where cognitive activities scaffold the creation of independent fictional entities. This stance, informed by the intentional binding paradigm and Libet's paradigm, acknowledges the complexity of cognitive processes and the creation of fiction. It suggests that our understanding of agency can be enriched by these paradigms, providing a nuanced view of how fictional entities emerge from mental activity, even without explicit intention.

Ontological Arguments for the Existence of Fictional Entities

To summarize, my metaphysical proposal as described in Section 2.2 is that fictional entities are hylomorphic artifacts that are comprised of episodic memories. The matter of fictional entities is constituted by these recollections; this is what differentiates my proposal from Evnine's (2016) idea that fictional entities are hylomorphic artifacts that have properties as their matter. In contrast, I suggest that fictional entities are composed by the rearrangement of the images of objects in the creator's memories. Fictional entities are individuated both by the creator's unique memories and personal imagery that result from these recollections, and the creator's intention. Finally, I suggest that fictional entities come into being when the creator synthesizes, disassembles, completes, or modifies the mental images that result in the new fictional entity, and that they can also cease to exist.⁴

This leads to a realist ontological view that assumes that fictional entities exist. However, they exist under an ontologically dependent condition; they would not have existed were it not for the existence of some real existing objects perceived, imaged, and rearranged by the creator, the existence of the works that tell their stories (e.g., books, comics, films), or the mental states of those consuming them (e.g., thinking about or remembering them). Again, this assumption that fictional characters exist is similar to Thomasson's (1999, 2003a, 2003b) ontological view of fictional characters. However, I believe that for Thomasson, these entities are ontologically dependent only on the existence of the works that tell their stories; for example, the words that Austen wrote to render Emma Woodhouse a certain way. Although I agree with this assumption, I argue that these works would not exist without some real, existing, antecedent objects or individuals perceived, remembered, or imaged by the creator and rearranged into the fictional entity.

I suggest that, contrary to Evnine (2016), the matter of fictional entities consists in the actual objects or individuals perceived, later remembered, imaged mentally, and then rearranged, not just certain properties selected by the creator. This idea is important in analysing some of the semantic arguments that have been proposed for the existence or non-existence of fictional entities. Although from a semantic standpoint the realism-antirealism debate began with the disagreements between Meinong (1904/1960) and Russell (1905a, 1905b), Frege (1948) set the tone when he stated that the fictional name Odysseus had a sense but no reference.⁵ The semantic argument for or against fictional realism centres on the question of whether or not the name of a fictional entity refers to something that exists in the real world.

I have highlighted some of the similarities and differences between my proposal and Thomasson's (1999, 2003a, 2003b) fictional realist proposal. My semantic arguments differ substantially. As mentioned earlier, Thomasson (1999) argues that fictional characters such as Emma Woodhouse are created when the author writes a sentence that shows them to be a specific way. Thomasson (1999) suggests that when Jane Austen writes the above-cited sentence describing Emma Woodhouse, she pretends to refer to a person and not a fictional entity. Through this referencing process, Austen creates a fictional character, Emma Woodhouse (Thomasson 1999). Emma Woodhouse thus exists as an abstract artifact and not as a pretty, intelligent, and rich twenty-one-year-old woman.

Brock (2010) calls this the Pretend Reference thesis and points out some potential problems with it. In The Nonexistent, Anthony Everett (2013, p. 3) argues against realist views of fictional entities on the grounds that thoughts about these fictional objects "take place within the scope of a pretense." I do not consider this to be a strong argument against the realist view. Moreover, I am arguing here for a realist account that does not require pretence. Even if I acknowledge Everett's (2013) criticism, my proposed view can account for the main aspect of fictional creation: imagination depends on memory. My focus on the matter of fictional characters (i.e., the memories and mental images of previously perceived existing objects) avoids the pretence problem.

Vyshedskiy (2019) describes the neurological process of constructing a novel composite mental image, such as a centaur. The composition of this mythical entity must have been imagined by the creator(s) by first mentally synthesizing parts of a man and a horse and then producing the product of this mental creation through drawing (Vyshedskiy 2019). In other words, the creator must first see a man and a horse. The creator must then segment the remembered image of the man and the horse and synthesize the man's head and the horse's body into a new object called a centaur. In philosophical terms, the centaur denotes the real existing man's head and horse's body, which are viewed and then remembered, segmented, and synthesized. More complex fictional entities such as Emma Woodhouse exist because the name refers to the objects or individuals that Jane Austen perceived and then rearranged into this character. Specifically, Emma Woodhouse exists from a semantic viewpoint because her name denotes the real things or individuals with specific characteristics (e.g., a woman in her twenties, prettiness, intelligence, wealth, and comfort) perceived, imaged, segmented, and synthesized by Jane Austen.

Brock (2010) contends that the creationist perspective lacks explanatory power and generates more questions than answers. He highlights the ambiguity in the timing and method of a fictional object's creation and questions the reliance on authorial intentions. It is therefore important to understand how a brain state, identical to some reworked memory amalgam, can produce an abstract artifact.

The neurophysiological mechanisms underlying the ontological dependence of fictionalia on mental states (i.e., reworked memories) likely involve the core network of the brain. This network includes important regions such as the medial prefrontal cortex, frontopolar cortex, ventrolateral prefrontal cortex, medial and lateral parietal cortex, precuneus, retrosplenial cortex, posterior cingulate cortex, lateral temporal cortex, and the hippocampal formation (Addis et al. 2007, Hassabis and Maguire 2007, Schacter et al. 2007; Schacter et al. 2012, 2017, Benoit and Schacter 2015, Arzy and Dafni-Merom 2020).

Each core network component contributes to the interplay between memory and imagination. For instance, the medial prefrontal cortex mediates decision-making and relies on the hippocampus for rapid learning and memory consolidation (Euston et al. 2012). The frontopolar cortex supports the exploration of novel behavioural options and higher-order cognition (Boschin et al. 2015). The ventrolateral prefrontal cortex aids emotional regulation and response inhibition (Sturm et al. 2016). The medial parietal cortex is involved in memory recall and visual scene processing (Silson et al. 2019), while the lateral parietal cortex plays a critical role in episodic memory (Davidson et al. 2008). The precuneus supports the sense of self, autobiographical memory, and spatial navigation (Cavanna and Trimble 2006, Freton et al. 2014). The retrosplenial cortex is associated with spatial functions (Cona and Scarpazza 2019). The posterior cingulate cortex is active during internally directed cognition and autobiographical memory retrieval (Leech and Sharp 2014). The lateral temporal cortex supports conceptual representations of action (Kable et al. 2005), and hippocampal formation is essential for spatial cognition and episodic memory (Maurer and Nadel 2021).

Studies using functional neuroimaging techniques have shown how these brain regions work together to interrelate memory and imagination. For example, Addis et al. (2007) asked participants to remember an experience or imagine a future event while being scanned using functional magnetic resonance imaging. The tasks were divided into a "construction" phase, in which participants generated an event in response to a cue, and an "elaboration" phase, in which they described the event. The results indicated that brain activity was similar during remembering the past and imagining the future, particularly in the right frontopolar cortex, left ventrolateral prefrontal cortex, and right hippocampus (Addis et al. 2007).

In their constructive episodic simulation hypothesis, Schacter and Addis (2007) propose that episodic memory permits the simulation of future events by flexibly extracting and recombining features of past experiences. This constructive process is prone to errors and illusions, indicating that past and simulated future episodes use similar neural processes. This hypothesis suggests that imagination and memory are deeply intertwined, a connection that facilitates the creation of usable episodic information.

The core network's role in supporting the interrelation between memory and imagination suggests that fictional entities can arise from spontaneously reworking memories. This spontaneous creation is a neurophysiological process whereby mental images of the future, based on rearranged memories, arise unintentionally during routine activities (Kvavilashvili and Rummel 2020), as described above. This process mirrors the creation of fictional contexts and entities, which can emerge without deliberate intent.

Although the exact causal mechanism by which brain states produce abstract artifacts may remain elusive, I focus here on the ontological dependence of fictionalia on mental states, specifically reworked memories. Importantly, fictional entities ultimately refer to natural objects and experiences in the world. The memories reworked to create these entities are derived from real perceptual experiences. Therefore, even when a fictional context or character is created spontaneously, it is rooted in the individual's autobiographical memories and perceptual experiences of the real world. This grounding in reality provides a coherent and relatable basis for fictional entities, ensuring that they are not entirely detached from the real world but are a reimagined reflection of it.

Therefore, I argue that the creation of fictional entities is not solely dependent on authorial intentions but is deeply rooted in the neurophysiological processes of the brain's core network. This network's ability to rearrange and rework episodic memories into imaginative constructs supports the idea that fictionalia can emerge spontaneously.

Although more work is needed to identify the specific neural mechanisms underlying the creation of fictional entities, the use of neurophysiological insights to inform the study of fictional creation is a promising approach to understanding how the brain generates complex, imaginative scenarios. This approach bridges the gap between intentional fiction creation and spontaneous mental constructions, expanding our conception of creativity and the role of memory in shaping our imaginative landscapes.

Conclusion

Drawing on neuroscientific studies on the relationship between the memory of perceived objects and imagination, I have presented and defended a realist account of fictional entities. I consider these entities (i) to be hylomorphic artifacts; (ii) to have as their matter real perceived objects remembered through mental imagery; (iii) to be composed of the recollection of those objects; (iv) to be individuated by the creator's unique memories and the subjective images of their recollection; (v) to be created; (vi) to come into being when the creator synthesizes, disassembles, completes, and modifies the mental images resulting from memory retrieval into a new fictional entity; and (vii) to have the capacity to cease to exist.

My theory has two advantages. The first is that I have shown that rearrangement, in this case, the rearrangement of memories, results in the creation of new objects. In Material Beings, Peter van Inwagen (1990) asks the reader to imagine a foreign legion in a desert that manipulates the sand to create a structure that can be used as a fort. van Inwagen (1990) believes that these people have not brought an object, the fort, into existence. According to van Inwagen (1990, p. 124), "they have...rearranged the furniture of earth without adding to it." Thus, no form of arrangement or rearrangement can bring an artifact into existence (van Inwagen 1990).

I believe that although it may be correct to state that rearranging the "furniture of earth" does not necessarily bring artifacts into existence, such as when I rearrange the furniture in my house, rearrangements may create a new object. Just as a sculpture cannot be created without the rearrangement of clay, a centaur cannot be created without the rearrangement of the creator's memories of a man and a horse. Thus, although rearrangement of the furniture of the earth is not a necessary condition to create all new objects, it may be a necessary condition to create artifacts such as fictional entities. In other words, the idea that rearranging something does not necessarily result in the creation of something does not mean that objects that require rearrangement to come into being (like a clay sculpture) do not exist. Unsurprisingly, van Inwagen (1990) calls his stance "anticonventionalism" (p. 3).

The second advantage of my theory is that it provides a realistic account of fictional entities without relying on pretence theories (e.g., Everett 2013). As I have argued, the proven relationship between memory and imagination supports the idea that fiction should be primarily explained in terms of the creator's memory of existing objects and not exclusively in pretence or simulation terms. We use existing objects to generate simulations; simulations are not independent of existing objects. I suggest that we must accept the premise that imagination cannot take place without a process whereby we perceive objects, remember them through mental imagery, and finally rearrange those images. According to this understanding, Kroon and Voltolini's (2018) nonexistence datum no longer seems intuitive. Considering the relationships between perceived objects, memory, and imagination, I conclude that an existence datum is more intuitive.

Footnotes