In this paper I deal with the question of how it is that we have emotional responses to things that we do not believe in the reality of, specifically things like characters and events in literature and film (the paradox of fiction). The direction in which I wish to take this query is not the traditional philosophical approach to this question, however. There has been much written on this particular approach, and because of this I think that it is becoming, in philosophical effect, stuck. What I wish to do in what follows is to approach this question from a different set of assumptions and from a different paradigm, with the hope that I will be able to produce a more constructive and perhaps even more accurate resolution to the problem.
aesthetics, paradox of fiction, narrative understanding, moral imagination, discursive reasoning
The more isolated I become, the more I come to like stories.
-----Aristotle, Fragment 688
To say that all human thinking is essentially of two kinds-reasoning on the one hand, and narrative, descriptive, contemplative thinking on the other-is to say only what every reader's experience will corroborate.
Philosophers often concern themselves with questions that seem either overly convoluted or downright obvious to laypeople. The question that I have concerned myself with for the past several years, that seems all too obvious to those around me, is this: "How is it that we have emotional responses to things that we do not believe in the reality of, specifically things like characters and events in literature and film?" The direction in which I wish to take this query is not the traditional philosophical approach to this question. There has been much written on this particular approach, and because of this I think that it is becoming, in philosophical effect, stuck. What I wish to do in what follows is to approach this question from a different set of assumptions and from a different paradigm, with the hope that I will be able to produce a more constructive and perhaps even more accurate resolution to the problem. I also hope that by using a different framework that I put forward, the kinds of discussions philosophers have about this problem might change as well.
The roots of the issue I will here discuss have their basis in Plato and Aristotle's discussions of the kinds of emotions that representations (or falsities) produce in us and whether such emotions are healthy. One recent form of this debate appears in what I will call the classical construction of the paradox of fiction. This is a paradox constructed to question how it is that we can have seemingly real emotional responses to the situations of fictional characters, when, under normal circumstances, we have to believe in the reality of the states of affairs described by propositions that describe certain situations in order to have emotional responses to them. It is clear that we do not believe the propositional statements that describe fictional situations and so it seems odd, or at least philosophically inconsistent, that we should respond in the same way that we do when we do believe the propositional statements.
What I wish to do here is to call into question the construction of the paradox itself by changing the nature of the question that prompts the paradox. I will focus the discussion on how it is that we cognitively order, construct meaning in, and understand the world around us. I will borrow from literary theory, narratology and cognitive psychology for different kinds of explanations of how we order our experience of the world and how we understand fiction. In the end I will argue how these two ways of ordering experience (of experiences with real intentional objects and those with fictional intentional objects) are more in line with one another than the paradox of fiction assumes. For the purposes of this paper I will look exclusively at the narrative descriptions that produce the seemingly odd, or philosophically unjustified, emotional responses and ignore for the time being the visual arts altogether. What I hope to show in the end is that by appealing to a different model on which to base the way that we order and understand our experience of the world, our emotional responses to fiction will not seem as odd to philosophers, and might appear as natural (and even as justified) as they do to the laypeople with whom we interact.
1. The paradox of fiction and its assumptions
The classical construction of the paradox of fiction is as follows:
1. We have emotions concerning the situations of fictional characters.
2. To have an emotion concerning someone's situation we must believe the propositions that describe that situation.
3. We do not believe the propositions that describe the situations of fictional characters.
Attempted resolutions to the paradox usually attack one of the three propositions. Pretense theorists deny the first proposition, suggesting that the emotions that we have with fictional characters as their intentional objects are not genuine but "quasi"-emotional responses, "as-if" emotions  or some other alternate variation denying the same quality or kind of emotional responses that we have to real characters. Those who deny the second proposition, generally known as thought theorists, advocate different views about what is required in terms of cognitive content in order to have an emotional response. They argue that we can have justified emotions as a result of propositional beliefs merely entertained but not believed. Thought theorists argue that there is not a strict existential belief requirement necessary about propositions in order to have a justified emotive response. Those who argue against the third proposition of the paradox of fiction are known generally as illusion theorists and they argue for a stronger variation of the proposition, which defends the readers' cognitive states. That is, they argue that we do, in some sense, believe the propositions that describe the situations of the fictional characters.
I want to suggest that there are two implicit assumptions working within this paradox that will continually constrain the kinds of possible resolutions available and will never allow the paradox to be resolved. First, as the paradox is constructed, the belief component is assumed to be a necessary part of a genuine emotive response. A reader must either believe the propositions that describe that situation or she must disbelieve the propositions. Or, for the thought theorists, the reader must at least seriously entertain the propositional statements. If there is no belief in the reality of the situation, that anyone was really affected or that anyone really experienced what happened in the fiction, then there is no belief, existentially, in its reality. If I were told that the information about to be imparted to me is false, and I am then informed that I won the lottery, it would indeed be odd for me to become ecstatic, mildly excited or affected at all-and I certainly would not begin to behave as if I had won the lottery by running up my credit card bills. If I do not believe the propositions that describe a given situation, it would, indeed, seem odd that I would behave or respond emotionally as though I did believe them. This very fact, that I respond emotionally (cry, feel fear, clench my fists etc.) to things I do not believe to exist or to be really happening is at the heart of the phenomenon that generates the paradox of fiction in the first place.
What further prompts the primacy of the role of the belief requirement in a justified emotive response is that until recently the prevailing theory of emotions did, in fact, depend on beliefs to locate the cause of the emotive response. The cognitive theory of emotions is based on a general model that claims that emotions are cognitively constructed by the beliefs and judgments that might prompt them. Recently, however, philosophers like Jenefer Robinson and Paul Griffiths have suggested a different model of emotions that does not depend on the cognitive prompts that other judgmentalist theories have previously held to be necessary. Robinson, for example, argues that the startle response is evidence that emotions (including primitive ones like the startle response itself) are at least partly physiologically, and not cognitively, based. Robinson argues that the evidence from neurophysiology shows that emotions do not depend upon cognitive judgments. She explains, "the formulations of beliefs, judgments, and conceptions about a situation are not necessary to an emotional response." Further, the cognitive belief components are neither necessary nor sufficient for an emotive response. Given the predominance of the judgmentalist or cognitive view of emotions, an emphasis on a belief seems warranted. Considering, however, that there are other viable theories that do not require the cognitive components once thought necessary, it seems only fair to call this into question.
The second assumption I wish to call into question is that we can clearly differentiate between what is fictional and what is real and that we always have knowledge of the propositions that describe a set of circumstances as being one or the other. The paradox assumes not only that we can distinguish between the fictional and the real, but it implies that they are indeed two completely different things altogether, both ontologically and epistemically. I will not argue the ontological distinction here, but I do want to call into question the notion that epistemically there is a clear distinction between the real and the fictional and that it is possible or even likely that we can identify that distinction. Although there may be two completely differentiable realms of fiction and reality, practically, we do not always have full information on which to base a distinction. Further, epistemically, having full information may not be what is really relevant in terms of understanding or differentiating between the real and the fictional. I will address this concern more fully later.
For the purposes of my argument, I wish to bracket these two assumptions: that a cognitive belief requirement is necessary for a justified emotive response and that fiction and reality are always distinguishable. I will deal with both of these later in the paper, but for now, I want to change the direction of my query. Rather than explaining why the paradox is unresolvable or why each of the different kinds of responses fails to resolve it, what I propose to do in what follows is to change the question from, "How is it that we can have emotional responses to fictional intentional objects?" to "Is there something about narrative that is different from other linguistic constructions and that makes us tend to respond emotionally?" With a different construction of the problem, I will more easily be able to formulate a different resolution that better reflects the way that we experience the world and is more reflective of how we order and understand the world, and makes our emotive responses to fiction seem reasonable.
2. A new paradigm: the narrative and the discursive
To begin to construct a new paradigm from which to approach the question "Is there something about narrative in particular, different from other linguistic constructions, that makes us tend to respond emotionally?" I first want to turn the discussion to storytelling. The way we tell stories, how we understand the stories that others tell us, the way we use stories to recall memory, and the role that storytelling and story-creation play in developing a theory of personal identity are all integral to the way we order and make sense of our experience in the world. The kinds of stories I am referring to are the most basic-the stories we relate to other people in our lives describing our experience of the world. We create stories out of the most mundane information to the most exciting, inspirational things that we experience. I will, from now on, refer to these kinds of everyday general retellings as stories or as being in story-form.
I wish to begin to build a new framework within which I can address the question of why we respond to fictions by making an argument that there are (at least) two fundamentally different ways that we order our experience of the world-the discursive and the narrative. Most generally, the discursive mode relies on logical, linear reasoning while the narrative depends on the narrative aspects of a situation to show how one state of affairs is motivated by a previous one (although not necessarily caused by it). Each of these two ways of knowing and ordering the world, the narrative and the discursive, has different operating principles and each has its own criteria for success. They differ fundamentally in the ways they go about verification. Narrative psychologist Jerome Bruner goes so far as to assert that "a good story and a well-formed argument are different natural kinds" even though both are used as ways of communicating information and of convincing others. He explains that the discursive "verifies by eventual appeal to procedures for establishing formal and empirical truth" while the narrative "establishes not truth but verisimilitude" and lifelikeness. We utilize both the narrative and the discursive to order our experience. Both are legitimate ways of understanding our experience of the world, but both are also produced cognitively by different means.
Although these two modes of understanding are complementary, it is important to note that they are neither reducible nor translatable to one another. It seems unlikely, and worse, futile, to attempt to either reduce or translate from one to the other since they are, at base, fundamentally different kinds of reasoning and ordering. This is not to say, moreover, that stories cannot exemplify both the narrative and the discursive. Many Sherlock Holmes mysteries, for example, are presented in the form of a narrative, but what motivates the reader in this case is not the narrative aspects of the novel as much as the discursive clues that one has to follow to figure out the mystery. Edgar Allan Poe demonstrated this also in the introduction to "The Murders in the Rue Morgue." Poe made clear early in the story that straightforward logical reasoning will not always be the best strategy for figuring out a mystery (in this case). Poe noted that "the analytical power should not be confounded with simple ingenuity; for while the analyst is necessarily ingenious, the ingenious man is often remarkably incapable of analysis." The story that follows shows how listening for the narrative clues will bring one to a fuller understanding of the story than strict logical reasoning. He says that "the difference in the extent of the information obtained, lies not so much in the validity of the inference as in the quality of the observation." Further, even though both the narrative and the discursive can exist in a single text does not mean they cannot be differentiated.
The logical proof, "Socrates is a man, all men are mortal, therefore Socrates is mortal," is an example of discursive reasoning. The logical and causal steps along the way are apparent, linear and straightforward. No one who understands even the folkways of logic could deny that the conclusion follows the premises, categorically. This is a completely different kind of reasoning than is expressed in the kind of motivation which prompts Creon to make a decision about whether or not he should let Antigone give her brother a proper burial. Although there are identifiable discursive components to Creon's reasoning, these are not what motivate the story, nor are they the elements that help readers to empathize with Antigone.
The kinds of causation used in the two different cases, of Socrates being proved mortal and Creon being forced to make difficult kingly decisions, are patently different. Further, the linking term then has two different functions in the two cases. For example, in the logical proposition "If x, then y" and in the narrative recit "The queen died and then the king died, of grief" then functions as a logical consequence in the former and as an explanation for mortal grief in the latter. The former looks for universal truth conditions and the latter looks for human connections and general explanations between events. The goal of causation in the two cases is construed differently, even though it exists in both ways of knowing; the goals are different for the two ways of knowing, and the cognitive processes that aim at these goals are fundamentally different as well.
The discursive mode uses general causes and employs them in order to be able to test empirical truth and to find universal truth conditions. Bruner explains that the discursive mode "leads to good theory, tight analysis, logical proof, sound argument, and empirical discovery guided by reasoned hypothesis." Further, he explains that the discursive mode "attempts to fulfill the ideal of a formal, mathematical system of description and explanation" and in order to do that, "it employs categorization or conceptualization and the operations by which categories are established, instantiated, idealized, and related to one another to form a system." The discursive mode allows us to see logical connections between states of affairs prior to being able to prove them empirically. It further allows us the faculty of abstract reasoning, which can go above and beyond the particular example of any given situation. This ability to abstract is essential to our functioning as humans but it is not what makes it reasonable to respond emotionally to particular situations we believe to be fictional. For that, we must rely on the narrative mode, and our interest in relating to others through stories in general. The narrative mode, in opposition to the discursive, will lead to good stories, and believable dramatic and historical accounts. The narrative mode looks primarily for human (and not necessarily or exclusively logical) connections between characters and events.
The discursive mode of reasoning, by and large, underlies philosophers' attempts to resolve the paradox of fiction. The logical construction of the paradox assumes that the discursive mode of thinking is the only mode available for use. The way that the three premises of the paradox add up entails the logical nature of the causal and linear way of thinking that is epitomized by the discursive mode. I want to argue that, seen in light of the narrative mode of understanding and reasoning, the question of why we have emotional responses to fictional intentional objects (assuming that we do not respond to things we do not believe in the reality of), the classical construction of the paradox of fiction is misguided in and of itself, philosophically misleading, and most importantly, inherently unresolvable. If it really is a goal of philosophers to come up with a workable and reasonable resolution to the seemingly odd emotional responses we have to fiction, we have to change the nature of the question. I wish to rely on the model of the narrative way of ordering experience in order to show how this can be used as a different context from which to look at the original question that prompts the paradox of fiction.
It is clear that if a reader were asked while reading a novel whether or not she had an existential belief about the characters in the novel, the answer would likely be "No" (she does not believe the characters really exist). This is not, however, what is at the core of the issue. What is at issue is the fact that, although we know the existential, propositional statements are false (readers know that the proposition "Cinderella exists" is false), we also have knowledge of events within the fiction that does not correspond to the same framework of discursive propositional statements. "Cinderella went to the ball" is an example of a proposition that we believe to be true in the fiction. Rather than separating what is true in the fiction from what is true in reality or outside the fiction I suggest that we continue to bracket that dividing line between fiction and reality and distinguish these by the discursive and the narrative forms of reasoning. That is, there are existential, logical, questions we have about Cinderella, but also narrative questions that can be understood only within the narrative mode. If the existential disbelief of the propositional statements were all that is at stake and the narrative really were reducible to the discursive, then the classical construction of the paradox really would be the definitive end of the debate. I want to argue, however, that the narrative really is a fundamentally different kind of meaning construction and entails a different kind of understanding and processing of information. The narrative mode of understanding is what remains constant between the stories we hear and tell about "real" events and the stories we know to be fictional.
Although it is not necessary for me to defend a particular theory of narrative, it will be useful to outline some of the features that are most commonly included in theories of narrative. Generally, narratives involve at least two events or states of affairs, which are in some way ostensively connected. The events usually concern the interests of at least one unified subject and there must be some sort of narrator. The temporal relations between the events are perspicuously ordered. Most times the events are sequenced in such a way that causal connections can be delineated. That is, earlier events in the narrative can be shown to be causally relevant to the effect. There is some implied causal link between events, which shows their connection beyond an episodic or chronicled catalogue of happenings. A narrative explanation of a series of events is focused on how change from beginning to end is produced and not just on how one event is predicted or deduced from another.
An annal represents historical reality, listed chronologically, as if the real events did not take the form of a story. A chronicle, on the other hand, represents in such a way that lists of events appear in the form of unfinished stories. "The queen died and then the king died" is a chronicle, but not a narrative. "The queen died and then the king died of grief" is a narrative, partly because it utilizes causation or consequence, but mostly because it shows a relationship in the story, which goes beyond a timeline and into the relationship between the characters. Narratives also require more of the listener if she is to understand how all of the parts come together to form a coherent whole.
For my purposes what qualifies as a narrative is less important than the ways in which and the frequency of which we process what comes to us in story-form. Roland Barthes explains the pervasiveness of narratives in our lives as follows:
"The narrative may incorporate articulate language, spoken or written; pictures, still or moving; gestures and the ordered arrangement of all the ingredients: it is present in myth, legend, fable, short story, epic, history, tragedy, comedy, pantomime, painting, stained glass windows, cinema, comic strips, journalism, conversation. In addition, under this almost infinite number of forms, the narrative is present at all times, in all places, in all societies; the history of narrative begins with the history of mankind; there does not exist, and never has existed, a people without narratives."
Narrative is also a concept that is used in a variety of ways. Narrative can be used to describe a kind or quality of a text (with text defined loosely as well). It can also be used to describe how we interpret a particular text. Further, what I call narrative understanding is the mental mechanism with which we understand previously constructed narratives, how we order and make meaning out of unordered information we take in from the world, and how we employ narration to express our views and experiences to others through stories.
4. Understanding stories
I want to argue that we make sense of our experiences of the world primarily in terms of stories. This is not to say that we do not remember, order, or make sense of outside data in any other ways. This is also not to say that external information comes to us in the form of stories; it is clear that whatever information that might be obtained by any given sequence of events arrives and is available for the makings of a story still needs to be crafted into a story. What I do want to argue is that the primary way that humans make sense of, and order their experiences in the world, is through the act of creating, telling and understanding stories. Barbara Hardy argues that storytelling plays a major role in both our sleeping and waking lives. She says, "we dream in narrative, daydream in narrative, remember, anticipate, hope, despair, believe, doubt, plan, revise, criticize, construct, gossip, learn, hate and love by narrative." Self-expression is done primarily through narrative construction and communication. Understanding of a significant part of the world is also made comprehensible through the stories we tell to others (and the ones we tell ourselves) and the stories others tell us.
In his work on artificial intelligence, Roger Schank goes so far as to assert that "storytelling and understanding are functionally the same thing" and that "intelligence is bound up with our ability to tell the right story at the right time." The ability to remember the gist of a story, to index and reference a multitude of stories, and to be able to express them well is an extremely important and useful aspect of intelligence.
We experience the world and craft stories based on that experience in order to make sense of it. Fictional narratives (good ones, or well-told ones), then, take only what is significant in an event and present that, without the unstructured events that we experience in our own lives. Understanding fictional narratives and understanding narratives about "real" events thus employ the same cognitive process. What is significant here is that we understand narratives in a consistent way and not that we understand or process narrative accounts of real and fictional stories primarily in terms of whether or not we believe the propositional statements that are descriptive of the narrative. Belief is secondary, at best, to the unified cognitive processes of what it takes to understand a narrative. The way we experience fiction, then, is based in the same cognitive processes as the way we respond to the stories we hear from others. The kind and degree of emotional response we have to fictional narratives and the narratives other people tell us will have to do more with the subject matter and the story construction than it has to do with whether we believe or disbelieve the propositions that constitute the story.
Schank argues that one of the primary ways that we make sense of our world is through storytelling. He suggests that when someone experiences something in the world, she has a set of scripts that she uses as general reference points. For example, I have a restaurant script that allows me to go to a restaurant and have a certain, reasonable set of expectations of how this event is going to go, and what the expectations are of me if I want to be served food. We have these general scripts for all kinds of activities in our lives, everything from brushing our teeth to going to the grocery store to engaging in romantic relationships. If we did not have these scripts, then every situation would be new every time, and we would constantly be bombarded with new information that is unstructured by habit, social mores, and practical knowledge. Scripts are used like a set of expectations about how certain familiar situations will play out. Schank explains "in a sense, many situations in life have the people who participate in them seemingly reading their roles in a kind of play." We have life scripts that we play out, but we also have scripts for almost everything else we do while interacting with others as well as while we function alone. The scripts through which we experience the world are relatively easily adaptable to certain new information and new variables that might be introduced. Life can be somewhat frustrating for those who cannot or choose not to be flexible with these scripts.
When we create a catalogue of stories, we also construct an indexing and referencing system. When I experience an event in the world, the data that I can take in about that event is not organized in any meaningful way. When I relate to someone else my experience of the event, I use the aspects that I consider to be relevant to my expressing the gist of what happens. I can choose to tell about my experience in a number of ways using a number of different aspects of what happened, and the aspects I choose to create a story out of might be very different from the ones someone else might choose. But a story is a much better way of communicating the gist of an event because stories illustrate events in a more effective and efficient way than just simply relating a chronology of things that happened or listing a set of beliefs or attitudes. If my story is good enough, I will not have to state the point of the story at all, and my beliefs will be made apparent along the way as well.
Not only are stories the most effective means of communicating with others, but stories also help us to form memories of particular events in our lives. One way that we organize all of the incoming information is into what Schank calls Memory Organization Packages (MOPs). He explains that a MOP covers a context dependent aspect of memory. Memories are partially based on scripts, but as we recall events later, we remember how the event perhaps varied from a script, or we remember how a particular event happened within a particular context. He suggests examples of MOPs as things like going on a family vacation or on a first date. Any MOP is made up of a set of scenes that are imaginable and easily recalled within the framework of the MOP. Schank suggests that "in this view of memory, episodes that occur get stored in terms of the scene in which they took place and can be connected to the larger context by reconstruction." In this way, then, we take in unorganized information and, through the process of creating stories, we both communicate to others as well as help to solidify the experience for ourselves. Much of our knowledge is then comprised of the set of experiences we have had in the world, and it becomes established through the stories we tell of our experiences. We also remember our experiences through stories, as we often find that one person's story will help us to recall an aspect of one of our own stories. Most people index and recall through stories rather than propositional beliefs or attitudes because they are easier to remember and to index. Our communication with others, however, is limited to the stories that we have to tell. On our own we have a storehouse of memories and experiences, but what we communicate to others are often stories. Thus, knowledge is experiences and stories; intelligence is the ability to tell an appropriate story containing the appropriate information and emotional pull, at an appropriate time, crafting just the right story from a storehouse of options, for the right occasion.
One of the ways that we create particular memories for ourselves, and how we create our own particular sets of experiences that make up our lives, is through the process of story creation and storytelling. However, given the fact that the story that I tell about an event could be vastly different from the story that someone else tells about the same event, the reality of the situation is that we often choose to tell what we want to tell and thereby manipulate our own experience into the way we want to experience it. Often, we tell stories to certain people, in certain ways, in order to give them a certain sense of the way we wish them to perceive us. What we keep in memory is the gist or the skeleton of a story and then we can tell it in different ways depending on what we are hoping the effect of the story to be on the hearer of the story.
What may change the stories and the memories even more than the different ways we experience situations is the post-story creation. For example, if I feel like I have had a particular kind of injustice done to me recently, I begin to see these kinds of injustices done to others more often than I have seen them before, and I then, in turn, talk about the prevalence of these injustices around me as well. Thus I am aware of it, so I tend to see it more, so I talk about it more, so it becomes more ingrained in my memory. Years later I can recall the time when I was so focused on that injustice. Someone else may not be looking for my injustice and so will not be telling stories about it or creating memories around it.
The kinds of stories we tell are also connected to the way we choose to see the world. Thus in much therapy, especially psychoanalytic therapy, the goal is not just storytelling, but storytelling that can be modified into healthier or more positive or productive life stories. Therapy that just allows one to rehearse negative stories would likely do more harm than good. When we tell people stories about our experiences, in the process of creating and telling the story the memory structure will be created at the same time. Story creation is a memory creating experience in itself. This memory structure produced the first time the story is told will contain the gist of the story that we remember long-term. Talking is memory creation and if one has an experience that she never once talks about, it is likely that she will not create any kind of long lasting memory of the experience. Rehearsal also helps memory; the more often a story is told the more prevalent it will be in memory.
The fact of the matter is that story invention is a process of massaging reality. All we have are the perspectives and stories that we tell, constructed through the lenses of the other experiences we have had and the beliefs and attitudes that we have about the world. The truth conditions that apply to propositional statements are not helpful in describing human experience in this way. Unfortunately (for some), Schank explains, "our stories, because they are shaped by memory processes that do not always have their basis in hard fact, are all fictions." This is not to say that these fictions did not happen or are not based on anything at all, but as far as they are constructed through the medium of narrative they are necessarily perspectival. It is this shared aspect of narration that literary fiction shares with the fictions or stories of our everyday lives. Even though the fictions or stories we create are based on lived experience and sometimes even on concrete events, the stories are still created through perspectives that do not allow "true reality," or the "truth" of what really happened, to come out. What really happened, aside from the way those who were there experienced it, does not meaningfully exist for us outside of our own standpoint. Epistemologically, we have no unbiased access to the events of our lives.
5. Narrative selves
A narrative based theory of personal identity helps to lay the groundwork for this more integrated notion of understanding narrative. Psychologists have given us the framework in which to tell stories in order to recreate and reinvent ourselves through talk therapy. Freud, moreover, suggested that we are like a host of characters in a play in and of ourselves, which we produce ourselves, internally. These aspects of our personalities should not, ideally, take over one another or violently wage war with one another, as happens when mental instability and illness arrive. The goal of psychoanalytic therapy is to help the analysand to recall and reinvent the stories that make up the self, which is a malleable, story-based, personal identity. Stories are the connectors in memory and are perhaps the linchpin to a coherent narrative based theory of personal identity.
In his ethical treatise, After Virtue, Alasdair MacIntyre develops a narrative concept of selfhood. He notes that experience does not come to us in the form of narrative, but rather we construct narrative in order to make sense of experience. The way we experience the world is easily understood through the construction of narrative; it is where we feel most at home, living our lives, telling our narratives to others and importantly, understanding others' actions in terms of narrative. MacIntyre explains "stories are lived before they are told - except in the case of [literary] fiction." Only in literary fictions do we see the narrative framework of an event already pre-packaged. Because our own senses of selves are located in these narrative constructions we relate to them very naturally.
The events that comprise our personal histories are linked together by narrative and our ability to share these histories with others is expressed through narrative as well. These histories linked together eventually become part of our own sense of self. We are the stories we tell. There is no such thing as a static, intuitively obvious, or essential singular self to be identified or pointed to. Humans are not, by nature, such fixed entities. I will never find my real self, if I strip away the roles that I play in my life or produce a list of beliefs that I hold. I am these roles and these beliefs and these come out in my interactions with others and through the stories that I tell. The sense of self is that which gets talked about and reflected through the stories we tell. We invent and reinvent our selves in order to meet the needs of every situation we encounter. Part of what keeps the identity stable at all are the beliefs and attitudes that stay with us at the core, and part of what keeps the identity stable is the storehouse of memories and stories we choose to tell ourselves and others. To tell about oneself is like making up a story about where one has been, what one has learned and why one is doing what one does.
In his work on narrative selves, Owen Flanagan argues for what he calls the self model, in which the self is "both a real entity and an abstract idea intended to pick out the most important features of an individual's life." He argues that there is an essential self to which we have no reliable access. He compares the core of the self to an event in history that did, indeed, happen and should not be construed as a fiction. Historians pillage through the collected versions of the stories that recount an event in hopes of finding the one that approximates the real event most accurately. Flanagan explains that an "omniscient historian" might be able to represent an event as it really happened, but in reality there are no omniscient historians, as there are no omniscient narrators of our own selves. Flanagan seems to confuse the ontological self and the epistemological self, however, not taking into consideration that we can only have knowledge of our selves epistemically and not ontologically-thus even an omniscient narrator would only be giving us epistemological insight. Whether or not there is a core ontological self is irrelevant to the discussion of the suggestion that our selves are created through narrative.
Jerome Bruner argues that without the capacity to make and tell stories about ourselves, there would be no sense of personal identity at all. He evidences a neurological disorder called 'dysnarrativia,' which is "a severe impairment in the ability to tell or understand stories" that is related to the other neurological disorders such as Korsakov's syndrome and Alzheimer's disease. One of the characteristic symptoms of dysnarrativia is an almost total loss of the ability to understand others' behavior and how others might experience various events emotionally. Sufferers of dysnarrativia have an imagination that is compromised by their inability to tell stories. No only do sufferers of dysnarrativia experience a loss of self, but also they lose their sense of empathy and all of their sense of the human experience of others. A medical study by Kay Young and Jeffrey Saver on various forms of dysnarrativia claims that selfhood is directly related to the ability to construct narratives. The construction of self is inherently incomplete without a capacity to narrate. The ability to tell stories is not just about communication with others; it is about creating a coherent script, life story, or integrated matrix within which to make sense of one's life. The ability to narrate seems to be then a necessary condition for personal identity, although it is not sufficient.
Richard Wollheim suggests another way of characterizing the mental states that help to describe imagining generally (and in our narrative imaginings) that could help to differentiate two different kinds of imaginings that underlie our ability to understand narrative. He suggests that we have iconic mental states [IMS], which are descriptive of a particular kind of imagining, specifically a visual imagining. IMS may be about individual people, but are more often imaginings of events that have their own characters. IMS regularly occur in sequences in themselves, the mental states follow one another sequentially, and the characters who act in the events are what Wollheim refers to as the dramatis personae. IMS have a central imaginer for whom the events of a story unfold into a narrative. The second feature of IMS Wollheim points out is that the individuals, events, sequences of events, and stories that iconic mental states are of, represent, iconically, the individuals and events themselves, whether real or imagined. Wollheim recognizes the difficulty in characterizing how exactly this kind of representation works, but suggests that the kind of representational relationship he is referring to is the relationship between the intentionality and the subjectivity of the IMS. He argues that we do not necessarily have to "observe our images before we can know what they are, any more than we have to listen to the words we think in before we can know what they are." This is a large part of how iconic imagining is of a different kind than non-iconic or propositional imagining. Perhaps these iconic imaginings are more reflective of, or iconographic, rather than strictly representational of, a one to one relationship: the iconic imaginings are intentionally related either way. Thirdly, and perhaps most important, Wollheim argues that "Iconic mental states have causal efficacy over the behavior and the mental dispositions of the person who has them." IMS have psychic force. Iconic imaginings are different from lightly entertained or even deeply held propositional beliefs because they are not propositional at all.
Wollheim gives an example of an iconic report as follows: "I imagined the horse's falling down in the street." Conversely, a non-iconic report would be: "I imagined that the horse fell down in the street." While the first (iconic) report is something that can be understood only in terms of a visual imagining, the second report is a propositional statement that can be understood discursively, and in terms of belief requirements. It is clear that these are two fundamentally different kinds of reports and that one is not reducible to the other, nor would one be an effective means of understanding or utilizing the other. Recognizing that there are two ways of representing the report leads to an importantly different way of characterizing our reports about what we believe or disbelieve, imagine, or report about our experiences in the "real" world as well as with our encounters with fiction. Beliefs are not necessary for emotional responses, but IMS are.
Iconic imagining might help to explain the significant drop in beach attendance the summer after the film Jaws was released. Entertaining or believing the proposition (non-iconic) "I imagine that there are sharks in the ocean" is very different from (iconic) "I imagine there being sharks in the ocean." The latter might command psychic force in a way that the former would not, still not requiring even an entertained belief. I borrow the Jaws example from Kendall Walton but with a different intended purpose. Walton explains that "Jaws caused a lot of people to fear sharks, ones they thought might really exist, and to avoid swimming in the ocean. But this does not mean that they were afraid of the fictional sharks in the movie." He goes on to argue that the fear experienced by those watching the sharks generates a fear of the depiction of the sharks and not of the sharks depicted. What I want to argue is not that the fear is caused solely by the depiction of the sharks or by the mere thought of the sharks. Rather, the depiction prompts the IMS, which accounts for a kind of imagining that elicits an emotion that commands psychic force. Iconic imagining by itself cannot account for resulting emotions but IMS can be a part of what accounts for the motivation that leads one to form real beliefs and desires that become part of the cognitive makeup of the emotive response. Iconic imagining can thus account for a number of different kinds of emotional responses to fictional intentional objects that continue to be problematic for other kinds of theories. Narrative descriptions that prompt iconic imaginings can produce affective, emotional responses that do not seem to be so philosophically inconsistent with apparent belief sets.
IMS help us to recognize, along with the narrative mode of understanding, that the discursive mode is not the only way that we can imagine, remember or entertain thoughts. Given the IMS option, it might be possible to begin to leave behind the belief requirement once thought necessary for a justified emotive response to anything, real or fictional. It is not even that we cannot have justified emotive responses without the belief in their reality, but rather it now appears to be the case that it is a category mistake to discuss a belief requirement at all. Belief and disbelief, in conjunction with the necessity of truth and falsity of propositional statements, can only work in a non-iconic, discursive framework. If narrative understanding is taken to be a viable option to the discursive, the narrative ordering of experience, along with IMS, can take the place of the once firmly held need for belief in the propositions that describe a fiction that were once thought necessary for justified emotive responses.
I want to return to the suggestion that I made earlier that fiction and reality are not as distinguishable as we might like to think. This is not to say that I cannot distinguish, or that I even have trouble distinguishing, between my life and Anna Karenina's life or my friend's life and Anna's life. What I am suggesting is that the way that my life, and personal identity broadly construed, is constructed is through narrative understanding, and not through discursive understanding. Fiction and reality are distinct ontologically and epistemically but they are both understood by means of the same cognitive process. My personal identity or my sense of self is created and understood in the same way that I order, understand, and make sense of Anna Karenina's life story. Conveniently, hers is already carefully crafted and pre-packaged for me so that when I read about her predicament I am provided with all of the relevant information that I need for narrative understanding. Thus the huge advantage of reading literature is that much of the filtering of unnecessary detail is left out and what is left is the narrative description that will (ideally) lend itself well to the narrative understanding. If only our lives and the lives of our friends arrived so concisely and orderly!
In conclusion, it should now be clear how the narrative mode can enhance the explanation of the way we order our experience of the world. I further think that on my view, utilizing the narrative mode of understanding in addition to the recognition of the way we utilize IMS, we can move beyond the paradox of fiction and begin to explain our emotional responses to fiction in these terms. Without the constraints of the paradox itself, the centrality of the belief component within the paradox, and the recognition that modes of understanding are more important than the acknowledgement of the fictionality or non-fictionality of a given story, our emotional responses to fiction seem less philosophically inconsistent. Introducing the notion of narrative understanding, the pitfalls of the paradox of fiction can be avoided from the start. We do not need to assume the truth of all three propositions because we do not need to begin exclusively from the assumption that the cognitive theory of emotions is the only way to produce genuine emotive responses. With narrative understanding provided as an additional way to account for narrative experience and narrative meaning construction, there comes an alternative with which to deal with the paradox differently. Narrative understanding will also enable the philosophical discussion concerning our emotive responses to fiction to resume in a way that is more consistent with our actual experience.
 See Kendall Walton, Mimesis as Make-Believe: on the Foundations of the Representational Arts (Cambridge: Harvard University Press, 1990).
 See Robert Gordon, The Structure of Emotions: Investigations in Cognitive Philosophy (Cambridge: Cambridge University Press, 1987).
 See also Gregory Curie, The Nature of Fiction (Cambridge: Cambridge University Press, 1990); and "The Paradox of Caring: Fiction and the Philosophy of Mind," in Mette Hjort and Sue Laver eds., Emotion and the Arts (Oxford: Oxford University Press, 1997).
 See Peter Lamarque, Fictional Points of View (Ithaca: Cornell University Press, 1996); and "How Can We Pity and Fear Fictions?" British Journal of Aesthetics, 21(1981) 291-331. See also Noël Carroll, The Philosophy of Horror or Paradoxes of the Heart (New York: Routledge, 1990).
 Derek Matravers, "The Paradox of Fiction: The Report Model versus the Perceptual Model" in Mette Hjort and Sue Laver, eds., Emotion and the Arts (Oxford: Oxford University Press, 1997).
 I will use the adjective genuine in reference to these emotions, although I know that the term is fraught with controversy. Genuine emotions, in the way that I am referencing them here, are emotions that result from events that we believe to be true. Other variations of these emotions are known as justified, legitimate, reasonable, or rational emotions. Part of the problem is that someone like Walton, for example, would argue that "quasi" emotions are genuine as well, but not the same as emotions prompted by real events.
 Jenefer Robinson, "Startle," The Journal of Philosophy, Vol. XCII, no. 2 (1995), 53-74.
 Paul Griffiths, What Emotions Really Are: The Problem of Psychological Categories (Chicago: The University of Chicago Press, 1997).
 See Ronald deSousa, The Rationality of Emotion (Cambridge: The MIT Press, 1987) and Patricia Greenspan, Emotions and Reasons: An Inquiry into Emotional Justification (New York: Routledge, 1988) for more on the cognitive theory of emotions.
 Robinson, p. 64.
 Psychologists Donald Polkinghorne (Narrative Knowing and the Human Sciences (Albany: SUNY Press, 1988)) and Jerome Bruner (Actual Minds, Possible Worlds (Cambridge, MA: Harvard University Press, 1986)) both use the term "paradigmatic" (and Polkinghorne uses "logico-scientific" in addition) for what I will prefer to call discursive reasoning. Discursive captures the logical, linear aspect of reasoning more accurately, I think, than paradigmatic or logico-scientific, both of which are too broad for philosophical purposes.
 Bruner, Jerome, Actual Minds, Possible Worlds (Cambridge, MA: Harvard University Press, 1986), p.11.
 Ibid., p.11.
 Edgar Allan Poe, "Murders in the Rue Morgue" in The Raven, The Gold Bug and The Murders in the Rue Morgue (Pittsfield, MA: The Caxton Society, 1909), p. 5.
 Ibid., p. 4.
 Ibid., p. 11.
 Ibid., p. 13.
 Ibid., p. 12.
 Kendall Walton would even go so far as to say that the sentence "Cinderella exists" isn't a proposition at all since it contains a name that does not refer to anything.
 Having a sentence be "true in the fiction" is reminiscent of Walton's pretense theory where there are statements like "Cinderella went to the ball" which are true in the fiction but not outside of the fiction.
 Noël Carroll, Beyond Aesthetics: Philosophical Essays (Cambridge: Cambridge University Press, 2001) p. 126. See also Hayden White, The Content of the Form: Narrative Discourse and Historical Representation (Baltimore: Johns Hopkins University Press, 1987).
 Roland Barthes, "Introduction to the Structural Analysis of the Narrative," Occasional Paper, Centre for Contemporary Cultural Studies, University of Birmingham, 1966.
 Barbara Hardy, "An Approach through Narrative," in Towards a Poetics of Fiction: Essays from Novel, a Forum on Fiction, 1967-1976 (Bloomington: Indiana University Press, 1977), p.31; Roger Schank, Tell Me A Story: Narrative and Intelligence (Evanston, IL: Northwestern University Press, 1990), p. 24.
 Loc. cit.
 Ibid., p. 21.
 This is parallel to the argument Plato makes in the Theatetus about knowledge being like birds in an aviary. It is one thing to "have" the birds, but another thing to be able to call one to come at the right time. Being able to access knowledge when it is needed is at least as important (if not more so) than having it in the first place (Theatetus 197c-d).
 Schank, p. 7.
 Autistic individuals have a particular difficulty with adapting these scripts. They have a tendency to engage in rigid routines, are interested in fact-based information to the essential exclusion of an interest in fictional narratives and have a lack of interest in pretend play. This lack of interest is generally thought to be due to a lack of imagination generally, but manifests itself in an inability to sympathize with others or to have very little ability to understand the mental and emotional states of those with whom they interact. See Steven Stich and Shaun Nichols, "Second Thoughts on Simulation," and Alvin Goldman, "Empathy, Mind, and Morals" both in Mental Simulation, ed. Martin Davies and Tony Stone (Oxford: Blackwell Publishers Ltd., 1995). Both papers address the question of empathy dealing in particular with the challenges autistic individuals pose to the cognitive paradigm, which explains empathetic responses generally.
 Ibid., p. 127.
 Schank discusses this in some detail in his chapter on "Story Skeletons."
 There will be some exceptions to this with cases like repressed memory syndrome and other traumatic experiences. My point is only that general experiences like brushing one's teeth will not be differentiated unless a story is told about a particular instance.
 The notion that there is not one ideal, rational perspective has been developed theoretically, quite extensively in aesthetic theory and elsewhere. All aesthetic theory that argues against the British Taste Theorists' ideal of a disinterested viewer of art will apply here as well. Feminist theorists and standpoint theorists would be especially prone to argue against one real, knowable reality, as well as against one real, knowable self.
 Ibid., p. 44, my italics. The suggestion that all stories are fictions includes the implicit assumption that all fictions are narratives and perhaps vice versa. I do not endorse this suggestion, but I leave it as it is since it conveys the gist that the stories we tell are only marginally based in fact.
 The following all contribute to the argument for a narrative-based theory of personal identity. Alasdair MacIntyre, After Virtue, 2nd ed. (Notre Dame, IN: University of Notre Dame Press, 1984); Charles Taylor, Sources of the Self: The Making of the Modern Identity (Cambridge MA: Harvard University Press, 1989); Richard Wollheim, The Thread of Life (New Haven and London: Yale University Press, 1984).
 Sigmund Freud, Delusion and Dream: An Interpretation in the Light of Psychoanalysis of "Gradiva," a Novel by Wilhelm Jensen, ed. P. Rieff (Boston: Beacon Press, 1956).
 Alasdair MacIntyre, After Virtue, 2nd ed. (Notre Dame, IN: University of Notre Dame Press, 1984), p. 212.
 Owen Flanagan, The Problem of the Soul: Two Visions of Mind and How to Reconcile Them (New York: Basic Books, 2002), p. 244.
 Ibid., 244.
 I will use the term "personal identity" here loosely, not wanting to neglect or be blatantly unaware of the rigorous history that philosophy provides for that concept. What I am proposing here is a sense of personal identity in the philosophical and in the psychological sense, but what I am suggesting is that narrative is a necessary component of personal identity but is not sufficient. I will leave the strict definitions of this to others.
 Amnesic Syndrome is also known as Korsakov's Syndrome. The symptoms include the impairment of recent and remote memory, preservation of immediate recall, absence of generalized cognitive impairment, and amnesia.
 Kay Young and Jeffrey Saver, "The Neurology of Narrative," SubStance, #95/95, 2001.
 Centrally imagining is imagining from the single viewpoint of one person. The person whom you centrally imagine (in most circumstances this is yourself) you imagine from the inside and you view all the other characters peripherally. Imagining acentrally is to imagine from the perspective of an outsider-as the figurative fly on the wall. See Wollheim, The Thread of Life, p. 74 for more on central and acentral imagining.
 Richard Wollheim, The Thread of Life (New Haven and London: Yale University Press, 1984), p. 63.
 Ibid., p. 63.
 Walton, p.202.
Sarah E. Worth
Department of Philosophy
Greenville, SC 29613