A Theory of Ideology
J.M. Balkin

Yale University Press
New Haven & London

Copyright © 1998 by Yale University. All rights reserved.

Online version

With the permission of Yale University Press, the author has made an online version of this book available under a Creative Commons Noncommercial Sharealike (by nc sa) license, accessible at the author’s website at


Our theory of how ideological effects are produced requires us to search for cultural heuristics with three basic features: First, they must be transmitted through social learning or communication. Second, they must be stored in the memory of many different individuals. Third, people must use them to reason about the social world. Narratives fit these three criteria particularly well. Nar­rative and narrative structures are ubiquitous and pervasive features of cultural life. They are easily transmitted through communication. They are deeply en­trenched in human thought. Indeed, as we shall see in this chapter, human thought uses narrative structures for a wide variety of purposes.

Narratives are pervasive forms of human thought because narrative struc­ture is a particularly efficient form of human memory storage. Our minds are comparatively well designed to remember and understand narrative sequences. For example, people are better able to recall complex sequences of events in stories than complex lists of words and numbers. Indeed, translating informa­tion into narrative form is often an excellent method of memorization. This fact explains the importance of bards and epic poets in oral cultures, where information storage through writing is difficult, costly, or unavailable, and memorization skills are at a premium.

Human beings pick up narrative structures easily from watching and ob­serving events. We naturally seem to create narrative explanations for events or abstract narrative structures from our experiences. We glean narrative struc­tures from life; we impose narrative order on the world. For all of these reasons, the memes associated with narrative structures find a particularly hospitable environment in the ecology of human minds.

Narrative memory is memory of expectations of events in time. It is more than an ability to recall strings of sequences of events; it also involves the ability to store expectations about what usually happens under certain conditions. These expectations are coded in narrative form. People recall that A happened and then B happened, but they also remember that C is usually followed by D.

Our comparative abilities for narrative memorization have probably been shaped by evolutionary forces. Narrative memory structures are particularly useful for remembering what kinds of things are dangerous or advantageous, making complicated causal judgments about the future, determining what courses of action are helpful, recalling how to do things in a particular order, and learning and following social conventions that require sequential or script­like behavior. In the struggle for survival, storing sequences of events and ex­pectations may have proved much more useful than storing isolated bits of information in propositional form.

Whether or not there is an evolutionary advantage to narrative memory, human beings have a particularly well-developed capacity for it. As a result, people use narrative structures for many different mental tasks and operations. These multiple uses are examples of cultural extapation or bricolage--a mental ability or characteristic developed for one purpose is now put to many different purposes. And this particular extapation has far-reaching effects on the devel­opment of human culture.

Here are only a few of the things we use narratives for:

1. Remembering events in temporal sequences.

2. Ordering and organizing the past.

3. Explaining human action in terms of plans, goals, and intentions.

4. Understanding our own selves and motivations through autobiography.

5. Giving causal explanations of events.

6. Creating expectations about the future.

7. Internalizing expectations about how to behave in social situations and interact with others.

8. Providing scripts that tell us how to understand social situations, en­gage in social conventions, and assume social roles.

9. Creating notions of what is ordinary and extraordinary, expected and unexpected, canonical and deviant in social life.

10. Accounting for deviations from what is ordinary, expected, or canonical.

11. Creating social myths and shared memories that unite groups we are a part of, frame their experience of contemporary events, and produce shared expectations about how the group is supposed to behave.

In short, narrative is simultaneously a method of memory storage, a method of framing and organizing experience, a method for indexing and retrieving information, a method of internalizing cultural expectations, and a method of explaining deviations from cultural expectations. Because narrative is such a ubiquitous tool of understanding, it can also be the source of many different and powerful ideological effects.



Narratives as Networks of Expectations

In general, narrative thought organizes the world into a sequence of events, involving characters and their actions. This is the "plot" of the narrative. The plot and its constituent elements define each other: the plot situates and makes sense of the characters, actions, and events, and these in turn help constitute the plot. Usually the characters in a narrative have reasons for what they do, and their actions have goals. The narrative either assumes or directly ascribes purposes, beliefs, and intentions to the characters. Nevertheless, purely causal stories-for example, the gradual creation of a canyon due to water erosion­are also narratives, although they involve no human characters. Often there are anthropomorphic elements in such stories--we ascribe actions to particular inanimate "characters," like a river, even though we do not believe that they have plans or goals or act with intention.

The words of a story are only surface phenomena of its narrative structure. Equally important is the set of cultural expectations behind a story; they make a story comprehensible to us and allow us to draw inferences from it. When we tell a story we do not mention everything that happened; much is left to implication. For example, if I say that I had breakfast with Mr. Smith at Joe's Cafe, I do not mention every mouthful of food I ate. My listeners naturally assume that we went to a restaurant, that someone took our order, that both of us ordered food, that we ate the food, and so on. We do not speak about such events unless there is a reason to do so. We always understand a story against a background of other expectations that are also organized and stored in narrative form.

Many simple propositional sentences are actually narratives in disguise. Con­sider the sentence, "Mr. Smith and I discussed the game over breakfast at Joe's Cafe." This sentence not only states a fact; it also tells a story. But it does so only because it implicitly draws on a whole set of cultural expectations--for example, how to have a discussion, how to eat a meal with someone else at a restaurant, what kinds of things one usually eats at breakfast, and so on.

Thus, at its most basic level, narrative structure is a structure of expecta­tions, which are embedded in and connected to larger networks of expectations. These expectations play a dual role. First, they frame our understanding of what is happening. They give meaning to events. We attempt to understand what is happening in terms of expectations we already possess. We recognize patterns of behavior as meaningful in terms of patterns we are already familiar with. We create a story about what is happening based on stock stories-­expected sequences of events--that already lie to hand. Second, the expecta­tions that frame our understanding create the possibility of deviations from what is expected. These deviations call for explanation, and we employ stories to explain them.

Thus behind all narratives lie understandings about what is canonical, ex­pected, and ordinary. These understandings are themselves narratively organ­ized because they are stored in sequences of actions and events: this usually follows that; this is done on Sundays and that on Mondays; this is how you are expected to behave under these conditions; and so on. But these cultural ex­pectations are Janus-faced: storing information in this way simultaneously de­termines what is deviant, unexpected, and extraordinary in a situation. It creates an agenda for what does not fit our stock of existing narratives and therefore has to be explained. That explanation, in turn, will be phrased in terms of a story that ascribes motivations, intentions, and beliefs to an actor and relies on other stock stories about human behavior.

This is the dual character of narrative thinking: it focuses on, frames, and uses what is expected in human life, and it bestows legitimacy and authority on the expected. At the same time, narrative thinking lets us organize the exceptional and the unusual into a comprehensible form.[1] It allows us to learn by letting us match and reconfigure old expectations in light of new experi­ences. In this sense, narrative thinking is a heuristic device. It is one of the most basic of cultural heuristics.

Narratives as Norms

Cultural expectations also act as norms. The word norm has two meanings--a benchmark of what is ordinary or average, and a standard of what is appropriate. Similarly, the word normal can mean what is expected and what is appropriate to a situation. Narrative thought combines these two meanings. Our cultural expectations help us understand what is happening by reference to norms of what is expected in a situation and what is appropriate to the situation. Thus cultural expectations, stored in narrative memory, help frame social re­ality. As Erving Goffiuan pointed out, the frame we use to understand events shapes what we believe is happening and what is socially real. Events that seem normal or obvious in one frame become bizarre or inexplicable in another.[2] When the frame becomes controversial or blurred, it loses its framing character and our sense of what is socially real is disturbed.

Human beings organize their cultural behavior around expectations because this strategy saves effort in thinking and in determining how to act. Much of what we call cultural know-how involves expectations about what kind of situation we are facing and how to proceed in such a situation. Cognitive psy­chologists call these expectations scripts.[3] A standard example of a cultural script is knowing how to order and eat a meal in a restaurant. Situational scripts save us time and energy in figuring out what is going on and what we are expected to do. They are ready-to-hand narrative constructions that we adapt to various social situations. They offer us roles to play and ways to behave. We can think of them as narratives in which we are one of the actors, plots in which we play some of the characters. Scripts do the work of Goffinan's cultural frames. They set up expectations about what things mean, and they offer a background against which events and statements can be understood. Because of our restaurant script, when the waiter says to us, "OK, what will it be?" we understand that he is asking about our order and not the nature of the universe.

Much conventional behavior is oriented around such scripts, which is an­other way of saying that much conventional behavior is organized around co­ordinated sets of cultural expectations. When people go into a restaurant, they know what is likely to happen and hence they know how to behave appropri­ately. Moreover, they assume that others will behave in similar or complemen­tary ways. They assume that the waiter will approach them to take their order and not to extract their wisdom teeth. Thus most cultural understanding begins with a postulate of "situational normalcy": unless there are good reasons to the contrary, people tend to behave normally in accordance with the social situation that they believe themselves to be in, and according to the social roles expected of them in that situation.

This rule of situational normalcy underlies Paul Grice's theory of conver­sational interpretation. His Cooperative Principle is really a baseline of expec­tations about communication: we assume, without evidence to the contrary, that communications will be brief, truthful, relevant, and perspicuous.[4] When people deviate from these expectations, they cause us to search for explanations. Because people have departed from the ordinary scripts of conversation, we must make sense of their behavior in some other way.

Social scripts offer background expectations about what is happening, what is ordinary, and what things mean. These expectations literally go without say­ing, and that is why we do not usually speak about them. That is why our replies to the waiter are brief and perspicuous, to use Grice's terminology. Indeed, it defies cultural expectations for people to attempt to articulate the nature of a script that they are following in detail. Suppose a waiter approaches us and we say: "I see that you are a waiter. You are here to ask me what food I would like. The piece of paper in your hand is a menu. Give it to me and I shall tell you what I would like to order." The waiter would think we were crazy, or obnoxious, or performing some sort of psychological experiment. To talk about frames is to make them lose their character as frames--to make them a possible subject of analysis and contestation, which must be framed by some other set of expectations. Much deconstructive argument--and many ar­tistic effects--involve shifting cultural frames or making us self-conscious about them in order to disturb our sense of normalcy.

What is ordinary about the ordinary is precisely that we don't comment on its ordinariness, don't feel that it needs explanation or explication. Only deviations from the normal are worthy of comment. Cultural know-how, in this sense, is the ability to understand the ordinary and have our expectations confirmed by experience. Conversely, to lack cultural know-how is to fail to recognize the ordinary as ordinary, to lack expectations about what is going on and how to behave. So if a person asked us, "Why did that man walk over to you and give you that piece of paper?" and we thought that she was sincere in asking the question, we would think that she did not understand the cultural norms involved in eating in a restaurant. The remark would be evidence that this person lacked a certain kind of cultural know-how.

Matters are different, however, with acts or events that seem to deviate from the ordinary or the canonical. These things create puzzles that need to be solved or given meaning. Here again narrative structure plays a dual role. Narrative structures offer norms that give meaning to human action, but they also create the possibility of deviations from these norms. People must also be able to make sense of these departures as meaningful human actions. They also use narratives for this purpose.

When we encounter a person who seems to be acting in an unusual or unexpected fashion, and we ask why, we usually get an explanation in terms of a story that ascribes reasons, beliefs, and intentions to the actors involved. That is how one might account for the earlier example in which a customer elabo­rately described his actions to the waiter: "He told the waiter all these things because he is a social psychologist," or "He said all these things because he is a jerk." Often these explanations are offered in terms of their appropriateness to some other script or set of cultural expectations: for example, how psychol­ogists test people's reactions by doing strange things, how uncouth people tend to tease others, and so on. These actions make sense in terms of these alter­native social scripts. Justifications and excuses are familiar forms of narrative explanations. To excuse or justify behavior is to tell a particular kind of story about beliefs, intentions, and actions.

Yet narrative is not only a framework for making behavior meaningful; it is also a framework for understanding the psychology of others and attributing mental states to them. Narrative structures organize our use of psychological concepts like purpose, desire, intention, and belief. When we explain people's behavior through narratives, we simultaneously ascribe purposes, desires, in­tentions, and beliefs to them: "He ran out of the restaurant because he heard that his house was on fire"; "He arrived at the meeting thirty minutes late because he wanted to get an advantage in the negotiations." Narratives ascribe mental states to others (or to oneself) to justify or account for deviations from what is culturally canonical or socially expected.[5] Conversely, our ascriptions of belief and purpose make sense because they implicitly rely on background cultural expectations. It makes sense to believe that a person has certain beliefs or desires because of the way she reacts against the background of existing cultural conventions. In this way narratives mediate between beliefs, desires, hopes, intentions, and actions on the one hand, and existing cultural conven­tions on the other. They understand the former in terms of the latter. In short, narrative structures are a medium through which three facets of human life are understood and explained in terms of one another: (1) cultural conventions, (2) human behavior, and (3) beliefs, plans, goals, and desires.



Narratives as Stock Stories

As forms of cultural software, narrative structures can be passed to others through communication, imitation, or other forms of social learning. Many narrative structures are transmitted through mass media, through artistic ex­pression, and through myths and legends. Much art is based on narrative struc­tures, and our understanding of art is based on absorption and appropriation of these narrative structures. Adults and especially children like to hear stories told over and over again, just as they enjoy hearing a song played repeatedly. Listening to a familiar story can be pleasurable because it fulfills our expecta­tions. Hearing new stories reinforces or alters the existing stock of narrative structures that we use to make sense of what is going on in the world. Art manipulates and plays on our storehouse of stories, retelling them with inter­esting variations and details. And art can also replenish and expand our stock of stories by exposing us to new narrative structures, new ways of behaving, and new ways of understanding.

Although art is a crucial method of memetic transfer, one of the most important ways that we assimilate scripts and social expectations is through watching other people. The transmission of narrative structures through ob­servation is a good example of how the spread of memes differs from a simple copying of information. Usually people do not transmit social expectations like messages that are coded and uncoded. Rather, watching others acting out social scripts in front of us creates expectations in our own memories. Moreover, because people have different bodies of experience and different sets of prior expectations, they carry away different things from their social encounters. They assimilate behaviors and produce expectations in slightly different ways. As a result, each person in a given culture will have a slightly different set of social scripts, with slightly different expectations, and therefore each will un­derstand and react to the behavior of others slightly differently. Furthermore, social scripts are not simply routines that we must invariably follow blindly. They are platforms for innovation and improvisation. Precisely because nar­rative structures give us a sense of the world around us, they enable further development. People play and experiment with narratives and social scripts, producing new expectations that can, in turn, be passed on to others.

People employing similar narrative structures will understand the world in similar ways. Shared social meanings and conventions are not supraindividual entities but result from the interaction of distributions of relatively similar memes. There is no grand restaurant script in the sky, only different but rel­atively similar restaurant scripts stored in each of us. Nevertheless, these scripts often have an interlocking character. Our expectations about restaurants in­clude expectations about what it is normal to expect from others and what it is normal for others to expect from us. We not only expect that some things are the norm, we also expect that other people also expect that they are the norm, and that other people also expect that we expect that they are the norm. Interlocking expectations can have a stabilizing effect on social conventions and keep them from diverging too widely.

As each of us grows up, we gain a library of social scripts and stock stories. At any time we have an enormous number of stories and parts of stories in our memory. When we want to understand what is happening in society, we try to understand events in terms of an existing story or script. Our recognition of events as an example of an existing storyline creates expectations about how events are to continue. Events may surprise us, and then we try to reinterpret them as following yet another story line. In this way, we try to assimilate what is new in terms of what is old, improvising and playing different stories off against one another to explain deviations from our expectations.

It is important to emphasize the creative aspect of this process. Narrative understanding is not simply a matter of rote; it is also a framework for impro­visation and growth. Our library of stories and scripts is constantly increasing. We modify stories and scripts in the light of new experiences; these modifi­cations become part of our memory, used for understanding subsequent events. Suppose, for example, that we go to an Ethiopian restaurant where no silver­ware is served and people eat with their fingers. Over time we may develop a special set of expectations for Ethiopian restaurants. We may even be surprised if we find a fork on the table at the next Ethiopian restaurant, and we may conclude that the owners are catering to the tourist crowd. (Note how we ascribe motivations to explain deviations from what has now become culturally canonical.) New experiences rewrite our storehouse of narrative expectations, and we improvise on old stories to respond to them. In this fashion our cultural software is continually rewritten. An increasing variety of narratives adds flex­ibility to our framing of events and consequently our understanding of them. A person who has "seen it all before" is a person who has many different stories to draw on.

Like other forms of cultural software, new stories are created from older ones through bricolage. Parts of stories or scripts may be combined or grafted onto each other to form new ones. As a result, many of the stories and scripts that we possess bear structural resemblances to one another, even if they are used for widely different purposes. In the same way, we should also expect that many narratives and scripts widely dispersed in the larger culture will be strik­ingly similar, because they are common descendants of older stories and scripts that have been adapted to new ends.

Narrative understanding is a simultaneous process of organization and matching. To see the present as connected to the past we must already have begun organizing it into narrative form. There is more than one way that one can do this, because every event has many different "hooks" or indices that can connect it to many different stories or scripts. In my memory of a dinner I may recall that the waiter seemed rude, while my friend the oenophile will remember the quality of the wine. I connect this meal to previous stories of rude waiters while she connects it to previous experiences of great wines. Thus it is possible for different people to remember the same events in different ways because each sees its similarities to different kinds of stories and stores it dif­ferently in her memory.[6]

Narrative structures shape our thought because they organize our mem­ory of experience and our methods of memory retrieval.[7] Narrative structures provide "boxes" into which subsequent events can be categorized, indexed, and stored for later use. Stories and scripts are linked to other stories and scripts through this process. Experiences that do not conform to our existing forms of memory storage are more likely to be lost from memory.[8] Indeed, large amounts of our everyday experience are discarded because they do not mesh with our modes of storage. Many aspects of life will be lost to us or re­membered in highly limited form if we have a limited stock of stories to serve as an interpretive matrix for categorization and memorization. Just as a pigeon cannot make sense of Hamlet, so a person with only a very small set of stories and scripts will not glean or recall very much information from her experiences.

Not only do we tend to retain memories that conform to our existing narrative structures, we also tend to alter our memories to conform to our canonical expectations about and representations of the social world. Memories that cannot be altered to fit to our expectations may be forgotten or may be deliberately highlighted as exceptions that need explanation. In one famous experiment, college students were asked to tell each other a Native American tale. They either forgot the elements that were unconventional from their own cultural standpoint or transmuted them into something more conventional and expected.[9] This result is hardly surprising. If we don't understand what is going on in an experience, it is more difficult to remember it; it is much easier to understand events in terms of our existing stock of stories. This is consistent with an evolutionary model of memetic development: our minds form an ecol­ogy in which certain memes are more likely than others to take root and thrive.

On the other hand, we can, with sufficient effort, gain new sets of cultural expectations. This will change our mental ecology. If we study Native American culture, the previously alien elements of the story gain significance for us and we can remember them when we retell the story. We may even highlight them to our audience as an example of what is distinctive about the culture.

Political and legal rhetoric gains much of its power from these features of narrative framing. People naturally attempt to explain gaps in events for which they have no direct evidence, or events that they do not wholly understand, in terms of familiar stories and scripts. Once they have settled upon a story to frame events, it can exercise great power over their imagination, leading them to make unwarranted inferences and prejudicial judgments.

During the Supreme Court confirmation hearings of Justice Clarence Tho­mas, Professor Anita Hill accused him of sexually insensitive conduct. Because of conflicting testimony, it was difficult to know whom to believe. Thomas's defenders, however, invoked the plot of a contemporary movie, Fatal Attraction, to paint Hill as a spurned lover who was seeking to destroy her former boss's reputation and career. Hill's accusations could then be reinterpreted as those of a calculating, unstable vixen. Thomas himself invoked a stock story about whites who attempt to keep "uppity blacks" in their place, and he accused the Senate Judiciary Committee of staging a "hi-tech lynching." These narrative framings had enormous rhetorical power and may have helped turn the tide in favor of Thomas's eventual confirmation.

Trial lawyers have long understood the power of narrative framing. They attempt to lay out a story of how events occurred during their opening argu­ments in the hope that the jurors will use the story to frame the evidence they hear. Getting the jury to accept one side's story as the most plausible frame­work for the events of the trial is often tantamount to winning the case. That is because once a story is accepted, it is used to filter and organize all of the evidence subsequently presented. Like most people, jurors tend to discount or ignore evidence that does not fit their organizing story, and they will alter or simplify information so that it does conform. Evidence that can be made to fit actually tends to reinforce the power of the story, because it seems to confirm it, even though the same piece of evidence could also be consistent with a very different story. Because narrative framing is so powerful, lawyers faced with the other side's story realize that if they are to win the case they must offer an equally plausible counterstory that also fits most of the evidence. Often the only way to dislodge a narrative is with another narrative that also fits most of the facts but shows them in a very different light.



Personal Narratives

Judgments of human character are also organized around narratives. We form expectations about people's behavior and ascribe attitudes to them that are consistent with these expectations. When we describe what individuals are like, we often do so in terms of stories about the sorts of things they usually do or the kinds of things we normally expect of them. We often judge people and explain them to others through anecdotes that reveal their characteristic be­haviors and attitudes. Family members often have stock stories that they tell about other members. These stories not only describe the character of partic­ular family members but also their place in the family, as beloved firstborn son or black sheep.[10] We also use anecdotes to characterize organizations and even entire cultures. Travel guides often contain a wealth of anecdotes that create expectations about how members of a given country are likely to behave.

Just as social scripts are made from fragments of older ones, our expecta­tions about people are created from expectations about other people who seem similar to them. Moreover, in our culture we have a well-developed set of stock characters-the miser, the ladies' man, the clinging mother, the neurotic in­tellectual, and so on--that we use to frame our understanding of others. These stock characters are templates of expectations that we use as building blocks to form our expectations of particular people that we meet. Although our expec­tations about people may change greatly as we learn more about them, the initial framing of individuals as fitting a certain stock character may have a significant effect on the future development of our expectations about them. That is because we will tend to behave toward them according to our existing expectations. In this respect, it may well be true that first impressions are lasting ones.

Often our pool of stock characters is tied to ethnic and gendered stereo­types. We have stock stories about how whites and blacks or American tourists and French waiters normally behave (and therefore are expected to behave) with respect to each other. We have literally hundreds of stock stories about how men and women behave, for example, with all sorts of variations--the macho man, the passive wallflower, the stupid hunk, the femme fatale, the henpecked husband, the ditzy blonde. Each of these stock characters can form a template for organizing and giving meaning to our encounters with others. They are important parts of stereotypical thinking and often have significant ideological effects.

Narrative structures are extremely important in interpersonal interactions, and especially close personal ones. Individuals in the early parts of an intimate relationship often talk incessantly with each other so that they can create a set of expectations about who the other person is and what he or she is like. They create cultural software in each other for mutual understanding. As time goes on they tend to spend less time in this sort of talk because the pictures are starting to become more fully formed and they have less need for new infor­mation. Getting to know people is, in a large sense, creating a set of expecta­tions about them and about their behavior, and ascribing attitudes to them on this basis. In the early stages of a relationship, surprises and unexpected be­havior may be desirable, because they are opportunities for learning more about the other person and developing new expectations. Later on in a relationship, surprises may have the opposite effect--they may lead to the unsettling feeling that we do not really know the other person any longer.

In times of crisis, people often need to reconfigure their views about each other, and they may start talking a great deal again. One reason people in long­term intimate relationships "never talk anymore" may be that they don't feel that they have to. But when ways of behaving are no longer satisfying--because of a crisis, for example, or because people have grown apart--their old expec­tations may no longer be adequate. At that point, the fact that people "never talk anymore" becomes a real problem. As a result, people may feel compelled to start intensive discussions about themselves again, in order to reconfigure their expectations about each other and preserve their relationship.

We use narratives not only to describe our personal experiences to others but also to understand them ourselves. Often people feel the need to talk about their experiences to others so that they themselves can comprehend them. The act of talking organizes experience into narrative form so that it can be under­stood and memorized. The need to talk and describe what has happened may be especially great with regard to emotionally powerful experiences. People must connect these experiences to narrative constructions that they already understand and to features of their lives that they already recognize. This may require considerable narrative work.[11]

Narrative construction of personal experience is inevitably partial; it selects certain features of experience as meaningful because narratives are organized in terms of what is already understood to be meaningful. What cannot be so organized is usually and eventually forgotten over time. As a result, narrative memories of the same event by different people can vary widely. Each person remembers what is most salient to her, given her existing cultural software and her special preoccupation.

Just as people ascribe purposes and motivations to the behavior of other people so that they can understand it, they do the same with their own behav­ior. When asked to describe themselves, people often give stories about what they have done in the past and why they did it. They offer anecdotes about their past that symbolize the kind of people they believe they are and the way they usually behave. They describe events that have shaped them and have made them who they are today. Equally important, people construct narratives of their lives. They understand who they are, what is happening to them, and what they should do next by means of narratives. These narratives are stories in which they are the protagonists. Such stories often fit well-established pat­terns--the Bildungsroman, the adventure story, or the picaresque novel. They portray people's lives as comedy, tragedy, or even farce.[12]

Personal narratives organize and give meaning to previous experience. They also provide a form of justification. We justify who we are to ourselves in terms of a story about what we were before and what we have gone through. Perhaps most important, autobiography is a form of prophecy.[13] Our personal narra­tives-whether comedies or tragedies--can be seen as a kind of script. And scripts are meant to be followed. A personal story has a trajectory, a trajectory that demands to be filled out through future action. A personal story is a set of expectations about the self that demand to be fulfilled in practice. If we see the story of our lives as a tragedy, we may understand what we must do next and what will eventually happen to us in tragic terms. The role of personal narrative in framing the possibilities of our future actions, and thus in limiting or empowering us, cannot be overstated. In extreme cases, we can become the slaves of our personal narratives.

Although autobiographical narratives are deeply personal, they also make use of stock stories and elements available in the surrounding culture. We understand whether we are successes or failures, good or bad persons, in terms of social roles, stock stories, and stock characters. Our own narrative under­standing of ourselves is composed out of elements that we get from the larger culture--from movies, television, family anecdotes, social mores, and cultural expectations.

Moreover, our individual narratives are strongly influenced by our cultural heritage. Our ethnic and religious identity forms a template of expectations about how to behave toward others, how to be a man or a woman, how to act toward our children or our parents, and so on. Our understanding of ourselves as Jewish or Italian, Korean or black, already preshapes and constrains the possible stories we build upon and the kinds of futures we feel that we can have. These effects on personal narrative are another example of the inextri­cable relationship between the personal and political, the individual and the cultural.

People use and develop scripts for their interactions with others. Just as people learn how to order meals in a restaurant by watching others, so, too, they learn how to get along with others, handle and avoid conflicts, love and be loved by watching their parents and others close to them. They develop narrative expectations for how to be a friend, a lover, and a parent. They modify and rework the scripts they have learned in previous social settings to form scripts for dealing with people in new situations. Our expectations about social relations with others are produced through bricolage from previous relation­ships.

People learn how to be parents, for example, from watching their own parents; this learning helps them create roles that they naturally slip into in their dealings with their own children. People absorb lessons about how to deal with and love others from their parents and others close to them; they apply these lessons to their subsequent relationships with others. These scripts may be particularly awkward examples of bricolage--old tools badly adapted to fit new social situations. Yet people cling to these scripts because they do not know how else to perform these roles.

We often employ the narratives of others consciously or unconsciously as models for our lives. We may absorb the stories of our parents into our own personal narratives, for example, using them as the raw materials to develop our own personal stories. As a result, we may feel unconsciously compelled to play out parts of these stories in our own lives. A man whose father failed at business may absorb this story into his own personal narrative and reenact it as part of his own life story. Although we are hardly doomed to repeat the narratives of our parents in exact detail, parts of their stories may still be im­portant elements in what we eventually do construct. And, as is so often the case with bricolage, even our modifications and innovations may bear the char­acteristics of what was built upon.

The narrative nature of human self-understanding and human social inter­action explains why psychological therapy has historically turned to narratives as a means of treating patients. Much psychological therapy involves recount­ing, interpreting, and reconstructing the patient's story and critically examining the scripts she employs.[14] Through recounting and revising stories about her­self, the patient begins to recognize how she became the person she is today. She learns to identify the sources of the scripts and expectations that underlie her reactions to people and events. She tries to see how living according to these scripts and expectations is keeping her from a happy, healthy, life. To­gether the patient and therapist try to modify her scripts and expectations by substituting new narratives for old ones.

In short, successful therapy teaches the patient to develop new scripts and write new cultural software through repeated narrative construction by patient and therapist. The patient learns to reorganize the past in new ways, to see previous events in a different light, and to form new expectations. The hope is that these new expectations will lead to new and more positive behavior.

From a purely physical standpoint, it has often seemed puzzling how the mere recitation of stories could effect any improvement in a person's mental condition. But if people's behavior is shaped by the narrative constructions that they use to understand themselves and interact with others, if much social thinking relies heavily on scripts and expectations, the idea of a "talking cure" is not at all far-fetched. Supplementing or replacing old narratives and scripts with new ones might change people's behavior for the better. The problem is that the means of doing this must necessarily be as much an art as a science. And it is by no means clear that there is only one route to the creation of new and healthier scripts.

There is an important analogy between the personal narratives and scripts that hinder our development and make us unhappy on the one hand and the ideological effects of cultural software on the other. Just as people's cultural software contributes to social injustice, it can also contribute to their personal unhappiness. Indeed, some conceptions of justice do not draw a sharp distinc­tion between these two concerns. Under those conceptions, the idea of justice also applies to the self, so that one can speak of people being unjust to them­selves. This injustice is not a matter of bad behavior toward others (although this can be involved). It is rather an injustice to the possibilities of what we could be.

The narratives of our lives and the social scripts that we employ in inter­acting with others can lead us repeatedly to act in self-destructive ways or in ways that prevent happy and fulfilling relationships with others. We may seek out lovers who abuse or manipulate us, for example, because we are replaying scripts about how to love and be loved that we began organizing in early child­hood. We may deal with conflict in ways that we worked out for situations that happened long ago, strategies that are inadequate to the situation that now faces us. We may generate unhappiness and conflicts with our children because we are following scripts of how to be a parent that we assimilated from our own parents.

These personal scripts and narratives and their unfortunate consequences for our lives are like the ideological effects of cultural software, except that they act at a very personal level, and our concern with them is not that they produce social injustice but that they hinder personal growth and personal happiness. Just as we must take an ambivalent attitude to our cultural software because it has the capacity to produce injustice, we must take an ambivalent attitude to­ward cultural software because it has the capacity to produce personal unhap­piness. Personal scripts and roles that we have assimilated may be partially adequate to deal with many of the problems and situations we face in our lives. That is probably why we developed and adopted them in the first place. But in new contexts and new situations, their inadequacies become increasingly apparent, and they begin to hamper our lives. The goal of successful therapy is to build newer, more adequate expectations out of older, less adequate ones.



Group Narratives

Just as individuals have stories that they use to understand themselves and the world around them, so do entire cultures and countries. Each society has stock stories drawn from its past that are told over and over, and, in this retelling, take on a mythic status. These stories symbolize what is most important to the society, its values, its sense of itself, and its relationship to the outside world. The stock stories of a society are abstracted and condensed through frequent retelling, and eventually can even be encoded as single icons, which can be either persons or events. Examples in American history are "Lincoln," "Pearl Harbor," or "the Alamo." Each of these icons condenses and thus represents a story with a rich set of historical associations, often contradictory and con­tested. Such narrative icons invoke not only a particular order of events but also a tradition of interpretations that grows up around these events.

These stock stories and icons form part of what is variously called collective memory or social memory. It is an excellent example of widespread memetic transfer and assimilation. Social memory is an example of an endemic cultural virus. Narrative memory is spread from generation to generation through com­munication and becomes part of the shared cultural software of a culture or society. Members of a society or culture repeatedly tell each other stories about important events in their history. These events often have deep emotional res­onances; they are still able to evoke anger, solidarity, pain or pride long after the events have passed. Examples are wars and revolutions, depressions, riots, strikes, famous trials, and genocides.[15] These important events become bench­marks for comparison with later events, an index through which to understand what is happening to the culture. As the memories are passed on through the generations, they are stylized, pared down, and altered, much like any other story. These social memories are thus at the disposal of storytellers, the mass media, and even the state, to rework and reorient.

People share social memories because they are members of a common meme pool. As a result, entry of new individuals into the meme pool can alter social memory; even though widely shared memories can be assimilated into the new members, they may also bring with them new stories from different cultures.

Social memory is distinct from other endemic cultural software in an im portant way. The memory of important events is not simply shared by members of a culture. It also unites them or divides them, gives them something in common or produces a bone of contention. Some especially divisive events, like the Vietnam War, the Dreyfus trial in France, and (most likely) the O. J. Simpson trial in the United States, retain their ability long afterward to invoke conflicting meanings and reinstitute old social and ideological divisions.[16]

Sometimes the divisions created by a momentous event will be resolved or mediated by the creation of a stock story that is roughly satisfying to most of the contending groups. American culture has produced such a narrative of the Civil War: the North fought for freedom, but the South is acknowledged to have fought valiantly and bravely, and the conflict is seen as deeply tragic. The noble figure of Robert E. Lee--who gave up the chance to command the northern armies in order to defend his homeland of Virginia--plays an im­portant role in symbolizing southern heroism against enormous odds, as well as the deep, familial connections between the two sides even in the midst of a vicious war. This narrative allows both sides to accept and even celebrate the outcome of the Civil War--through the staging of mock battles and the col­lection of memorabilia, for example. Of course, this narrative is most appealing to American whites, since it sees the conflict as one between honorable brothers who are eventually reconciled. It operates only by downplaying the evils of slavery and the subjugation of an entire people. Thus it is not a completely effective narrative resolution. And it breaks down precisely in moments of heightened racial awareness.

Just as a person draws on a stock of stories to frame and understand what is currently happening, so members of a culture draw upon its stock of stories or myths to frame and understand what is happening to them. These myths organize experience and the culture's reaction to that experience. The value of myth is that it helps us understand what is new in terms of what is already understood. As we saw previously, stories save time and energy in figuring out what is going on, what is socially real, what it is appropriate to do, and what is likely to happen in the future. The myths of a culture reorganize the world to take on the appearance of a story that is already well understood. Thus an attack on American interests is readily envisioned as another Alamo or Pearl Harbor. If the new experience fits the old story well enough, the myth will be strengthened and confirmed. But experience is often recalcitrant. Although we try to understand social events in terms of stories that lie to hand, the world will not always so easily conform. Viewing every conflict as the moral equiv­alent of Pearl Harbor and every war as World War II is a recipe for disaster. So a culture's myths are gradually adjusted and reconstituted in order to take account of the changing world it faces. The stories of the past are given new glosses and new meanings. In this way old icons like "the Winning of the West" or "the Civil War" take on multiple and conflicting associations. Widely shared stories serve as a springboard for future improvisation. Like other memes, myths are transformed as they are communicated to others in new circumstances. Thus the myths of a society are not simply handed down; they are subtly remade in each generation for its own purposes, even as that gen­eration is guided and shaped by what it works with.[17]

Occasionally new stories and icons replace older ones. "Vietnam" has be­come a permanent fixture in American consciousness, with a highly conflicted and ambivalent set of meanings. Yet as with all cultural bricolage, the new is always built on the old. A culture's new stock stories are constructed out of variations on older narratives, which in turn were based on still older frame­works originally used to understand the problems of the distant past. Thus it is hardly surprising to find that the stock stories of different cultures have many elements in common, even if these elements have now come to be used for widely different purposes and have quite different meanings.

Because a culture's stock stories are a shared way of understanding the shared past, present, and future of its members, media of mass communication take on a special importance in the creation and development of social myths. They are a crucial determinant of the ecology that narrative memes face. The stories and genres produced by mass media are evidence of a culture's most pressing and abiding concerns. Yet the products of mass media cannot be taken as the endpoint of analysis. They are not simply and faithfully absorbed by members of a culture. Like all memes, narratives take root depending on the existing cultural software of the people who absorb them. Viewers and listeners appropriate and reinterpret the narratives produced by mass media. They are the raw materials for the work of cultural memory performed by individuals. And the sum of these individual appropriations and reinterpretations of stories, in turn, shapes the way that the mass media will present these stock stories in the future.

Richard Slotkin has argued that a central myth of American consciousness is the story of the frontier: a story of repeated separation, regression, and re­generation through violence.[18] According to this story, Americans leave civi­lization--whether Europe or (later on) the cities of the Eastern Seaboard--and spread out into the wilderness, where they find that they must live more prim­itively and simply. There they encounter savages who threaten their survival and whom they must conquer. American progress comes from leaving the old world behind, temporarily regressing to a more primitive or natural state, tam­ing the wilderness, and defeating the savages. To succeed, Americans must both separate themselves from the older authority structures and class privilege of civilization and defeat the savage of the western wilderness. The familiar icon of "Cowboys versus Indians" is only one of the many variants of this central myth.

In this account, the American hero is one who confronts the wilderness and makes it his own. The American hero is one, like Natty Bumppo or Daniel Boone, who "knows Indians" and absorbs this knowledge to transform himself and civilized society. For these American heroes, the war against savages is mirrored by a struggle in their own souls in which they conquer and discipline the savage side of their own natures.[19]

The story of the frontier has several corollary stories, including the myth of bonanza: even though the frontier is dangerous, it is often a place of un­bridled opportunity in which great wealth can come from comparatively little effort. Freedom comes from conquering the frontier and making use of its manifold opportunities.[20] Repeatedly, events in American history--from the California Gold Rush to the Roaring Twenties to the deficit-financed boom of the 1980s and the hi-tech expansion of the present--have been conceptu­alized in terms of the boundless wealth of the frontier. In the myth of the bonanza, all things are possible and everyone can grow rich, at least until the gold reserves dry up, the stock market crashes, or the savings and loans go bankrupt.

The myth of the frontier has been used repeatedly by Americans to un­derstand who they are as a nation and how they should behave in the many crises that have faced them. The story of the "savage war" has been used not only to justify the expropriation and extermination of the Indians but also to understand many other crises and problems, both domestic and international.[21] Labor conflicts, race relations, the Cold War, and the war in Vietnam have all been readily envisioned as new versions of the mythical "war against the sav­ages," in which annihilation of the enemy is the only acceptable way to preserve American civilization. The conception of the savage war projects the difficulties of American life outward onto some hypothesized Other--whether it be Native Americans, labor unions, immigrants, international communism, or the con­temporary black urban male-who is seen as a dangerous element that must be defeated or controlled, and "who becomes the only obstacle to the creation of a perfect republic."[22]

One might contrast this stock story of American life with the stock story of Judaism. Because Jewish culture is much older, it has a wider range of nar­ratives to draw from. Nevertheless, the most enduring stock story of the Jews is a cyclical myth of dispersion, persecution, and redemption. It is, roughly speaking, the story of the Exodus. In this story, God reveals himself to the Jews and promises them his everlasting protection if they will obey him and spread his wisdom to other countries. The Jews are then dispersed from their homeland (usually because of their previous misbehavior) into foreign coun­tries. They are, as Moses says when he flees into Midian, "stranger[s] in a strange land." Like Joseph, they benefit their adopted lands and rise to prom­inence. Eventually, there arises "a new King ... who knew not Joseph"-who does not understand the value that the Jews bring to civilization. The new political forces persecute the Jews: "He set over Israel taskmasters to afflict them with burdens. And he made them serve with rigor." God hears the cry of the Jews and remembers his promises to them. He delivers them from bond­age "with a strong hand and an outstretched arm, with signs and wonders." He then reveals himself to them, gives them his Law, and delivers them to their Promised Land. This story is implicitly cyclical. The Jews begin and end in their homeland. Once in the Promised Land, they can then be dispersed again for their misbehavior, and the cycle begins anew.[23]

This story has enormous pull over Jewish culture and thought. Many Jewish holidays are organized around the myth or various stages within it. The most obvious is Passover, when Jews are directed to tell their children the story of the Exodus. The Passover Haggadah even instructs them that they are to regard themselves as if they personally had been delivered from Egypt. Two other major festivals, Succot and Shavuot, celebrate different aspects of the redemp­tion story-the dwelling in the desert and the reception of the Law. The story of Mordecai and Haman in the Book of Esther is yet another version of the myth, as is the story of the Maccabees celebrated at Chanukah. The fast day of Tisha B'Av commemorates the destruction of the First and Second Temples and the beginning of two different diasporas. The most recently added Jewish holiday is for the commemoration of the Holocaust. In one sense the Holocaust fits well into the governing myth of Judaism, because it is a catastrophic ex­ample of oppression, which is followed by the founding of the State of Israel. In another sense it severely strains the myth, because the sheer scope of the persecution involved in the Holocaust seems to dwarf any previous misbehavior by the Jewish people and any promise of eventual redemption, even the creation of a Jewish homeland. The Holocaust is an example of a historical event that has tested and reshaped the fundamental myth of a culture.

Each of the myths just offered is derived from previous sources. For ex­ample, the American myth of the frontier probably borrowed from the story of the Jews in the wilderness who gain the land promised to them by God. Note, however, that these two stories take on very different trajectories and that the wilderness serves very different purposes in each of them. If anything, the myth of the American frontier seems to borrow most heavily from the story of the Jewish wars to conquer Canaan and the divine injunction to destroy the Amalekites.

The great danger of myths, like personal scripts, is that they not only frame ur understanding, they also invite us to play them out in our lives. Narratives are not only tools of understanding but also tools of action. As heuristics, they save time in understanding a situation because they provide ready-made social meanings to events and ready-made roles to play in them. Narratives are scripts, and scripts are made to be followed. Yet cultures are by no means destined to play out their constitutive myths over and over again. Any culture that has existed for long periods of time has many different stories and myths to choose from. Moreover, social myths are not simply a script to be followed blindly; they provide opportunities for reinterpretation and a platform for in­novation. Each generation has reread and rewritten the dominant stories of its past to serve the needs of the present. Social myths lend themselves to this reinterpretation precisely because they have such deep resonances and such a wealth of associations. People can and do draw on countermyths; they can reinterpret, reread, and rewrite existing stock stories to meet present-day con­cerns and crises. For example, Americans have not only the myth of the frontier but also the story of Thanksgiving, in which the Indians aid the Pilgrims and the two share the bounty of the land and jointly give thanks to divine provi­dence. This version of the Thanksgiving myth, one might think, would be much more helpful to a country that is trying to live down the sorry history of previous persecutions and trying to accommodate people of many diverse cultures, than is the myth of the savage war.

These examples show how narratives function as cultural heuristics. They help constitute much of our social understanding and make possible much of our social existence. Yet at the same time it is clear that they can have serious ideological effects. They can produce stereotypical thinking or lead us to reen­act them in wholly inappropriate situations. Hence our attitude toward them, as toward all cultural software, must be ambivalent.

Moreover, as heuristics, narratives are necessarily partial. They can describe and store in memory only certain features or aspects of a situation. The world is too complicated and multifarious to be captured in a single narrative account or even in a series of narrative descriptions. This inadequacy of narratives is the flip side of their advantage to us. Narratives are useful memory structures precisely because they select and organize our experience--they categorize and store events into scripts or indices that we can use for later comprehension and comparison. Narratives are useful tools of understanding because they create social expectations that frame our understanding of what is and should be hap­pening; without such expectations, we literally would not know what to expect. Because narrative structures work in these ways, they necessarily lead our un­derstanding in some directions rather than others. They categorize future ex­perience in terms of preexisting indices and expectations. This produces the familiar trade-off of any heuristic-although these expectations may be good enough for some purposes, they may seriously hinder our understanding and promote injustice in others.



Narratives and Justice

Narratives are intimately connected to questions of justice. We use narratives to describe human plans, goals, and intentions, which are often necessary to judgments of what is just and unjust. Moreover, our explanations of what is just or unjust in a situation often depend on a narrative account of how that situation came to be. Narratives connect-and sometimes fail to connect-the misfortunes and inequalities of the present to the events of the past. The fact that my wallet is empty looks very different depending on whether I spent all my money or it was taken from me; my injuries have a different meaning depending on whether I have tripped over something or have been kicked.[24]

To be sure, theories of justice need not be based on the events of the past. Under some theories of justice, we can judge something as just or unjust simply on the basis of the existing distribution of resources. But the explanation of why that distribution is unjust inevitably will rest on some narrative account that describes either the meaning or the consequences of the distribution. For example, one might hold that a distribution of resources is unjust because it conveys a certain meaning about the worth or dignity of the individual or because it is likely to prevent equal chances for human happiness in the future or is likely to have oppressive or wrongful consequences. In each case, we still offer a narrative account of the evils of the present state of affairs. Our sense of justice inevitably has a narrative character, whether our concerns are cor­rective or distributive, whether our theory is deontological or consequentialist, and whether our vision of justice is forward or backward looking. Indeed, the very fact that our notion of justice looks in any direction at all means that some narrative underpins our accounts of what makes a situation just or unjust.

Because of the connections between narrative and justice, social memory is an essential framework for judging questions of justice and injustice, and con­trol of social memory is an important ideological tool. The history that has produced present holdings and present injuries is usually important in assessing whether people have been treated justly or unjustly. If people forget the past or if it is disguised from public view, the world seems a blank slate or a level playing field: past injustices are forgotten and present debts are wiped out. A person's or a group's social situation and respective life chances are more easily seen as no one's fault in particular and as primarily the responsibility of each person or group. If narratives of previous injustices are forgotten, distorted, or replaced by false narratives of reconciliation and recompense, serious injustices may go unacknowledged and unremedied. Loss or distortion of social memory can bury past injustices and make present distributions of power, wealth, and other social goods seem unfreighted with previous wrongs. It is no accident that the words amnesty and amnesia come from the same Greek word meaning "to forget."[25]

Yet the memory of previous injustices is not always an unqualified good. If past injustices become deeply and pathologically lodged in our identity, they can adversely affect how we deal with others. Fixation on the memory of past wrongs may be necessary to preserve the memory of unremedied injustices or as an important lesson for future generations, but it can also hinder our per­sonal growth. Both forgetting and remembering can be pathological.



Narratives and the Canonical

Narratives can also have ideological effects because they shape our sense of what is canonical and normal, and hence our understanding of what is different and deviant. Our sense of the canonical includes not only expectations about behavior in restaurants but also expectations about how blacks and whites or men and women are likely to behave. Race and class relations are organized around stock stories about members of different groups. Debates about welfare policy and immigration are often based on anecdotes about the behavior of welfare recipients and immigrants.

Expectations about social groups have both a descriptive and normative character; they can describe not only how things usually are but how they are expected to be. Situations that conform to canonical expectations require no special explanation or justification. But deviations from the canonical--like a black man walking in an all-white suburb or two men kissing in a public park--­call attention to themselves as unusual and special and demand either expla­nation or justification.

Moreover, people often understand majority or superordinate groups as canonical in both descriptive and normative senses. Racial minorities, women, homosexuals, and the physically and mentally disabled are generally viewed as being "different." This "difference" arises from an implicit comparison to ca­nonical norms of whiteness, maleness, heterosexuality, and lack of disability.[26]

Understanding people as "different" often means identifying their differ­ences from the canonical as the cause of any disruption of settled expectations. What is canonical forms the baseline of expectations; because it is regarded as normal, it is not in need of special justification. Thus a person in a wheelchair cannot climb stairs to enter a building, while "normal" people can. The cause for this situation is located in the disability of the person in the wheelchair, not in canonical social expectations and settled social institutions. Accommo­dating disabled people may require adding ramps, widening doors, and providing elevators. Such accommodations may be understood as special treatment even when they are done in the name of equality. The notion that equality entails "accommodation" or "special treatment" preserves a sense of what is canonical even as it authorizes deviations from it. It locates difference--devi­ation from the canonical--in the minority or subordinate group rather than in the social relation between groups.[27]

Our narrative expectations about what is canonical and ordinary are far more than isolated sets of beliefs about men and women or about whites and blacks. They are forms of cultural know-how in the truest sense, for they offer us scripts about how to possess a particular identity in a culture and expecta­tions about what to expect from people with different identities. Cultural know-­how is in large part the ability to understand what is canonical and to have our expectations confirmed in social practice. This is not necessarily an unmitigated good. Some of these expectations are essential mechanisms in the preservation and reproduction of unjust hierarchies. Considerable cultural knowledge is nec­essary to behave according to the dictates of a status regime, including an unjust one. One of our most finely tuned social skills is the skill of understanding our status relative to others and the consequences of that comparative regard. This skill is used over and over again in our everyday interactions.

Children are taught how to operate within status regimes through social learning and imitation. They absorb new cultural software that makes them socially competent to be racist or homophobic. People who lack the requisite cultural software may commit social faux pas because they do not understand the social meaning of situations; they do not discriminate in the same ways that others will.

In the previous chapter I compared racism to a virus or disease transmitted through social learning. But such a "virus" is also a form of cultural know-­how. Hence one must also understand racism as a social skill, often deeply ingrained in the cultural software of individuals. It is not a lack of cultural knowledge but a particular manifestation of it. We often say that racism is born of ignorance, but in another sense this is hardly so. The truly ignorant person is the one who does not understand the system of racial caste and therefore does not know how to behave within it. To participate in a regime of racial status requires delicate and complicated social understandings. It demands a considerable degree of fluency in the language of injustice.

Narratives That Make Themselves True

Narratives do more than simply distort or limit understanding. They also have the ability to "make themselves true" through their use. Because social meaning is part of the social world and is constructed in part by narrative understandings, narrative organization is folded into the social world, becoming part of its fabric and shaping its future evolution.

First, shared narrative structures help create intersubjective social meanings. People use narratives to understand the nature of a social situation; when many people share the same set of expectations about what is happening--for ex­ample, when they employ similar conventions and similar assessments of social meaning--their interlocking expectations establish what is the case socially. If people in a society share the view that people with darker skin have lower status, for example, then people with darker skin do have lower status in that society, whether or not this is just.

Second, narrative structures reconfigure and add to the meaning of past events. Narratives cannot change the past, but they can change how people remember the past and what the past means to them. People remember the past in terms of a set of narratives. These narratives bestow meanings to past events that the participants in those events may not have shared. The American Revolution is a good example. The Revolution, like any other part of the past, does not arrive in a premade narrative package with premade social meanings. The meaning of the Revolution must be constructed out of the memories and stories of the persons who participated in it, those who witnessed it contem­poraneously, and those who come afterward. These narratives do not exhaust what happened during the Revolution. They are partial in both senses of the word-they are both incomplete and biased in their organization and charac­terizations. Yet they add something to the social meaning of the American Revolution. The narratives produced about the Revolution become part of the understanding of the people who lived through those times as well of those who came later. These understandings are then passed on to others, who in turn add their own interpretations and stories about these events.

In this way a tradition of understandings and narrative expectations about the American Revolution grows up. This tradition is always in flux, shedding old meanings and gaining new ones. Its associated memes develop and mutate as the tradition is passed on from person to person and from generation to generation. The resulting narrative structures are folded into the tradition and become part of the social meaning of the American Revolution. They need not form a homogenous whole and may be in conflict with each other. As a result, the social meaning of the American Revolution may be contested and contra­dictory. Nevertheless, the cumulative social meanings of the Revolution are quite real, even if they embellish or mischaracterize events. The existence of a tradition of representations and the palpability of its effects are distinct from the accuracy of the representations themselves.

Third, narrative structures can make themselves true in practice because they subtly direct the actions of people who frame experience according to these narratives. People tend to characterize situations that they face according to the scripts or narrative accounts that they possess. These scripts not only describe the nature of what is going on but also offer the nature of a proper response. Depending on the narrative structures that we possess, the same be­havior can be interpreted as a mere social slight, a misunderstanding, an ag­gressive action, or a vital threat to national security. This interpretation can shape our response; the response, in turn, can induce behavior from others that confirms our worst fears.

Antagonistic cultural groups and nations often understand each other's ac­tions according to a previously prepared script of expectations that both reflects and reproduces distrust. Their internalized story lines about the social meaning and the likely course of their relationship may exacerbate tensions between them. Their expectations may lead to mutually self-destructive behavior unless the parties learn to interpret each other's actions and intentions in different ways.

Stock stories are one of the most powerful sources of social and personal prejudice, not merely because they frame the nature of events but because they are mechanisms of self-fulfilling prophecy. People with particularly powerful stock stories have expectations that tend to dominate their interactions with others. They tend to make the world fit these stories by understanding events as confirming examples of their powerful and well-worn narratives. Seeing the world and the behavior of others in this way tends to place others in the roles designed for them. In this way scripts can sometimes "make themselves true." If we see the actions of others continuously in terms of expected slights to our ego, for example, we may well behave in ways that bring about what we fear most. People whose personal scripts end with rejection or abandonment may orchestrate events so that rejection or abandonment becomes likely. If we have stock stories in which a certain group of people is worthless and undeserving, they are likely to treat us with disrespect and hostility, confirming our bad views of them. Nations whose social memory is organized around certain forms of conflict and defeat will often find that they create the very sort of enemies that they expect and deserve.

Of course, no script, no matter how powerful in our imaginations, can completely reorient the behavior of others. But others are also attempting to understand our behavior according to a set of narratively coded expectations that they already possess. Hence our responses to them can often redirect the ways they are likely to respond to us. The people whom we treat as threatening may not become threatening simply because we expect them to; but our ag­gressiveness toward those people may cause them to act aggressively toward us in return, which then confirms our estimation of their dangerous tendencies.

I noted earlier that a familiar set of American cultural expectations understands opposition and adversity in terms of a "savage war" in which Americans must dominate and conquer their enemy. Such a script followed to its logical conclusion often tends to provoke violent responses from the group that is assigned the role of the "Indians," thus providing the country with the conflict it expects and deserves. Yet this behavior may have disastrous consequences, as in the case of the war in Vietnam: Although the United States responded to the crisis in terms of its traditional script, the Vietcong were not Indians, the American soldiers were not Cowboys, and the result was not the winning of the Old West.

Ironically, narrative framing can make itself true because narrative structure is irrelevant to truth. What is true and what might be true are both expressed, understood, and memorized in narrative form. In addition, narrative structures are both a set of frames for experience and a set of directions for action. Nar­rative structures do double duty in social memory and social convention. The distance from explanatory story to cultural script is not very far, because both are constructed from the same cognitive materials.

Because both history and conventions for social behavior are stored in nar­rative form, there is an inevitable tendency for the two to nourish each other and be confused with each other. The importance of history to the human mind is precisely its tendency to make us want us to reenact it, to follow its lead, to see the path ahead of us in terms of the path that was lately trod. Professional historians deliberately resist this impulse--for they are interested precisely in discovering and showing the strangeness of other times and lands. But this is an acquired tendency that not even the professional historian can fully adhere to. History inspires; it inspires us to reenact it, to see its relevance to our own time. It presses its events and expectations on us like a dancer whose bodily movements entice us to imitate them, like a musician whose playing energizes us to beat time to its rhythms. History's narrative construc­tion draws us subtly and inevitably into a web of imitation and mimicry, a conflation of history and script, memory and expectation. Santayana had it precisely backward, for it is those who learn from history, who absorb the narrative structures of the past, who are most drawn to and destined to repeat them.

Even the student of history who studies it to avoid its mistakes, as Santayana suggested, is drawn into this web. For when we learn from history, we still engage in mimicry. We imagine ourselves at the scene of the battle so that we can remake the fateful decision. We view ourselves in the situation where the mistake was made, and not in a wholly new one. We see a law that binds together the past situation and our present one as of the same general sort. We postulate a cause and effect that occurred in one setting and that will occur again if we do not choose otherwise. In this way we are still repeating history, still framing our expectations about what will happen in terms of the narratives of the past. We are simply trying to tell the story differently from a certain point on. And the scripts of the past still have a hold on us. For they suggest that it was this mistake that we should learn from, and not another--one that might have occurred to us if we did not compare our present situation to that particular one but to yet another not in view. Perhaps we think we will not make the same mistake as Napoleon at Waterloo, but perhaps we have already made a mistake in thinking that Waterloo is the appropriate analogy.

Yet even as our expectations are played out, they are in the process of changing. We do learn from experience; that is the flip side of our ability to memorize and repeat. So our myths and stories mutate partly in response to recalcitrant experience, an experience always mediated by our narratives and frameworks as well as those of the others we interact with. Narrative shapes the way history will proceed, but the procession of history is absorbed into memory and reconfigures our expectations and reconstitutes our governing myths. Americans may always think of themselves as Cowboys battling Indians, but the Vietnam War was not the Wild West. Nor was it even Custer's Last Stand. The war in Vietnam has reconfigured American memory and American myth in important ways; it has become its own cultural icon: Vietnam. Just as Americans remind themselves to remember Pearl Harbor, they now assure themselves that there will be no more Vietnams. Thus when Americans fought the Persian Gulf war against Iraq, they developed the military doctrine of "overwhelming force," so that they would not suffer yet another defeat at the hands of the Vietcong. America has surely learned lessons from Vietnam; what is unclear is whether it has learned the right lessons.

The more pervasive and powerful a form of cultural software in understand­ing the world, the more pervasive and powerful its potential ideological effects. Narrative thought is an excellent example of this phenomenon. Because nar­ratives are so central to our thinking, they create particularly compelling ide­ological mechanisms. Moreover, narratives produce ideological effects not only because they present a partial or misleading picture of the social world but because they are ways of intervening in the social world and of influencing the responses of others. Narrative structures do not simply reflect the world badly, they shape the world to their own distorted lens. They are not only illusion but prophecy. It is as if one could make one's face become ugly by looking at it repeatedly through a funhouse mirror. Indeed, the optical metaphors of dis­tortion are entirely inadequate to describe the variety of ideological effects that narrative thought can have on the social world.


[1] Jerome A. Bruner, Acts of Meaning (Cambridge: Harvard University Press, 1990),


[2] Erving Goffman, Frame Analysis: An Essay on the Organization of Experience (New York: Harper and Row, 1974)

[3] Roger C. Schank and Robert Abelson, Scripts, Plans, Goals, and Understanding (Hillsdale, NJ.: Erlbaum, 1977).

[4] Paul Grice, Studies in the Way of Words (Cambridge: Harvard University Press, 1989), 26-27.

[5] Bruner, Acts of Meaning, 49-50.

[6] Roger C. Schank, The Connoisseur's Guide to the Mind: How We Think, How We Learn, and What It Means to Be Intelligent (New York: Summit, 1991), 187-99.

[7] Roger C. Schank, Tell Me a Story: A New Look at Real and Artificial Memory (New York: Scribner's, 1990), 114-15.

[8] See Jean Matter Mandler, Stories, Scripts, and Scenes: Aspects of Schema Theory (Hillsdale, NJ.: Erlbaum, 1984).

[9] F. C. Bartlett, Remembering: A Study in Experimental and Social Psychology (Cam­bridge: Cambridge University Press, 1932), 63-94.

[10] Schank, Tell Me a Story, 46; Elizabeth Stone, Black Sheep and Kissing Cousins: How Our Family Stories Shape Us (New York: Times Books, 1988), 96-108, 165-95.

[11] Donald Polkinghorne, Narrative Knowing and the Human Sciences (Albany: SUNY Press, 1988), 150.

[12] Bruner, Acts of Meaning, 121.

[13] Ibid., 121-22.

[14] Donald Spence, Narrative Truth and Historical Truth: Meaning and Interpretation in Psychoanalysis (New York: Norton, 1984); Roy Schafer, "Narration in the Psychoan­alytic Dialogue," in On Narrative, W. J. T. Mitchell, ed. (Chicago: University of Chicago Press, 1981).

[15] Mark J. Osiel, "Ever Again: Legal Remembrance of Administrative Massacre," University of Pennsylvania Law Review 144 (1995): 463-704, at 475-77.

[16] Ibid., 476.

[17] Richard Slotkin, Gunfighter Nation:  The Myth of the Frontier in Twentieth-Century America (New York:  Atheneum, 1992).

[18] Ibid., 10-12.

[19] Ibid., 14.

[20] Ibid., 16-18.

[21] Ibid., 10-21.

[22] Ibid., 13.

[23] Indeed, the story of Exodus contains a cycle within a cycle: The Jews are almost within sight of Canaan and are about to receive the Law when they stray from God's teachings and worship the Golden Calf. As a result, God punishes them by condemning them to wander forty years in the desert before permitting them to enter the Promised Land.

[24] As Holmes pointed out, even a dog understands the difference between these two states of affairs. Oliver Wendell Holmes, Jr., The Common Law (1888), Mark DeWolfe Howe ed. (Boston: Little, Brown, 1963), 7.

[25] See Osiel, "Ever Again," 470 n. 23.

[26] See Martha Minow, Making All the Difference: Inclusion, Exclusion, and American Law (Ithaca, N.Y.: Cornell University Press, 1990).

[27] Ibid.