I believe in the moral value of narrative. A story, whether fictional or historical, is a coherent description of a set of events. Its coherence is not simply causal, such that the first event causes the second, which causes the third, etc. Instead, narrative coherence can take many forms, including: unity of character (one agent does a set of things sequentially); unity of community (a set of connected agents do a set of things); teleological unity (a set of events build up to a significant conclusion); or thematic unity (many things with similar meanings are described). Often more than one form of unity applies.

I would like to mention four features of narratives that make them useful for moral reasoning:

1. Narratives enable ?thick descriptions.? In Gilbert Ryle?s famous example, we may either say that someone ?contracted his eyelid? or that he ?winked conspiratorially.? The former is a thin description; the latter, a thick one. Thick descriptions often have moral significance. Contracting an eyelid is neutral, but winking conspiratorially is morally dubious. If it turns out that the contracting eyelid was a signal to commit murder, then that even thicker description marks the act as prima facie immoral.

What justifies a thick description is almost always a story. For example, a video camera would record a wink as a wink, whether it was a signal to commit murder or the result of biting a lemon. We know that it is one thing rather than the other because of what comes before and after it. But we don?t consider every prior and subsequent event, nor do we focus exclusively on actions that cause the wink or are caused by it. Rather, we ?thicken? the description by placing the event within a coherent narrative. This brings me to the second point ?

2. The selection of events in a coherent narrative is moral: Human institutions and actions are always dramatically overdetermined; they arise because of many events that are insufficient but necessary parts of unnecessary but sufficient (INUS) conditions. It is a common ambition of social science to measure as many of these factors as possible in order to assess their relative contribution to the outcome. For instance, we try to predict the decision to vote in terms of factors like the voter?s demographics, the nature of the election, and the voter?s opinions and preferences. Only an unreconstructed positivist would claim that this approach is value-neutral. Social scientists must always omit some contributing factors, and they must always decide how to measure the factors that are included in their models. (For example, demographic background includes race, which is a morally contested category). Nevertheless, social science aspires to neutrality and comprehensiveness. Ideally, every contributing factor goes into the model. If the morally significant factors play no explanatory role, so be it.

In contrast, a historian almost always emphasizes factors of moral significance?especially the intentions of human beings. (So does a novelist, in constructing fictional narratives). Writers of narrative combine causal explanation with moral judgment by making salient those causes that they deem most morally weighty. They are not engaged in retrospective prediction; their goal is much closer to moral interpretation. I think social science is extremely useful, because it allows us to assess causes that may not be deliberate or intentional. But if we want to make judgments and decisions, we need to tell stories.

3. Narratives help to ascribe responsibility for collective actions: Chris Kutz argues that we make the following assumptions: ?I am accountable for a harm only if what I have done made a difference to that harm?s occurrence.? And ?I am accountable for a harm?s occurrence only if I could control its occurrence, by producing or preventing it.? Unfortunately, we may belong to groups that do very serious harms, yet each member of the group can rightly say, ?I made no difference to the outcome, and I couldn?t control what happened.? In these cases?which probably create the bulk of the world?s evils?no one is responsible or accountable for the wrong.

You need not will an end to be responsible for it; you only have to be knowingly part of a group that is moving toward some end. And it doesn?t matter whether the predictable or intended outcome of the group is actually reached: you are accountable if you associate yourself with a group that has a bad telos. Unfortunately, it is often unclear whether a person is an intentional participant in a group. It?s one thing when I voluntarily join a defined and formal body. For example, if I choose to buy stock in a company whose negligence kills people, that is my problem (morally), even if I had no reason to know about the company?s behavior. But there are many harder cases, especially ones involving loose social networks.

When we consider whether someone morally belongs to a group, the form of our reasoning is a narrative. We want to know whether people are intentionally part of a set of coherent actions that lead toward some telos. Novelists are good at showing that sets of characters are linked in morally salient ways; indeed, such linkages often provide the main themes of ?bourgeois? novels. Like novelists, historians tell stories that link people together for teleological reasons. Their methods, which we also use in ordinary life, are the only means we have for ascribing responsibility for group behavior.

4. Stories have themes. A theme is usually a concept or situation that is significant and that repeats throughout the narrative. Determining the theme of a story is a dynamic process. We become gradually aware that a concept or situation is going to be repeated. As we look for themes, we also decide what is literally going on in a text. For instance, in the first scene of King Lear, is Cordelia proud and hurt, or young and very shy, or perplexed by the formal ritual? Our answer does not determine the words she utters, but it decides much else (her tone, body language, location, expression). The only way to determine how she literally behaves is to consider what Lear is about as a whole. Thus Roger Seamon argues that a story?s theme is not some general proposition that we derive (validly or invalidly) from the words on the page. Rather, our emerging sense of a theme helps to tell us what literally happens.

The importance of thematic interpretation has at least two moral implications. First, themes are essential to rhetoric. We deliberate by telling (putatively) factual stories that have themes; therefore we need to know how to tell good thematic stories and how to judge their quality.

Second, it was Hannah Arendt?s view that modern history has no causal coherence. The terrible events of her century could not be retrospectively predicted by measuring the factors that jointly created them. We must understand these events, but their explanation beggars the mind. At best, we are capable of identifying repeating motifs in history. That is why Arendt?s Origins of Totalitarianism is not a causal explanation of Hitler and Stalin, but rather a search for relevant themes in preceding history. It describes ?certain fundamental concepts which run like red threads through the whole.? If we can identify the major themes of our own time, we are doing the best that can be done.