Inside the Narrative System: Communicating Across the Divide

My previous blog post explored a new model of how concepts emerge from people's everyday experiences, based on findings from contemporary neuroscience and artificial intelligence (AI) research. Paul Churchland's book Plato's Camera argues that the concepts used by the Narrative System are really maps of neural activation patterns, taking sensory data and summarizing them at an abstract level. People's sensory experiences are never exactly the same from one moment to the next, but higher-level neurons in the chain of abstraction come to recognize new experiences as similar to an existing map. This post will explore how we might be able to use Churchland's new understanding to communicate better with our fellow humans, especially in situations where our narratives are quite different from one another.

Let's take an example from politics: I may see a political figure as essentially untrustworthy, not protecting the environment, and putting corporations ahead of individuals. A friend may see the same politician as someone who understands ordinary people, promotes local businesses and communities, and fights against government interference in everyday life. First, it's interesting to note that both of our perceptions are based on our experiences (however indirect) of the person in question. Second, it's often the case that we interpret the very same data in completely different ways -- it isn't as though one of us knows the politician well and the other has only heard of him by reputation. To me the potential consequences of the politician's actions are the primary issue, while my friend may see his words as more important, yet we both have the same set of information available for interpretation. Finally, the situation creates clear challenges for me and my friend, who may have a hard time staying civil with one another when talking about this public figure. Churchland's model of the Narrative System helps to explain how this situation can arise, and suggests some potential ways we might be able to resolve it.

Churchland's model explains how people can start from the same set of observations but arrive at very different Narrative conclusions, because what we think of as an "abstract concept" is not fundamentally a set of statements, but instead a sensation of familiarity. The feeling of familiarity occurs when a given set of neurons is activated again and again in similar ways. Thus, the more we are exposed to a particular pattern of thought, the more likely it is to come up again in our future thought. This phenomenon is known in social psychology as Zajonc's "mere exposure effect," and has been used in political communication as well as advertising. Public relations consultants caution against repeating talking points that you disagree with, even when your intention in repeating them is to dispute them, because you may inadvertently reinforce them instead. The reason that we sometimes seem to "speak different languages" from our neighbors is that each person builds up their own set of concepts from experience, and those concepts are based on different habitual activation patterns in our neural networks. When our habits of thinking are different, we can have very different interpretations based on the same set of stimuli, and may selectively notice different pieces of data.

It follows that one important way to develop good mental maps is to cultivate helpful ways of thinking. As this post on mindfulness suggests, some habits of thinking can help us to cope more effectively with the challenges of everyday life including conflicts. Of course paying attention to reliable sources of information will help our minds to construct more accurate views of reality. But we also need to be aware of the potential for bias in our own thinking, because repeatedly thinking about an issue in the same way -- no matter how well-founded -- inhibits our ability to see that issue in any other way. Mindfulness practices again may help, by developing the habit of cognitive flexibility and encouraging us to pause and reconsider before reacting. In our political example, this means questioning whether our own sources of information are good, whether there are more charitable interpretations of our least-favorite politician's behavior, and whether there might be a few of his actions that we could even agree with. A helpful habit of mind is to always question one's own assumptions by deliberately searching for data incompatible with one's conclusions. Another important strategy is to ask open-ended questions, and to be curious about the answers. I shouldn't be satisfied with the belief that I understand someone else's views, because I probably don't.

Once we have explored our own habitual way of thinking as a possible barrier to communication, we can begin to explore someone else's. Language allows us to share pieces of our internal maps, in order to transmit knowledge between people and over time. Even though each person creates their own view of reality (the constructivist view), Churchland still posits that there is an objectively knowable reality shared by everyone. Any two maps of the same territory can ultimately be transformed into one another, as in the figure below. This means that a shared understanding is possible. We just may be representing different parts of the map, or drawing it in different levels of detail. In the politician example, this means looking for places where our cognitive maps agree with one another, and using those points of agreement as touchstones to fill in the rest.

A major current challenge in transforming one map into another is that people trust different sources of information and discount others. But because there is an objective reality that is the target of all our efforts to understand, there is also always a yardstick available by which to judge the success or failure of our concepts. Our cognitive maps are designed to navigate the territory of our shared world, with better maps allowing for more successful navigation. To change someone's map, then, we need to highlight areas where it leads down blind alleys. It's not enough to show how the map and the territory differ from each other; instead, we need to explore places where the map gets people lost, times when they themselves are not happy with the results. In our political example, this might involve asking about the consequences of a particular policy in my friend's life, or times when he wasn't satisfied with the politician's ability to effect changes. Maps are also generally better when they draw clear contrasts (to avoid falling into ambiguous areas) and when they have a key -- an explicit connection to higher-order concepts, which in the political example might be things like "the rule of law" or the belief that "all people are created equal." The best public speakers use these strategies to sway us to their cause.

Churchland's neural-network view of narratives also suggests that they are about patterns and processes, not logical propositions. We do better when we tell stories than when we argue points. In the realm of health promotion, this explains the success of public health telenovelas that show rather than tell people how to cope with health risks. In the realm of political discussion our stories are especially effective when they are personal, when they show the consequences of some public policy in our own lives or in the lives of people we know. Two Minds Theory identifies social imagination as a particular strength of the Narrative System, and Churchland suggests that this is because our conceptual maps are particularly attuned to people and to things in motion. This is another reason to prefer stories over fact sheets, because stories illustrate what happens to people and highlight what has changed for them over time. I can ask about my friend's experiences over time, exploring his own story and what he makes of it; and I can tell my own story and how the politician's actions have affected me personally.

Of course, the greatest strength of the Narrative System is its ability to engage in rational, emotion-free arguments about facts and consequences. And if our goal is to achieve a more accurate view of reality there are well-established procedures to root out error such as those used by scientists or attorneys. But Churchland argues that the exercise of pure logic is a relatively rare occurrence even among otherwise civilized adults: We reserve it for specialized venues like classrooms or courtrooms, where the goal is to ferret out truth at all costs. In most circumstances narratives don't come from logical postulates, they come from experiences. If you can arrange for a formal debate after Thanksgiving dinner, with ground rules and an objective referee, then you can perhaps get people to set aside their prejudices and engage together in an unvarnished search for truth. But in order for this to work the terms of the debate need to be agreed on in advance, and the rules of conduct enforced; even the Narrative system, Churchland argues, is pretty Intuitive most of the time.

Perhaps most importantly for our ability to get along with our neighbors, Churchland's view provides a gradual way for Narratives to improve. This includes both our neighbors' narratives and our own, and suggests that we might each benefit from a sincere attempt to compare notes. Just as maps can show different levels of detail, the judgments of experts differ from those of novices mainly in terms of the nuances that are represented and the assumptions that are avoided. Scientific theories can be understood as maps too, with each new theory providing a broader explanation that still encompasses truths from the prior way of seeing things. Like experts and scientists, we improve our own views by testing them against contradictory hypotheses and honestly looking for ways in which they might be faulty. Even though all currently available theories are likely to be eventually proven false, most have a kernel of truth that can be incorporated into future, more general explanations.

Finally, we should be aware that repeated activation of the same neural network is needed to produce a lasting effect; the Narrative System needs practice in the same way that the Intuitive System does. To sway a friend or neighbor to our way of thinking, we are going to have to connect with his or her pre-existing experiences, or else help him or her to have new ones. Even when addressing the Narrative System, we can't expect to achieve agreement just through well-constructed arguments. And we will probably need to have the same conversation multiple times, in multiple ways. Ongoing, respectful dialogue is the best tool at our disposal to improve both others' narratives and our own.