Published on Webdiary - Founded and Inspired by Margo Kingston (/cms)

Is ignorance bliss?

By Project Syndicate
Created 03/02/2006 - 04:28

Frank KeilFrank Keil is a Professor of Psychology at Yale University. His research asks how adults and children track the causal complexity of the world around them and how they work with only partial understandings of that complexity. There are many Webdiarists interested in science and technology and this article may appeal to them alone, but I was also interested in the metaphor vis a vis our understanding of political economy and human nature. In Keil's words, "we are not nearly modest enough about our ignorance." Hamish.

by Frank Keil

The age of the Renaissance man is long gone. No one thinks it is possible anymore for an individual to grasp, fully, all areas of science and technology. Popular software contains millions of lines of code. Mechanisms of the immune response for just one kind of lymphocyte take up thousands of pages of scholarly journals. An iPod's elegantly simple appearance masks underlying technology that is understood by only a tiny percentage of its users.

But, despite the vast incompleteness of our knowledge, recent research suggests that most people think that they know far more than they actually do. We freely admit to not knowing everything about how a helicopter flies or a printing press prints, but we are not nearly modest enough about our ignorance.

The easiest way to show this is to have people to rate the completeness of their knowledge on a seven-point scale. For any question, a “7” denotes the equivalent of a perfectly detailed mental blueprint, and a “1” implies almost no sense of a particular mechanism at all, just a vague image. People happily, and reliably, assign numbers to their understandings of everything from complex machines to biological systems to natural phenomena such as the tides; but these ratings are usually far higher than their actual knowledge.

We can measure the discrepancy between what we think we know and what we actually know by simply asking people, after they have given their initial ratings, to tell us how some things work in as much detail as they can and then to rate their knowledge again in light of their attempt to explain.

Similarly, we can ask them to answer critical diagnostic questions (for example, “How does a helicopter go from hovering in place to flying forward?”) Or we can simply provide them with a concise but meaty expert explanation. In all of these cases, people somewhat sheepishly confess that their level of understanding was far worse than they originally thought.

People are often surprised and dismayed at their ignorance, but we are not generally bad at estimating how much we know. Instead, we have a special deficit with regard to our explanatory understandings. We are good at estimating how well we know simple facts (such as the capitals of countries), procedures (such as how to make an international phone call), and narratives (such as the plots of well-known movies). But we seem to have a specific “illusion of explanatory depth” – the belief that we possess a more profound causal understanding than we really do. We can be appropriately modest about our knowledge of other things, but not so about our ability to explain the workings of the world.

Several factors converge to create this illusion of knowledge. When Leon Rozenblit and I uncovered the illusion and its specificity, we ran an extensive series of studies exploring why explanatory understanding is so vulnerable to a false sense of knowing. All of the factors that we identified are less influential for facts, procedures, and narratives.

One important factor underlying the illusion of explanatory depth arises from the richly hierarchical nature of most complex systems, which means that they can be understood at several levels of analysis. One can understand how a computer “works” in terms of the high-level functions of the mouse, the hard drive, and the display while not having any understanding of the mechanisms that enable a cursor to move when a mouse is moved, or allow information to be stored and erased, or control pixels on a screen. This hierarchical structure of complex causal systems seduces us into a sense of understanding at a high level, which is then mistaken for having an understanding at a lower level.

A second factor is the false comfort we derive from seeing the parts of a system. The more parts you can see, the more you think you know how those parts actually work. Thus, the illusion is stronger for objects with easy-to-inspect parts than it is for objects with more invisible, inaccessible, or microscopic parts. For example, we may think that we understand the mind much better than we do when we see images of glowing brain regions.

Finally, we often figure out things on the fly when they are in front of us, but then falsely assume that we came to the object with a full understanding in our heads rather than using and manipulating the object to decipher its mechanism.

There may be a silver lining to our inflated sense of understanding. The world is, of course, far too complex for any lone person to fully grasp. If a gnawing sense of ignorance kept us diving deeper in our quest to understand everything we encountered, we might suffocate in the details in one area and miss other areas completely.

The illusion of explanatory depth may stop us at just about the right level of understanding, one that enables us to know how to get more information from others when we really need it without being overwhelmed. It would perhaps be better if we recognised the limits our own explanatory ability, but there may be some adaptive value of those limits as well.

Copyright: Project Syndicate, 2006.

Source URL: