TWO EXTRAORDINARY NONSCIENTISTS

TWO EXTRAORDINARY NONSCIENTISTS
Almost 20 years ago, I briefly knew a man by the name of Craig. Although he died about a year after I met him, I’ve thought about him ever since. Craig had this uncanny ability to converse with a person for a few minutes, and then announce what make and model of car they drove. Neither I, nor anyone I ever spoke to, had ever seen him get it wrong. Craig was never able to explain how he did it, and his unique ability followed him to the grave.
What Craig had perfected was an impressive skill—perhaps even an art—but it was not science. It was not science because it was not a procedure that he could verbally communicate or write down, so that other people in other places or other times could do it also. One of the defining features of scientific activity is that it generates a body of knowledge and techniques that can be communicated and utilized by others in this way.
I also knew a woman named Tula with an equally impressive ability.
Tula was able to predict how well a person’s day would go, based on the shape, size, and color of the aura they emitted in the morning. And in contrast to Craig, she could even explain the specifics of her method. For example, if your aura was round and blue, you would have good luck all day. But if it was square and yellow, then you’d best go back home and stay in bed. Tula had prepared a chart with the complete
relationship between properties of your aura and the upcoming features of your day, so if you had a copy of the chart, you just needed a daily reading of your aura. Although Tula’s success rate wasn’t per17
18 Machina
fect (like Craig’s was), it still compared favorably with standard medical,
meteorological, and macroeconomic predictions, and most of her friends would stop by each morning for a quick reading of their aura, and then go away to consult their chart.
By constructing and distributing her chart, Tula had codified and communicated features of her technique in a way that Craig never could. But since Tula was the only one who could see these auras, what she was doing still was not science. An activity is not science unless it involves techniques that others can also apply as well as variables that others can observe.
The purpose of this chapter is to examine one of the most important
theoretical constructs of modern decision theory—namely, the concept of states of the world or states of nature—from the point of view of these and similar scientific considerations. Are states of nature inherently descriptive or prescriptive objects? Do individuals making choices under uncertainty face these states of nature, or do they create them? Are states external and independently observable, like an indi-vidual’s commodity demand levels, or are they internal and not directly observable, like utility or marginal utility levels? In addressing these questions, I will offer an overview of how researchers have sought to represent the concept of uncertainty, from the original formulation of probabilities and “objective uncertainty” in the seventeenth century, through Leonard Savage’s twentieth century formulation of states of nature and “subjective uncertainty,” to current work which seeks to eliminate—or at least redefine—the distinction between objective and subjective uncertainty. The following section presents some scientific issues common to all theories of choice, whether under certainty or uncertainty. The next two sections sketch out the current theories of choice under objective and subjective uncertainty. After that, I address the question of whether states of nature should be considered descriptive
or prescriptive constructs, and then I consider scientific issues related to the observability and measurement of states of nature. The final section concludes with current work on the relationship between subjective and objective uncertainty.