I propose a hypothetical device modeled on Rawls' original position: the scientific original position. In the scientific original position a rational group of scientists chooses a scientific methodology under a veil of ignorance. What the scientists are ignorant of is which potential hypotheses are correct in the world they occupy. They do not know whether Classical Mechanics or Quantum Mechanics is correct, whether String Theory or General Relativity is correct, etc. More than this, they recognize that both their identification of potential theories and their assesment of the likelihood of potential worlds may be inadequate.[1] Under this veil of ignorance the scientists must pick a suitable scientific methodology.
In Rawls' original position a rational group of people chooses a social contract under a veil of ignorance,
"No one knows his placein society, his class of natural asset and abilities, his intelligence and strength, and the like... More than this, I assume the parties [to the social contract] do not know the particular circumstances of their own society. That is, they do not know its economic or political situation, or the level of civilization and culture it has been able to achieve." [2]
In formulating his original position Rawls explicitly makes two assertions that are difficult to reconcile:
2. "The veil of ignorance excludes all knowledge of likelihoods. The parties have no basis for determining the probable nature of their society, or their place in it. Thus they have no basis for probability calculations." [4]
These two assertions are difficult to reconcile because a premise of rational choice theory is that one's preferences are known. The choice being made in the original position is among alternate contracts, but excluding knowledge of the "probable nature" of society removes any basis for preferring one contract over another. This can be made vividly clear by taking to the extreme the ignorance of likelihoods. A rational person is presented two contracts and told to choose between them. However, she is told no more than this. There are two contracts. She must choose between them. Rational choice theory cannot be applied to such an indeterminate situation.[5] Harsanyi makes a very similar critique of Rawls' original position, and points out that a decision maker following only a few compelling rationality postulates cannot help but to act as if he is using subjective probabilities, at least implicitly.[6]
In order to provide a viable version of Rawls' original position from the standpoint of rational choice theory, I will re-interpret assertion (2), regarding likelihoods, in a manner that I believe is consistent at least with Rawls' intent his original position. It is not clear that the "probable nature" of society is at all the same sort of thing as a likelihood. For example, futurists often provide divergent descriptions of future society without making any claims as to the relative likelihoods of each description. I use society here in a broad sense, incorporating not just the mechanisms for how people interact, but also the technologies the society possesses and, even, the natural events that have shaped the society's development[7], and so I choose to use the word civilization instead of society. In order to use rational choice theory one must at least have a list of permissible civilizations (or a means of generating such a list, which may be infinite). If such a list exists it is rational to pick any civilization from that list and consider the original position.
One further consideration in the interpreation of the original position is whether it should be interpreted as an instance of social or individual choice. Either seems reasoanble, but in this essay I choose to interpret it as instance of individual choice. A major advantage of this is that no consideration of Arrow's impossibility theorem needs to be made for an individual choice, as would be necessary for a social choice. Nevertheless, should one choose to interpret the original position as an instance of social choice, Fleurbaey provides some compelling reasons for rejecting the "Independence of Irrelevant Alternatives" assumption of Arrow's impossibility proof.
Following Angner, I interpret Rawls' original position as a decision under uncertainty.[8] I define a set of permissible civilizations, {C}, where each civilization Ci in {C} is composed of N individuals (N is the same for each civilization). Each civilization I define a probability function and utility function over individuals.
Permissible Civilizations
Contracts
Choose a civilization and define (a) a lottery over contracts and (b) a preference relation over those lotteries.
Define admissible decision rules in the obvious manner, yadda yadda yadda
Note: think of these mappings for now as if there is no set of permissible Civilizations. There is only one permissible one.
Social Contract <-> Scientific Hypothesis
Person <-> Reality
Utility <-> Predictive Power
Civilzation <-> Scientific Community
Removing the veil of ignorance <-> Scientific Methodlogy
...consider whether Rawls' original position is a meaningful device; that is, I consider whether the unattainability of Rawls' original position makes it useless as a model for rational behavior. Analogously, I examine the attainability of the scientific original position. I argue that a reasonable consequence of rejecting its attainability is the choice of a particular empirical interpretation of scientific inquiry
The veil of ignorance faced by the scientists is different than that faced by Rawls' social contractors in that it is not at all contrived. The goal of science is to determine what is true about the world but not yet known and thus, unlike Rawls' veil of ignorance, the scientific veil of ignorance does not disappear when the scientists have picked a methodology.