I. Natural Liberty: People are rewarded for their skills, talents and efforts which are in turn effected by social class and natural "accidents".

II. Liberal Interp: Compensate for or cancel the effects of social class, (scholarships, headstart)

III. Democratic Interp: Compensate also for lack of natural talents (additional instruction in school, e.g. ?). Don't pay higher wages for a talent unless its exercise benefits everyone. (This is Rawls)

If the right institutions have been established and regulated so as to be just, (see p. 87) then fair opportunity will be established by pure procedural justice. For example, if education actually does compensate for inequality of social class and talents, then the actual outcome of procedures for selecting people for the positions this education prepares for need not be monitored. These outcomes are fair "by definition," i.e., because fairness in this respect means "done according to certain fair procedures after the playing field has been leveled by education."

I. Circumstances of justice: moderate scarcity (people want more than they can all get), both identity and conflict of interests. Parties take not interests in other peoples interests (egoism, or mutual disinterestedness). Rawls claims that this is a weak assumption, and that altruism would be a strong (less easily justified) one. Parties are assumed to represent at least two generations of people.

II. Principles decided must be general, (not about specific people or groups), publicly known, final (not revisable).

III. Veil of Ignorance: parties do not know their social status, natural assets, or conception of the good. Each is convinced by the same arguments, must decide unanimously. Each prefers more to less, but no envy or spite. Strict compliance assumed. Parties will be rational in the minimal sense that they will make rational choices of means to give ends; ends not assumed to be given by reason.

IV. The reasoning leading to the 2 principles:

Maximum expected benefit strategy: only works when probabilities can be estimated. Suppose that we know that the probability of being in Situation 2 is .6 and the probability of being in Situation 1 is .4. Consider the following payoff matrix:

Payoff Matrix

Situation 1 Situation 2

Choice 1 10 Units 40 Units

Choice 2 20 Units 30 Units

The expected payoff of Choice 1 is .4 X 10 + .6 X 40 = 28 and the expected payoff of Choice 2 is .4X20 + .6 X 30 = 26. Maximum expected benefit strategy dictates choice 1.

Maximin strategy: Choose so that the worst thing that can happen is a good as possible. In the above payoff matrix maximin would indicate Choice 2, since the worst thing that can happen is 20 Units of benefit.

Maximin better because: can't estimate probabilities in original position, extra benefit not very important, not much known about payoff matrix, etc.

NO GAMBLING ON THE PRINCIPLES OF JUSTICE.

Reasons to prefer 2 principles to utilitarianism:

1. Strains of commitment: utilitarianism can require you to give up stuff to benefit others.

2. Self-respect not realize in case 1.

Concept of liberty: negative and positive: Rawls has both, but see top of p. 203

worth of liberty: important dodge, effect by poverty, ignorance.

Kantian interp of j. as f. : moral principles chosen by a person are expression of his nature as free and equal rational being. To express one's nature is to act on principles that would be chosen when that nature is the deciding element. Assumption that altruism is not part of human nature gives greater choice in the system of ends that someone may choose compatible with those ends. Note that eliminating the assumption of rationality would also increase freedom to choose.