Nozick (henceforth referred to as N.) is concerned with the functions of principles, in other words, with what they do.
They have intellectual, personal, interpersonal, and intrapersonal functions, as well as teleological, epistemological, and ontological functions.
Intellectual
Somehow, principles group things such as actions. Those things include things in the past, the present, and the future, as well as hypothetical things.
In most jurisprudence, a judge is required to find a principle which yields her decision. If there is no principle that yields the decision, that indicates that the decision is mistaken or needs further study. Any true particular judgement must result from a true general principle.
A PRINCIPLE IS A DEVICE THAT TRANSMITS PROBABILITY OR SUPPORT which flows from known data, via the principle, to judgements/predictions about new or somehow unknown data.
In ethics, what enables us to make judgements on principle is that our principles have lawlike characteristics. Part of being lawlike is that they do not contain particulars. Even a full examination of every single actual case might not be sufficient to support a principle: a principle also would cover hypotheticals and be supported by indirect evidence.
Interpersonal intellectual p.6
Principles justify to others by 1) attractiveness of the principle itself, and 2) taking into account accepted cases.
Precedent
Judges are constrained by precedent, because there is great utility to having legal decisions that are consistent with each other, and because precedent constrains the personal predilections of the judge.
Perhaps principles constrain individuals in the same way. Principles can protect individuals from their own individuality.
Filtering
In deciding this particular case, we want to consider all and only relevant other cases. A principle can filter out which cases are relevant and can cause us to consider not-yet-noticed relevant cases. We act as if we are sure that all relevant cases will conform to a general principle (principles have to be general). So principles can protect people from deciding two similar cases differently (why is that wrong? because one property of formal justice that is generally agreed on is that it will decide similar cases similarly).
In addition, of the many options open at any given time, given our limited capacities, principles are useful in that they filter out options that violate principles and rank other options (from p14).
Interpersonal binding p. 9
Principles help assure other people that we will overcome temptations (and that is also an intrapersonal function: they help us overcome temptations within ourselves). Being reliable is valuable, because it enables us and others to take risks (knowing that you will show up for class leads to my willingness to prepare for class, which leads to your willingness to invest your resources in the class in other ways, etc.). Perhaps our emotional attachment performs that function better within our family, but with people we do not know well, principles are what enables us to rely on others' behavior. So PRINCIPLES BIND US.
ANNOUNCING PRINCIPLES incurs reputation effects, but we also want people to actually believe their principles. Even if there are no correct principles at all, it would still be useful to hold principles because it makes us reliable not just for others, but for ourselves, and ACTUALLY BELIEVING in principles makes us more reliable. Perhaps evolution favors those who have and believe in principles.
Resolution p. 11
Principles can be more or less general. Killing is wrong is quite general. Killing is wrong except in self defense is less general. We can continue to fine tune the principle. But at some point, the fine tuning makes things so complex that others cannot see the principle behind actions, and so perhaps paradoxically, there is more utility in having coarse principles than in having fine-tuned principles (think of politicians: simple messages are easier to convey than complex ones). THUS PRINCIPLES MAY HAVE TO BEND to fulfill their functions, and those functions may conflict (intrapersonally, complex principles may work, but interpersonally, they may not).
Identity
Principles serve to integrate our actions over time and to keep us "true to ourselves." It would be consistent to have no principles at all, so consistency is not the only value of principles. They also serve to make one coherent, to integrate our selves into a whole. Even if you define your self not through principles, but through goals, principles are still useful constraints upon your self to keep it on track towards goals. An important reason why we are willing to start long-term projects is that we have some reason to think that we will continue them: by enabling us to rely on our future behavior, principles enable us to take on long-term projects.
Overcoming temptation p. 14
This is an explanation of "acrasia," incontinence.
We "discount" rewards that are in the future. That is a fact about how humans operate. Perhaps discounting future rewards is evolutionarily speaking selected for: it is a way to build in a rough approximation of statistical/probabilistic calculations.
Refer to the graph on p. 16.
From which standpoint should we judge the two actions? If we are in time interval B, we will choose the lesser reward. But if we are in A or C, we will choose the greater. N. Suggests that the reason why we should not trust the view from interval B is that A + C is a greater interval.
So how do we get through B without taking the smaller reward (i.e. without giving in to temptation?). Among several options, one is to adopt a principle. A principle "TIES THE DECISION ABOUT WHETHER TO DO AN IMMEDIATE PARTICULAR ACT . . . TO THE WHOLE CLASS OF ACTIONS OF WHICH THE PRINCIPLE MAKES IT A PART." (p. 17)
N. thinks that at any given time, I am in interval B with respect to one or a few things, but I am in interval A or C with respect to very many things. Thus it is attractive to me to formulate a principle not to give in to temptation. It will tie me to a class or classes of actions. The mechanism works because PRINCIPLES CHANGE THE UTILITY OF AN ACTION. If doing an action violate a principle, then doing the action is not an isolated event. It also involves a whole class of actions. We can violate principles, but that reduces our reliability and puts our self into question. Hence there is great disutility to violating principles. Because any violation of any principle violates the principle that we not violate principles, there is even more disutility to violating principles ever.
But there is also disutility to creating principles which we are sorely tempted to violate, since violating a principle has such great repercussions. Every principle must have an "all" or "every" built into it in order to be a principle, but it can be carefully chosen: "Every week skip most desserts" NOT "skip all desserts."
Jumping ahead to p. 25 for a moment, there is another way to get over temptation that involves adopting principles. That way is to tell people who are tempted to consider what they would recommend to others in their situation. By removing themselves from the particular situation, they are able to see that the same principle of choice should apply to them as to others.
Sunk Costs p. 21
There is a principle in economics that says that in considering what your next step should be, you should never consider "sunk costs." "Sunk costs" are the monies you have spent that you will never recover (e.g. the half of the value of a new car that disappears when you drive it off the lot). It is a strongly held principle in economics and also in decision theory. N. however rejects it in the realm of personal principles. Our effort to build up relationships or to start projects are not of no account to us and should not be. They help to define ourselves and hold our lives together as wholes (i.e. they make your life one sausage instead of a bunch of separate links). Keeping our principles in the face of temptation is in a way honoring sunk costs, for our principles represent our judgements about groupings of past actions we have done. They put some value on continuing those actions, and those actions are "sunk costs" (we already have their rewards and we will get no further reward from them).
The irrationality of violating a principle can be countered by the "irrationality" of putting a great deal of value on the sunk costs of holding a principle. Even if honoring sunk costs is irrational, N suggests that it is useful, and hence that rationally we should not stop honoring them, if by honoring them we increase future rewards. We stick to our principles because not doing so makes it unlikely that we will stick to other principles, and that makes future rewards less reliable: hence the sunk costs of holding principles should be honored.
But what if we have no irrationality in us to counter? N. claims that in that case, we still might want to value sunk costs of principles, because doing so retains the interpersonal utility of those principles.
Drawing the line p. 25
There are gray areas, and it would be valuable to have a device which could "draw the line" at which one has passed beyond what is acceptable. Principles can do that and hence are useful. Someone who has been sorely tried might find principles much more useful than someone who has had it easy all his or her life.
Symbolic Utility
Adopting a principle makes each act stand for or mean every other act that falls under that principle. It does so because a principle must involve a generalization. N. asks whether that exhausts the ways in which a particular act is tied to other acts by a principle and says, "No." N. thinks that "symbolic utility" is a further factor.
In decision theory, the fact that an action produces a certain result with certainty makes it the case that that act has the same utility as its result. So tuning up my car has the utility of the resulting long reliable life of my car.
N. wants to add "symbolic utility." To prove that such a thing exists, N. produces an example. Minimum wage laws do not help the poor, but they have "symbolic utility" in that they symbolize helping the poor. People value minimum wage laws in the face of strong evidence that they in fact hurt the poor.
Another way to look at "symbolic utility" is as "expressiveness." Minimum wage laws express a concern about the poor, a desire to help, etc. Whether it is "utility" or "expressiveness," Nozick thinks that just as tuning up my car has all the utility of a reliable long lasting car because utility flows from the effect to the cause, so actions that have a symbolism involve a utility that flows from what they symbolize back to the acts themselves. Instances include actions that express independence from parents (breaking curfew, e.g.), actions that express being free of guilt (washing hands, e.g.).
N. acknowledges that we are apt to think that "symbolic utility" is irrational, but he points out that holding ethical principles has a symbolic importance for us. If we "respect others" as a principle, then we are "on the side of" respect for others. Things that holding ethical principles might symbolize, suggests N. include: being a rational creature that makes its own laws, being a lawmaking member of a kingdom of ends, being an equal source and recognizer of worth and personality, being a rational disinterested unselfish person, being caring, living in accord with nature, etc. BEING ETHICAL IS A WAY OF SYMBOLIZING WHAT WE MOST VALUE.
Anthropology holds that we are symbolic creatures, and evolution made us so. It must have some adaptive function, but it need not stay mired in that function: it can transcend it. Believing in symbolic utility may enable us to symbolically at least achieve things that are otherwise causally and conceptually impossible. Symbolic utility can be misused or lead to bad things (drug prohibition, compulsive handwashing), but it can also be useful (if harming someone for revenge is bad, then there might be some value to having a mechanism whereby we can do an action that symbolizes revenge).
Symbolism is created by individuals and also given to individuals by cultures or other individuals. Whether we think symbolic utility is useful and should be kept or not, it seems to be a fact of human decisions, and so decision theory should take it into account. There are two ways to see it: as symbolism that comes from individuals and is transmitted to cultures, and as symbolism that flows from cultures to individuals.
Symbolic utility does not obey the same regular laws as causal utility does. In other words, the fact that if i do X, Y is 90% probable might make doing X attractive to me because the utility of Y is transmitted to X, and presumably it is transmitted with some discount that is a function of 50%. But if X symbolizes Y, and Y is 90% probable, it might be the case that X gains no attractiveness. The difference in symbolic utility between something that is .9 and 1 is much greater than the difference between something that is .8 and something that is.9, because 1 represents certainty, and that itself has symbolic utility.
Teleological devices p. 35
Teleology is concerned with goals. So to explain why a plant does a particular action, you might refer to its goal (sunlight, nourishment, etc.).
Epistemology is concerned with knowledge, belief, and the senses. So epistemological explanations explain how we know things, what is knowable, etc.
Ontology concerns being. So ontology tells us what is. For example, we might think that everything is material, or we might think that thoughts are not material but exist. So our ontology could consist of just matter, or maybe thoughts plus matter, or maybe just thoughts.
Deontology concerns what ought to be. Hence it concerns morals.
N. says that principles can be epistemological devices in that they tell us what a thing is by grouping it under a general heading: they transmit support or probability from a group of cases or a type of case to a particular case. They can also be deontological devices, because they tell us what we ought to do; they might, for example, help overcome temptation by transmitting utility from one action to another. They can be teleological in that they help us to achieve our goals for that same reason.
Principles are fallible. A way to refute a principle is to look at its results. Thus principles are tested empirically. A principle that seemed fine before Hitler's Germany took it to an extreme (obey superiors) is refuted by its results.
Decision theory enables us to see principles as devices, and to compare them with other options for principles as well as options that are not principles at all. Principles are devices for achieving certain effects. We can judge them by their effects.
Perhaps truly good moral principles would require actions that we could not comprehend, and so we are better off not using them as principles. Physics has identified laws that are beyond ordinary comprehension and inapplicable in the ordinary physical situations we find ourselves in, and so we are better off adopting principles that are not true!
Rationality and principles p. 40
Kant held that conforming to principles just is our rationality. Thus there is not need to look for the functions of principles in order to see why we should adopt them. But why should we value rationality so much? Is there something about it that is more than just an instrument?