OK - this may not be an entirely direct answer - but possibly an interesting one...

In my mind epistemology is rather simple - it is whatever fits a theory. Where then is ontology? Why that is how the theory fits within itself - the structure or co-causal relationships between the parts of the theory (or model, schema, whatever cognitive construct you prefer).

Now... an ethic might also be considered a cognitive construct because (hypothetically, at least) it serves as a guide for behavior. Thus, epistemologically, within any given ethical system, some behaviors are good or right and other are bad or wrong.

Good enough... but how do we judge the ontological validity of the ethical system itself?

My studies show that the structure of a theory is related to the effectiveness of that theory in practical application. Therefore, an ethic might similarly be understood as being more effective when it has a higher level of structural integrity - greater internal ontological interrelatedness.

In this article (some reviewer left up on the web) I take on that challenge:

http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=3&ved=0CCgQFjAC&url=http%3A%2F%2Fxa.yimg.com%2Fkq%2Fgroups%2F13907307%2F579241524%2Fname%2FSRJ%2B-%2BA%2BRobust%2BSystemization%2Bof%2BGandhian%2BEthics.doc&ei=CPbKTryJFInWiAL8msHHCw&usg=AFQjCNE9ggeHbG8-OBYI8bCUq-23F-eOGg&sig2=COyCF5E-FXmkUvGxItnxCA

Wallis, S. (2010). “Developing Effective Ethics for Effective Behavior.” Social Responsibility Journal Vol. 6 (4).

A novel approach - Let me know what you think!

Thanks,

Steve

Thanks for raising these questions!

How do you solve the obligation issue? 

Speaking from the point of view of a naturalist, I tend to think of game theory as providing an explanation of some of our moral feelings or tendencies to engage in certain behaviors.  Cooperation may be collectively rational but that does not seem to meet the standard for individual rationality.  Game theory might provide an explanation for why we would evolve a tendency to be cooperative but that does not make it rational to be cooperative when one sees clearly that one can take advantage of someone who is weaker.  I think of game theory as providing an explanation for why we have the moral feelings/tendencies that we do.  However, I have doubts concerning whether this provides adequate justification for those feelings.

Why should we see as rational, justified, or just as a good thing to be motivated only by what you deem as advantageous for yourself? Normally we consider being too much self-possessed and anxious as a very unhappy condition. At least we see as a most boring one the inability to take a walk, a meal,  to read a book, to cooperate and to be responsive to other people  -or even to a dog-  without asking if it is to your advantage. Rationality doesn´t require pettiness.

Yeah, I guess I don't think the idea of rationality employed in game theory requires that the participants in the game be motivated by what is advantageous to themselves, or that we hold that it is necessarily rational to will one's own happiness/utility/advantage as an end in itself. It just requires that the participants will some end(s) to be the case and seek to maximize the end(s) that they will rather than end(s) which they do not will. (That much seems rational). Mightn't we construct an interesting game between players who were all very altruistic and selfless and cared nothing for their personal advantage or happiness, but who willed conflicting things? For example, A might will nothing but to make B feel happy, and B wills nothing but to make A feel happy, but the situation which makes A maximally happy is incompatible with the situation that makes B maximally happy, so certain compromises might be rational for A&B collectively that would not be rational individually (relative to the goal of making the other feel happy).

I was not advocating ethical egoism. In fact, the ethical egoist will probably miss out on some of the most important goods in life.  My point was that, in my opinion, game theory will probably not provide a justification for some of our important ethical beliefs, which is not the same as saying that they may not be rationally justified in some other way.

Unfortunately, nobody has directly answered my question thus far.

What I'm asking about is the ontological status of game theoretic norms.  Are they necessary (i.e., do they hold for rational agents in any possible world)?  Do they hold in all cultures?  Are they 'natural'?  And (to bring in epistemology) are they known a priori?

The only reason I mentioned morality is so that people would better understand the point of the question.  (David Gauthier is one advocate of the view that there is some significant connection between game theoretic norms and moral norms.)

Could you rephrase the question using language that is more easily accessible to readers from a variety of different disciplines and sub-fields?

I assume by a priori you mean identifying or manifesting conditions, and not the parts so identified or manifested. Unfortunately, "rationality"  seems to conflate the two by referring to the skill for manipulating parts and the entirely different skill (or power) for identifying which parts can be manipulated. For example, to make a bouquet we must skilfully manipulate the parts - a set of flowers - according to our awareness of the manifesting or identifying condition - a bouquet. A bouquet cannot be identified or constructed from a description of the parts - the set of flowers - alone. So we need to make the distinction between parts, and the identifying condition for parts, and decide which of these the ability or power of "rationality" refers to.

Form identifies content, and it is the manipulation of content, and not the application of form that I take "rationality" to be. This might seem prescriptive, but then "rationality" is applied in game theory, and such theory is generally identified with the manipulation of content and parts. But like the bouquet of flowers, the parts themselves cannot be identified without their manifesting condition. For game theory, it is not enough to be rational in the manipulation of parts, we also need to be cogniscent of the meaning, or aware, of the framework that identifies the parts.

So while we may manipulate, rationally, the parts or elements involved in a game, we need to have in place a condition that shows us where we are going and what parts to select to get there. We don't find that from reading off "parts".

Regarding the ontology of game theory, then,
1) its elements or parts take on the ontology of mathematical objects, and so is physical. Correct me if I am wrong on this point. However, the ontology of the form that identifies the parts is entirely different,

2) the ontology of the identifying condition (such as a moral framework) of the parts of a game is phenomenalistic. It vanishes and appears without the sort of physical redress that would be demanded by the appearance and destruction of the parts themselves. For example, the bouquet appears and vanishes without compensatory redress from its parts, the set of flowers. Likewise, a moral framework vanishes and appears without compensatory redress from the elements it identifies, whether such elements are represented in a game or not.

Any statement that makes substantive truth dependent on analytical complexity, or which ties form to content, whole to parts, identifying condition to parts so identified, would result in, as Wittgenstein put it, a "pseudoproposition"... a book whose words, or syntactic particulars could identify their manifesting condition, in this case ethics, "which really was a book on ethics" would "...with one explosion destroy all the other books in the world".

This might point to a problem with game theoretic procedure. A game may seem to neglect its manifesting or identifying condition - it's whole point: the ontology of these identifying conditions are reduced to the elements they identify, this reductionism arbitrarily forcing through the physicalist ontology of the mathematical/logical abacus (do we want to do that?). The ambiguity of the term "rationality" in these contexts can only add to that neglect.

What do you have in mind?  Could you give me an example of a game-theoretic norm?  (Perhaps the Nash Equilibrium and other solution concepts?)
I am by no means an expert on game theory, so here are just my amateurish impressions.  Please feel free to correct me!

Originally rational choice theory and game theory were treated as a priori disciplines, but that seems to be going out of fashion.  That is because the behavior of actual agents does not seem to conform to these models, e.g., according to the standard model one ought to defect in a one-off PD situation, but this does not predict or explain the tipping behavior of customers engaged in one-time business interaction at a local establishment that they are just passing through.  So, to account for these "anomalies" in actual agents, there has been an empirical turn, that I think has two strands.  One focuses on the cognitive limitations of actual agents, resulting in the growth of behavioral economics, models of bounded rationality, and so on (see Herbert Simon, Kahneman and Tversky, etc.).  The other strand is evolutionary game theory that replaces rational choice with the optimizing role of natural selection, and Nash Equilibrium with evolutionarily stable strategy (maybe Robert Axelrod, Herbert Gintis and Brian Skyrms).  This latter strand need not account for cooperative behavior merely in terms of cognitive shortcomings of actual agents, it seems to me.  (In the first strand, you have experiments, and in the second strand, computer simulations.)

Anyway, what I would question most about standard game theory is its adherence to utility maximization and uniquely rational solutions.  Utility maximization is not the only rational norm.  And there are no uniquely rational solutions.  Reason is not the almighty problem-solver that it is so often taken to be,and there are several different standards of rationality that come into conflict with one another,accounting for all the moral conflicts we find in our normative landscape.

I view ontology as an ancient, antiquated, obsolete and wholly useless line of inquiry, which no longer has relevance to modern philosophy or to any philosophically minded person who understands the transformative power of empirical and now scientific inquiry to which we now add knowledge of how the brain works (the Enlightenment was not far off on this). 

The basic line is of argument is that our capacity to know anything about the outside world requires sensory input which must be transformed from some kind of energy to electro-chemistry and then assembled in the parts of the brain that create thoughts, judgments and understandings if it is to be accessible to self-consciousness.  This assembly process if fraught with potential mistakes.  (Game theory is relevant here, at least metaphorically.)  The process of rooting out the mistakes involves (ala Kant) seeking objective validity in our understandings which is tantamount to doing our best to identify and root out long lines of as yet unrecognized and unthought inferences in discursive fashion.  It is the Socratic model.  Quine's notion of the web or beliefs is reasonable way to think of the object in the Socratic project and what Kant means by the synthetic unity of understanding and apperception.  False beliefs (beliefs assigned an incorrect truth value) contradict or are incompatible with assignment of truth value to some other belief that logically connects them via a possible line of inference.  Contradictions or incompatibilities are intolerable to understanding and to philosophy, but not to imagination or mere thought. 

Nowhere in this process does 'being' or ontology enter as a significant or even credible player.  "Being" is a higher order cognitive construction we loosely and unreflectively slide in under a wide range of empirical beliefs involving objects with cause and effect relationships.  But being wholly reduces to those beliefs and adds no independent or substantive content of its own.  Rather it function to corrupt and to inhibit better empirical (scientific) understandings.  Being on its own explains nothing, tells us nothing. It is the chief scourge in Kant's paralogisms and antinomies.  To quote him:

"From perceptions knowledge of objects can be generated, either by mere play of imagination or by way of experience, and in the process there may, no doubt, arise illusory representations to which the objects do not correspond, the conception being attributable to. . . a delusion . . . and sometimes to an error of judgment. . .  To avoid such deceptive illusion, we have to proceed according to the rule:  Whatever is connected with a perception according to empirical laws, is actual.  (Kant 1787, A366, etc. etc.) 

The rules of judgment and understanding can be thought of as central components in the philosophy or epistemology game.  Playing with ontology and being (ala Aristotle to Scruton) is to play some other kind of game. 

Loving this discussion, just FYI. 

I came across something that reminded me of this forum today when reading about game theory in David Friedman's book Price Theory.   It doesn't have as much to do with ethics as it does to do with "meta-economics"...  but I hope you enjoy!:

Half of the assumption in my definition of economics was rationality; the other
half was that people have objectives. In order to do much with economics, one
must strengthen this part of the assumption somewhat by assuming that people
have reasonably simple objectives; with no idea at all about what people's
objectives are, it is impossible to make any prediction about what people will
do. Any behavior, however peculiar, can be explained by assuming that the
behavior itself was the person's objective. (Why did I stand on my head on the
table while holding a burning $1,000 bill between my toes? I wanted to stand
on my head on the table while holding a burning $1,000 bill between my toes.)

I liked the implication that economics is really a deontological question, not a teleological one.  He puts this statement better by saying:

In the course of this example, I have subtly changed my definition of
rationality. Before, it meant making the right decision about what to do--voting
for the right politician, for example. Now it means making the right decision
about how to decide what to do--collecting information on whom to vote for
only if the information is worth more than the cost of collecting it. For many
purposes, the first definition is sufficient. The second is necessary where an
essential part of the problem is the cost of getting and using information.

Responding to Joseph Wagner, and straying a little from the topic of game theory (though there are some relevant points below I think) I would note that a failure to acknowledge ontology in the scientific endeavour has led to the deaths of thousands and untold misery for many more, and has blotted the rational copy-book of science in a way that will be remembered for centuries to come, I am sure.

I refer to the ontology of objects delineated in the brain sciences and psychology. Here the confusion between the ontologies of physical and phenomenological objects is rife and acceptable. In their literature the physical ontology of external objects cross seamlessly to phenomenological objects such as colours, sounds, and perceptions, supposedly allowing an epistemological exchange between them. The irrational object(s) responsible for this bogus transition is called "the senses".  The senses are not a bona fide object because they have no coherent or unambiguated ontology.

As there is no ontology that facilitates the transition from the ontology of external objects to internal objects then there can be no epistemological exchange between them. This was Kant's point I think. Claims made to the contrary by these sciences are bogus. For example, we are told that mental illnesses are caused by certain brain states. No such claim can be made as we are dealing with objects of different ontologies. Yet thousands have been suckered in, if you like, by such assumptions and medicated on the basis of what is an ontological circularity (e.g. "we know that they are symptoms because there is a physical disorder in the brain, and we know that there is a physical disorder in the brain because there are symptoms).

Ontology is important because by knowing the ontology of an object we can distinguish between types of causality (reciprocal, non-reciprocal); we can distinguish between matters of existence, knowledge, and identification; and, by being able to make these distinctions we can be clearer and less accident-prone about the claims we make in respect of these objects of science. But it isn't just the sciences that are making these real-time mistakes. I might add that "qualia" are a good contemporary example of philosophical drought. Here the ontology of physical objects is cast upon objects, like colours, objects that simply do not share that ontology.