Friday, December 16, 2011

Some Thoughts on Game Theory and Literature

Game theory is one approach to helping one uncover the rules of complex social systems. It shows that complex social systems can be understood as games, meaning they have rules. In postulating that games have rules, game theory goes against certain postmodernist-anarchist views that insist on opposing the very concept of rules. They see rules as limiting, as preventing freedom. They do not understand that it is the very presence of rules that give us “degrees of freedom.”

Nietzsche points out in Beyond Good and Evil that rules are absolutely necessary for every form of morality and art form has used and needed rules. “What is essential and inestimable in every morality is that it constitutes a long compulsion: to understand Stoicism or Port-Royal or Puritanism, one should recall the compulsion under which every language so far has achieved strength and freedom—the metrical compulsion of rhyme and rhythm” (188). He then goes so far as to say that “all there is or has been on earth of freedom, subtlety, boldness, dance, and masterly sureness, whether in thought itself or in government, or in rhetoric and persuasion, in the arts just as in ethics” developed only because of rules – and that the use of rules lies in nature itself, that rules are natural. It is through living by rules that we make it “worth while to live on earth; for example, virtue, art, music, dance, reason, spirituality.”

Nietzsche rejects living without rules. But which rules? The tacit question asked by game theory is: “what rules make for the best games?” But it also asks: “what rules would evolve to ensure survival of the game?” – whether that game is a species or a ritual, an economic system or a work of literature. Further, “Game theory shows how people make decisions about what to purchase and when and the rationale for seeing goals or rewards” (Richmond, et al, Science 11 July 2003, 179). That is, “Our sense of which behavior to choose to reach a goal or obtain a reward is based on the perceived value of the reward, the effort needed to obtain it, and our previous experience about the likelihood of success” (179). Which raises the questions of what is the “goal” of a work of art, and what “reward” that work of art gives us, since behavior’s existence suggests there is a goal and/or reward to be achieved/received that must have been important enough for us to have been pursuing it from prehistory to the present day. We will not act if we do not perceive that the reward we will receive is sufficient.

Action is always directed toward the future; it is essentially and necessarily always a planning and acting for a better future. Its aim is always to render future conditions more satisfactory than they would be without the interference of action. The uneasiness that impels a man to act is caused by a dissatisfaction with expected future conditions as they would probably develop if nothing were done to alter them. In any case action can influence only the future, never the present moment that with every infinitesimal faction of a second sinks down into the past. Man becomes conscious of time when he plans to convert a less satisfactory present state into a more satisfactory future state. (Mises, Human Action, p. 100)

We would not create works of art of literature or participate in viewing/reading/listening to art/literature/music if it did not reward us. That is why l’art pour l’art is neither achievable nor desirable. But each of the questions raised by game theory are really the same question. Formulating it the first way makes it clearer regarding how it can be applied to art and literature. It helps us to see the critic as the uncoverer of the rules the artist used (consciously or not) to create their work of art or literature. Formulating it the second way helps us understand how game theory can help us understand the source of rules, from the laws of physics to the rules of grammar. It shows that more rules are needed for more complex games. Only a few are needed at the quantum level, but with each movement up in complexity, more rules emerge – and are needed – until one gets to complex human social systems, which need thousands, if not millions, of rules. And it shows how necessary rules are if one is going to have any sort of game at all. It is the existence of rules that give us freedom – making us more creative, often far more creative than we are otherwise. Many good rules (note the word “good” here – it is not the number of rules so much as the kind, those that generate more moves, not less) give us many more degrees of freedom. Chess is a better, more complex game, with many more degrees of freedom, than checkers, though both are played on the same board. It is better because more complex. Complexity gives us more freedom. Rules are necessary, but the more complex the system, the more and more complex the rules that are necessary.

This brings us to the question of what distinguishes Rules from Laws (this is similar to Hayek's distinction between law and legislation). Both Rules and Laws are used to delineate what one does. However, Rules are flexible, which means they can be bent; they are prescriptive, which means they say what you can do (as, say, the rules of chess) and, as such, are positive in nature; they act as strange attractors, meaning they are dynamic, they deepen and grow more complex over time, and they increase your degrees of freedom, giving you more possibilities. Action is impossible without rules; rules create actions, possibilities of and for actions. Laws, on the other hand, are inflexible and cannot be bent, but only broken; and they are broken under threat of punishment (laws can be changed – but in the sense that they are changed, they no longer exist as laws and other laws now exist); thus, they are restrictive, saying what you cannot do; they are static, unchanging (especially in philosophy), they decrease your freedom by being restrictive, and give you fewer possibilities. Action is cut off with Laws; laws prevent actions, possibilities of and for actions.

In Individualism and Economic Order, F. A. Hayek points out the dangers found in the radically individualistic view of human nature – showing that it can and usually does lead to the collectivist view (too fine a texture looks like a solid color). Hayek shows that taking the exclusively individualistic view of human nature (vs. the social-individualistic view of human nature) leads to bad games (social systems, economic systems, government), since no information can be shared among players. A good game-system is one where communication – and, thus, community – is possible.

Hayek suggests that there are two kinds of individualism, one based on rational philosophy, which started with Descartes and was further developed by Rousseau, Voltaire, Kant, Hegel, Schopenhauer, and by the existentialists, including Heidegger, Sartre, and de Beauvoir, and which I will call Cartesian Individualism, and the other based on the Scottish philosophical tradition of David Hume, Bernard Mandeville, Josiah Tucker, Adam Ferguson, Adam Smith, and John Locke, and further developed by Edmund Burke, Alexis de Tocqueville and Lord Acton (which is the social-individualistic agonal view). Cartesian Individualism is based on the idea that man is rational and can thus rationally construct society, while the Scottish tradition does not see man as being fully rational, but also, perhaps primarily, influenced by his drives and wants and needs of the moment. These quite different views give rise to quite different forms of individualism. Perhaps the best way of explaining the differences would be to put the two traditions of individualism side by side in a table showing what Hayek sees as the difference between the two traditions, and the consequences of each of these traditions:

Scottish (Social-Individualistic) Individualism
the individual is found within the social, leading to free markets

man is not always rational, or even capable of always being rational – man also has impulses and instincts

since man is not rational, he cannot design or plan something like a society or economy

the individual participates in the social (cooperates) through being selfish

“If left free, men will often achieve more than individual human reason could design or foresee” (11).

It is not necessary to find good men to run the society, meaning anyone can play

it is not necessary for us to become better than we already are, making it easy to enter the game to play it

freedom is granted to all

no one group never always wins, which keeps people playing

reason is seen “as an interpersonal process in which anyone’s contribution is tested and corrected by others” (15)

inherently unequal people are treated equally

inherent inequality allows diversity

hierarchical – intermediates encouraged

Cartesian (Radical Individualistic) Individualism
radical individualism, leading (ironically (?)) to socialism

man is rational and has no instincts and can always control his impulses

since man is rational, he can create through planning the ideal society or economy

individual vs. the social – i.e., selfishness vs. cooperation – therefore need coercion

“social processes can be made to serve human ends only if they are subjected to the control of individual human reason” (10)

only the best can or should run society and make economic decisions – few can play

men need to be improved (presumably made more rational) before a good economy or society can be created – hard to play

freedom granted only to the good and wise

the “good and wise,” “rational” rulers always win – no reason to play the game

reason found in the individual, especially in certain “good and wise” individuals

people are made equal in actuality – thus, have to arbitrarily assign tasks

only State and Individual, thus flattening society – intermediates suppressed

We can see in this comparison that the Scottish form of individualism, by being simultaneously social, provides us with a much broader, more inclusive set of game rules. Anybody can be involved in the social and economic games – making these systems more complex, containing as they do more constituent parts acting in coordination and cooperation. Man does not have to be “improved” for the kinds of systems that would be set up using Scottish principles as he does using Cartesian principles (historical examples of attempts to “improve” man to make him more suitable for “rationally” designed societies include the Holocaust in Nazi Germany, the Terror of Revolutionary France, and the slaughters of millions in the Marxist states of the Soviet Union, Maoist China, and Cambodia, just to name a few). In the Cartesian view, there is only one rationality, but in the Scottish view, there are many rationalities, which can often come into conflict.

Scottish philosophy gives us far more complex social game rules than does the (radical individualist) rationalist philosophical tradition. One may think this rationalist approach would allow a given individual’s influence to extend throughout a society and create a more interesting game, but what it actually does is flatten out society, making it less complex, less interactive. A radically individualistic world view leads, ironically, to a collectivist outcome. “All unity is unity only as organization and co-operation: no differently than a human community is a unity – as opposed to an atomistic anarchy; it is a pattern of domination that signifies a unity but is not a unity” (Nietzsche, WP 561).

If we take the Scottish view that a person’s knowledge and interests are limited, making our actions limited to a tiny sphere of influence – our family and friends, our churches and schools and businesses, the intermediate social groups the rationalists suppress and the Scottish encourage – we see a highly complex society emerging, with the individual influencing the small social groups, the small social groups influencing the individual, and both interacting to influence larger social groups, which themselves feed back to the smaller groups. We have a series of nested hierarchies where each person acts in a social-individualistic way through the communication of information to other individuals to create smaller cultural subsystems of the larger culture. The same individual can have an effect on a school, a church, a business, and a local government, each of which will have larger effects on the society at large. More people have more influence over society. And man does not have to be “improved” because the worst among us can be canceled out by the best. These principles, upon which the free market is based, are “an effective way of making man take part in a process more complex and extended than he could comprehend” (Hayek, 14-5). One does not have to have perfect knowledge to participate. One can participate while having a considerable amount of uncertainty, and still do well. Which is good, since no person is omniscient. We can reduce uncertainty through education, increasing our own individual knowledge, but we will still be left with a plethora of things which we will never have the time to learn.

There needs to be a way for individuals, with their limited information, knowledge, etc., to enter into a highly complex game, to be able to participate in the game itself. The way to allow someone into a highly complex game is by simply not having barriers to their entering and playing the game in the first place. And, if you do choose to play, and to take large risks while playing, you should be able to reap a correspondingly larger reward. To have a good game,

any workable individualist order must be so framed not only so that the relative remunerations the individual can expect from the different uses of his abilities and resources correspond to the relative utility of the result of his efforts to others but also that these remunerations correspond to the objective results of his efforts rather than to their subjective merits. (Hayek, 21)

And the game must not be constructed of iron-clad laws/legislation, but of more flexible rules. These are also good guidelines for creating works of art and literature, and for writing works of philosophy, theory, and criticism.

An example of good game rules are our “traditions and conventions . . . [which] evolve in a free society and . . . , without being enforceable, establish flexible but normally deserved rules that make the behavior of other people predictable in a high degree” (Hayek, 23). Most social rules should be those agreed upon and practiced by most of the people most of the time, enforced by subtle social pressures, not the use and threat of physical force. “In the social sciences the things are what people think they are. Money is money, a word is a word, a cosmetic is a cosmetic, if and because somebody thinks they are” (Hayek 60). They are rules because we agree they are – they are socially constructed. With these kinds of rules, those we find in the free market, we have various choices – while with orders or iron-clad laws, we get no real choices. This is what Nietzsche is getting at in his “On Truth and Lies in a Non-Moral Sense” when he says words are metaphors we have forgotten are metaphors, not Truth (words are not congruent with things – they are not attached to things through iron clad laws). Any choice is better than none. “It is better to have a choice between several unpleasant alternatives than being coerced into one” (Hayek 24).
Of course, just because the world has a socially constructed element, it does not follow that all the world is socially constructed. To claim it is brings us to the problems with pragmatism, where no system at all can be constructed. Hayek says pragmatism is “the preference for proceeding from particular instance to particular instance,” where the rule-maker “decides each question “on its merits””(1). With pragmatism, expediency and compromise lead us “to a system in which order is created by direct commands” (1). “Without principles we drift,” and we are led “to a state of affairs which nobody wanted” (2). Pragmatism makes it possible to change the rules with each move in the game – one could imagine some game master watching a game being played between two people, and changing the rules whenever he wished. This would lead to the game players in each move trying to gain the game master’s favor. They would end up trying to bribe the game master rather than paying attention to playing the game at hand. If this sounds like how too much business is conducted, with the government as the game master, we can see why. How much money do businesses waste trying to influence “pragmatic” government officials? With the use of basic principles, everyone is clear what the rules are and that they cannot – or, at the very least, are very difficult to – change. The game players concentrate on the playing of the game itself rather than coming up with strategies to influence some game master. With the use of general principles, the game master can all but be done away with.

There are a set of “basic principles” that are not socially constructed, a reality that exists even if we are not around to observe it which we have to deal with (though our attitude toward it, meaning our perspectives on it, are certainly socially constructed and thus inherited and modified based upon that inheritance). This is physis. On this world we have increasingly superimposed, with the introduction of such technologies as (especially fiat) money and writing, a socially constructed reality. This is nomos. This social reality, these social facts, “are accessible to us only because we can understand what other people tell us and can be understood only by interpreting other peoples’ intentions and plans. They are not physical facts, but the elements from which we reproduce them are always familiar categories of our own mind” (Hayek, 75). We have this socially constructed reality because “we all constantly act on the assumption that we can . . . interpret other people’s actions on the analogy of our own mind and that in the great majority of instances this procedure works. The trouble is that we can never be sure” (64). Which is what makes it all a game in the first place. But if we want this socially-constructed reality to work best, we need to allow it to structure itself as the rest of the world is structured – as a complex, dynamic emergent system.

Hayek has given us strong evidence against taking a collectivist-exclusive (unity-only) or an individualist-exclusive (pluralist-only) view. Hayek gives an alternative in his argument for a combination of individual and social – and even of a naturalistic and a socially constructed reality – that create a hierarchy of social interactions. What he argues for is a social-economic system that is in fact a system – a dissipative-structure system scalarly similar to every other system found in the universe, with the principles/game rules as the strange attractors of that system. It is a social system that reflects Francis Hutcheson’s definition of beauty unity in variety and variety in unity – which should not be surprising.

No comments:

Post a Comment