Limits to rationality in genetic and cultural evolution--论文代写范文精选

我们提出一个简单的模型来展示博弈论方法的局限性,并概述进化平衡无法维持的条件。来自生物相互作用的研究似乎支持的证据认为,行为往往是不平衡。似乎也是人类文化特征,具有迅速传播尽管他们对生殖有负面影响。下面的paper代写范文进行叙述。

Abstract 
In studies of both animal and human behaviour, game theory is used as a tool for understanding strategies that appear in interactions between individuals. Game theory focuses on adaptive behaviour, which can be attained only at evolutionary equilibrium. Here we suggest that behaviour appearing during interactions is often outside the scope of such analysis. In many types of interaction, conflicts of interest exist between players, fueling the evolution of manipulative strategies. Such strategies evolve out of equilibrium, commonly appearing as spectacular morphology or behaviour with obscure meaning, to which other players may react in non-adaptive, irrational ways. We present a simple model to show some limitations of the game theory approach, and outline the conditions in which evolutionary equilibria cannot be maintained. Evidence from studies of biological interactions seems to support the view that behaviour is often not at equilibrium. This also appears to be the case for many human cultural traits, which have spread rapidly despite the fact that they have a negative influence on reproduction.

Introduction 
The traditional way to analyze social strategies is through the application of game theory (Maynard Smith, 1982; Fudenberg & Tirole, 1992). The objective of game theory is to find and describe strategic equilibria, often referred to as Nash equilibria. At such equilibria no player can gain anything by using an alternative strategy. The evolutionary justification for this is that natural selection will favour change of strategies until such a solution is reached. Individuals using an equilibrium strategy appear to make adaptive or “rational choices” in order to maximize their reproductive success, selecting the best course of action from the set of possible strategies. It follows that, if evolutionary processes are at equilibrium, predictions about behaviour can be obtained simply by asking what is the most profitable way to behave, without considering the dynamics of the evolutionary process (Parker & Maynard Smith, 1990; Grafen, 1991). However, for evolutionary change to occur at all, strategies must exist out of equilibrium at least some of the time (Maynard Smith, 1978). 

The question remains open as to how persistent and wide-ranging such non-equilibrium conditions are in nature. It is commonly assumed that for much of the time strategies are at a stable equilibrium. New variants that arise are penalized by natural selection and the original situation is restored. Occasionally, a new variant does succeed in invading the population, resulting in a brief, transitional period of change, terminating in a new stable state. If this view is correct, we may be somewhat justified in ignoring the relatively short bursts of evolutionary change, and analyze behaviour purely in terms of optimization theory or game theory. Consequently, it then becomes possible to perceive almost every trait and behaviour as adaptive.

The idea of rationality or perfect adaptation has been criticised on the grounds that many constraints apply to behaviour mechanisms (Simon, 1955, 1956; Maynard Smith, 1978; Gould & Lewontin, 1979; Binmore, 1987; Rubinstein, 1998), that adaptation takes time (Maynard Smith, 1978), and that it is not always compatible with genetic mechanisms (Karlin, 1975) or evolutionary dynamics (Lande, 1981; Dieckmann & Law, 1996; Eshel, 1982). Here we consider a further factor that may prevent an equilibrium being reached: the nature of the interaction itself. When an advantageous trait evolves in one player this can be to the disadvantage of other players, and vice-versa. This scenario may result in endless cycles of adaptation and counteradaption among the different classes of player, with the result that evolution proceeds out of equilibrium for much of the time, with behavioural strategies in an almost continuous state of flux (Parker, 1979, 1983). Under these conditions strategies may emerge and persist that cannot be part of a game theoretical equilibrium. 

The players in such games seek to manipulate one another, and behaviour evolves that appears to be irrational when judged against optimization principles. Given these two very different views of evolution it is important to understand which outcome is most likely, and under what conditions equilibria may be maintained. We explore this question by first presenting a simple game theory model of behaviour that generates equilibrium strategies, but is unrealistic in a number of respects. In particular, the number of possible strategies is severely restricted and individuals in the game are assumed to be omniscient about the environment. When more realistic assumptions are adopted, the equilibrium disappears and exploitative strategies dominate. The results of our review challenge the notion that the analysis of behaviour can be achieved purely by the application of game theory. Indeed, many interactions appear to be outside the scope of game theory. We suggest that in many circumstances evolution is likely to proceed "out-of equilibrium" for much of the time. 

The strategies that emerge in such games are more appropriately viewed as staging posts on the road of an evolutionary race (Dawkins & Krebs, 1978, 1979), rather than as stable end-points predicted by game theory. Although many of our examples come from animals, our conclusions are equally important to the study of human interactions. Because new innovations appear much more quickly in cultural than in genetic evolution, it can be argued that human behaviour is much more susceptible to invasion by manipulative strategies compared with the situation in other species. The evolution of traits "out-ofequilibrium" may therefore also account for much of the richness observed in human culture.

A problem for game theory 
Some problems arising in game theory can be illustrated by a simple game between two players, an actor and a reactor. We call this the Game of Presence. The actor is either present (ν = 1) or absent (ν = 0) but has no choice of actions. The reactor, based on whether or not the actor is present, decides upon an effort x (x > 0). In the presence of the actor the return on this investment to the reactor is first increasing and then decreasing with x. The benefit to the actor of the reactor’s effort is ever-increasing with x. In the absence of the actor, providing an effort x > 0 returns a negative payoff to the reactor. Note that, as in most games, there is a conflict between the two players concerning the amount of effort x to be made by the reactor.

Since we have (deliberately) assumed that the reactor is the only player that has a choice of actions, we solve the game simply by finding the effort that maximizes the reactor’s payoff. If the actor is present the optimal effort is x = 0.5. Actor and reactor then receive in return 0.5 and 0.25 respectively. If the actor is absent, the optimal effort for the reactor is x = 0. This solution is a “Nash equilibrium” (no better response strategy exists ) and it is also evolutionary stable because if the reactor’s effort drifts away from the optima of 0.5 and 0, selection will return it to these levels. This game may seem so trivial as to not warrant a formal analysis. However, it nicely illustrates some of the problems of applying game theory to reality, problems that pervade the entire game theory approach to behaviour. First, in reality, information about the presence or absence of the actor is not automatically provided to the reactor but must be inferred from sensory input (figure 1). The presence or absence of the actor is detected by a mechanism that reacts to the stimulation or physical energy (e.g. light or sound) that reaches the reactor. The actor must be recognized when present even when viewed from different distances and angles, in different light conditions and against different backgrounds.

Second, the actor may take on a variety of appearances, and because such appearances give rise to different stimulation, they may elicit quite different reactions from the reactor. In game theory, it is common to restrict, consciously or unconsciously, the number of strategies considered, e.g. by limiting the number of appearances to the minimum needed to convey the relevant information, or by considering variation along a single dimension only. Furthermore, strategies available to players are prescribed in advance while evolution is an unfolding process in which new strategies become possible as a result of evolution itself. There are several reasons for making these simplifying assumptions. One is to make mathematical analysis possible. Another motive is to limit the number of solutions to the game or to eliminate those considered implausible (van Damme, 1987). However, as we shall see, these restrictions cannot always be justified from an evolutionary point of view.(论文代写)

51Due网站原创范文除特殊说明外一切图文著作权归51Due所有；未经51Due官方授权谢绝任何用途转载或刊发于媒体。如发生侵犯著作权现象，51Due保留一切法律追诉权。(论文代写)
更多论文代写范文欢迎访问我们主页 www.51due.com 当然有论文代写需求可以和我们24小时在线客服 QQ:800020041 联系交流。-X(论文代写)