Fundamental assumptions of game-theoretic solution concepts

点击翻译按钮获取中文摘要

Both decision- and game-theoretical assumptions of rational choice theory have come under criticism in recent years, mainly due to experimental research. At the same time there is an apparent convergence of new sociological action theories, social-psychological dual-process theories and economic theories of bounded rationality, which gives birth to a new synthesis in interdisciplinary action theory. This synthesis could serve as an alternative to the standard theory, provided the theory matures in terms of its degree of formalization. The core idea of this alternative is to conceptualize rational behavior and in particular Zweckrationalität as but one decision procedure among others, and to differentiate it from more automatic forms of decision-making. The new synthesis provides a variety of hypotheses concerning the conditions, under which instinctive rather than contemplative, automatic-spontaneous rather than reflecting-calculating or system-1- rather than system-2-behavior comes about. This alternative theory exists already as rudimentary decision theory; its extension to a theory of strategic interaction, i.e., the answer to the question of how a new game theory, which is based upon these principles, might look like, still misses.In this project two core ideas of this new synthesis will be examined in the context of strategic interaction. First of all, the centerpiece of the synthesis, i.e., the body of hypotheses referring to the situational conditions under which either automatic-spontaneous or reflecting-calculating decision making occurs, shall be tested experimentally. To this end, we will recur on the most explicit version of these hypotheses, which was provided by Esser and Kroneberg as part of the new sociological action theory. A second series of experiments will be directed at an empirical test of another core idea of the new synthesis, which has not yet been included in new sociological action theories. This is the notion that system-2-like, reflective thinking shares some kind of "energy pool" with other, higher cognitive functions. Specifically, we will test experimentally whether and to what extent the performance in strategic interaction is affected by classic social-psychological techniques of manipulating this pool (e.g.: Memorizing sequences while making decisions; performing an activity that demands self-control; etc.). The sketched research project contributes to the interdisciplinary understanding of human decision-making, since these two essential principles of the new synthesis have never been tested systematically in the context of strategic interaction.