The Cybernetic Theory of Decision

This book was the result of emigration across disciplinary boundaries.The point of departure was cognitive psychology and its philosophical underpinnings, a realm of meticulous experimental observation and highly generalized interpretative reasoning.The eventual destination was international security policy whose reigning doctrine-the concept of deterrence-was derived from an idealized vision of human mental operations that basic psychology would suggest is regularly violated in everyday life.In the initial stages of this journey I was astonished that security policies of such consequence could be constructed in apparent ignorance, or at any rate cheerful defiance, of what I understood to be some of the more important facts about how the human mind actually works.Since that remarkable instrument provides the foundation of all organized behavior, it seemed to me, its known features should inform any activity as serious as the deployment of nuclear weapons.In the course of writing the book and thereafter I became more accustomed to prevailing security logic and more tolerant of the role it has come to play in the management of inherently dangerous technology.The chief advantage of the logic of deterrence is that so many people are readily willing to believe in it and to accept its asserted requirements.If the underlying suppositions of the deterrence doctrine are difficult to reconcile with major characteristics of human behavior, they do nonetheless reflect how most people prefer to understand the business of determining threat and devising protection.Organizing ideas that can command widespread acceptance are an essential element of any coherent policy Nuclear weapons and the broader issues of international security definitely require reliably coherent policies.Despite that practical accommodation, I have sustained my original sense of concern.The fundamental conceptual distinctions presented in this book identify general aspects of organized behavior likely to affect the formulation and implementation of virtually any public policy.They are especially important in understanding the interaction of military organizations and the much broader set of activities that determine the state of global security.They illuminate problems not readily visualized in standard assessments of these matters.Three decades later the distinctions remain relevant and, I regret to say, insufficiently appreciated.The core distinction has to do with the difference between adaptive processes and direct outcome calculations derived according to what is herein termed analytic logic.Blind adaptation through the process of natural selection is currently understood to be the basic method of evolution.It is capable of producing exquisitely refined and highly successful behavioral strategies that in other species are assumed not to have been intentionally designed.In contrast, intentional design-and indeed rational design-is considered to be the principal characteristic of organized human behavior.International security, it is supposed, is conducted through conscious calculation.Misjudgments are admitted, as are unintended consequences, but explicitly formulated and systematically pursued intentions are assumed to be the driving force of human interactions-as distinct from the interactions of nature.The argument of the book suggests that adaptive logic affects human behavior more significantly than is generally admitted.The central claim is that awareness of the distinction between adaptive and analytic logic and practical skill in using it will improve comprehension of most situations.Understanding of adaptive logic and of the remarkable characteristics of recursive processes has advanced substantially in a variety of disciplines since this book was originally published.The central distinction in question can now be pursued through extensive literatures on subjects such as complexity, chaos theory, catastrophe theory, and adaptive agent dynamics.The latter body of work, enabled by dramatically improved access to advanced computational capability, is especially relevant.Adaptive agent modeling has demonstrated that coherent outcomes, suggestive of human experience, can emerge from adaptive interactions without any attributed intention or analytic outcome calculation. 1 In one instance it has been demonstrated that such a process can track reasonably closely the actual behavior of a historical human society over a thousand-year period. 2 This body of work has not dislodged the habit of attributing intentions and is not likely to do so any time soon.Nonetheless it has strongly reinforced the