The View from Within: Normativity and the Limits of Self-Criticism

Chapter 1. Setting the Stage: The Problem with Rationality

CHAPTER ONE Setting the Stage The Problem with Rationality Two Concepts of Rational Action Theories of rational action fall under two main headings. The first pronounces an action rational if, given the circumstances, it is perceived to be the right or proper thing to do. To act rationally is to act in ways that, given the actor’s needs, desires, and, situation, meet expert approval. On this view, what is scrutinized in assessments of rationality are the merits of the actor’s actual performance rather than the quality of the data at her disposal, her level of expertise, or the quality of her reasoning. It is the end result that counts. To be considered rational in this sense, it is enough for an act to be deemed effective or worthwhile.What is appraised is performance ,not agency.Action taken or withheld thoughtlessly,instinctively, or unwittingly may also be considered rational as long as it is deemed appropriate . Rationality thus construed is, therefore, not limited, in principle , to human action. Instinctive animal behavior can be deemed rational in this respect. Likewise, properly functioning computer programs, automatic pilots and thermostats,programmed to“sense,”“assess,”and respond to a situation effectively, are often said to perform rationally without doing too much violence to the term. It is a notion of rationality used widely in talk of skillful feats one trains to perform instinctively without 1 2 The View from Within having to ponder on them each time anew. It matters little how the skill was originally acquired. Often, as Michael Polanyi famously argued, the learning process may itself remain tacit.1 Most of us learn to perform highly complex feats of logical reasoning, for example, without ever having studied logic as such. The move from tacit human knowledge acquisition to preprogrammed, hardwired animal and artificial skilled response is then quite natural. To paraphrase Bernard Williams’s well-known distinction, it is not X’s reasons for φ-ing that are being judged but the extent to which φ-ing is considered the reasonable thing for X to do.2 Rationality on such a showing is hence a matter of situation analysis. It has to do with the weighing of possibilities and calculation of possible outcomes. Due to the supposedly calculable, algorithmic nature of its underlying logic, this notion of rationality has attracted enormous attention in recent years, especially among decision and game theorists. Thus construed, rationality still requires reflection and sound assessment , but it is not the actor who is required, or, in most cases, even quali- fied to do so. To be considered rational she is expected to act in conformity with what is considered appropriate by the best means available but without necessarily having to have worked it out for herself. Indeed, unless the actor happens to be the referring expert herself, her own account of her actions will be of little if any relevance to their rationality.3 Nor need the required approval be unanimous, or granted immediately .Approval,and with it assessments of rationality,may be conceded by some and not by others and, even then, only long after the event. Rationality , in this sense of the term, is time- and context-dependent, in principle , and is hence interestingly relativized despite its supposedly objective ,calculable nature.Actions deemed inappropriate and hence irrational by some may come in time to be considered fitting and perfectly rational by others—and vice versa. Although the rationality of the type of action in question may have been ascertained long before the specific act under consideration was performed (as in the case of standard countermoves in chess, or an appropriately applied well-honed mathematical technique), evaluations of this first category of rationality are always backward-looking.What is deemed rational is forever what the actor ended up doing or refraining from doing. One does not embark on a process of deliberation that can be deemed rational before it is completed and its outcome independently assessed. Rationality, in this sense of the term, is always decided after the event. It is not a prospective evaluative category of acting but a retrospective evaluative category of types of actions.What makes for the rational is not what goes into practical reasoning, if you wish, but what comes out. It is the move the actor actually makes that is assessed regardless of the manner in which she reasoned or conducted herself prior to making it...