A Cautionary Note against “Precautionary Reasoning” in Action Guiding Morality*

Abstract

This paper responds to a thesis advanced in a paper by Deryck Beyleveld and Shaun Pattinson (Beyleveld and Pattinson 2000), and developed in subsequent works by them and Roger Brownsword (Beyleveld and Brownsword 2007; Pattinson 2006). It addresses the question of “precautionary reasoning,” which is provided as a practical moral measure to guide agents when faced with traditionally troubling issues of epistemic uncertainty (such as whether a fetus or a person in a persistent vegetative state have equivalent moral status to a “normal” adult human). These authors' commitment to the “Principle of Generic Consistency” (PGC) (Gewirth 1978), and its attendant moral structure, gives them an internally coherent action guiding framework with which to analyse the morality of law and, more generally, the behaviour of agents (see Beyleveld and Brownsword 1986, 2001, 2007; Pattinson 2006). This can be done by reference to an action's effect on agents' generic rights. Briefly: Generic rights are to the generic features of agency (those capacities an agent needs to be able to act at all with any general chance of success, whatever its purposes might be). Interference with, or deprivation of, a generic feature of agency will interfere (or tend to interfere) with an agent's capacity to act or to act successfully, regardless of what the action envisaged is. (Beyleveld and Pattinson 2000, n. 4). The principle holds that an agent denies his own agency if he fails to respect the agency of others. Thus, Gewirthians such as Beyleveld, Brownsword, and Pattinson can rest assured that they are acting in a morally satisfactory manner so long as they do not unjustifiably breach another's generic rights. For cases where it is not evident that we are dealing with an agent, we are advised that precautionary reasoning becomes the morally prudent course of action (Beyleveld and Pattinson 2000; Beyleveld and Brownsword 2007, chap. 4; Pattinson 2006, chap. 16). This resembles “Pascal's wager”: It dictates that as we do not “know” for certain that something is not a moral agent, we do well to presume that it is. Failure to do this may result in our committing a moral harm. If in fact the thing is not an agent, we do no moral harm to it by treating it as if it were. In their exposition of the consequences of this line of reasoning, Beyleveld and Brownsword formulate their view in the following way: Even though X is apparently only a partial agent, there remains a risk that if I suppose that X is not an agent, and act accordingly, X is an agent, and I will have deprived X of the protection of the PGC to which X is categorically entitled. Thus, it is categorically necessary, all things being equal, to do whatever one can do to avoid this consequence, and this will require me to recognise duties not to harm and to protect X. (Beyleveld and Brownsword 2001, 122) Although there is no inherent intellectual flaw in the idea of precautionary reasoning, apart from a possible tendency towards conservatism, we seek in this paper to demonstrate its failure to provide a useful tool for guiding action. There are two predominant threads to our analysis. The first relates to the capacity to recognise instances when precautionary reasoning is appropriate. We contend that this is a matter that has not been given full consideration by advocates of precautionary reasoning, and on reflection they have not been able to come up with a satisfactory answer. The second relates to the fact that in practice the rights of a possible (or “non-ostensible”) agent must yield to the rights of an apparently definite (or “ostensible”) agent. As this will commonly be the situation in cases involving the sorts of disputes that precautionary reasoning is aimed to assist in resolving, it can be seen only to help us (if at all) in that tiny scintilla of cases where our actions will have absolutely no influence on anyone other than the possible agent. We will begin with a short analysis of choice (or wil