Robert Audi, The Architecture of Reason: The Structure and Substance of Rationality.
Audi's book stands out among works on rationality in that it offers a theory of both theoretical and practical reason and of the extensive parallels between the two. The title itself--The Architecture of Reason--reveals a crucial aspect of this theory: in both the theoretical and practical spheres there is a superstructure and a foundation on which it is based. The foundation is the same in both cases: experience, broadly understood. Beliefs are the basic elements of theoretical rationality and desires the basic elements of practical rationality. Beliefs provide us with a map of the world, desires with an itinerary. Just as noninferential beliefs are justified by experience, intrinsic desires--desires for something for its own sake--are warranted by rewarding experience. In this way, maps are shown to be correct and itineraries worthwhile.
Naturally it is impossible to do justice to a work of this complexity in a short review. Since Audi's theory of the rationality of intrinsic desire is perhaps the most significant accomplishment of this work, I will focus on it here. Audi contrasts his own view with instrumentalism. According to pure instrumentalism, there are no rational intrinsic desires. An action is rational provided that it satisfies my intrinsic desires whatever they are. According to procedural instrumentalism, actions are rational relative to intrinsic desires that I would have if I had reflected adequately under certain conditions. In contrast, Audi's own view is a substantive foundationalism: the substantive features that render intrinsic desires rational are revealed in experience as their foundation.
A noninferential belief is not rational simply because I have it, but rather because it is supported by experience. Similarly, an intrinsic desire to do something is not rational simply because I have the desire or have gone through a certain reflective procedure. Rather, it is rational because certain aspects of the activity are experienced as rewarding. Several explicative comments are in order. (1) To say that an experience is rewarding is not to say that it has some general quality of "rewardingness." Rather, this is merely a way of summarizing different ways in which an intrinsic desire is rendered rational. Audi wants to be quite open about the substantive criteria of rationality, but he mentions a few. Rational persons want their own happiness (eudaemonistic principle), pleasure, avoidance of pain (hedonic principle) and things they are justified--among other things on account of their experience--in taking to be intrinsically good (valuational principle). Thus, the meaning of saying that something is a rewarding experience will vary greatly from case to case. (2) When I have a rational desire to do something (say, to swim) for a certain rewarding experience (say, for pleasure), I do in most cases not want to perform the action as a means to the rewarding experience. Rather, I want to do something for a certain quality of the activity (say, for the enveloping presence of water in swimming) that renders the activity rewarding (say, pleasurable). (3) Audi wants to leave it open that there may be things of intrinsic value other than experience. The important point is that an intrinsic desire for a thing cannot be rational unless it is based on an awareness of its value through experience. (4) Rational desires warranted by experience ate not just rational in the minimal sense of the absence of irrationality, but rather in the sense of being consonant with reason. If the relevant experience would really be rewarding, there is an objective ground for doing what I desire to do.
Audi argues that the rationality of wanting things for certain qualities is grounded in the experiential qualities themselves, not in the qualities "conceived as experienced by me" (101). For example, it is the enveloping presence of the water that grounds the rationality of my desire to swim, not this quality as experienced by me. This idea is central to Audi's two arguments for the rationality of altruistic desires (I ignore the argument from reasonableness). The analogy with theoretical reason also plays an important role. A visual impression of a tire nearby provides a reason for believing that there is tire nearby irrespective of whether it is my or your visual impression. Thus, my awareness of your visual impression provides a ground for me to believe this. Analogously, the rewarding experience of the enveloping qualities of the water provides grounds for the desirability of swimming irrespective of whether the experience is mine or yours. Thus, my awareness of your rewarding experience provides a reason for me to desire that you have those experiences.
But Audi's argument is flawed. Assume that my awareness of your visual impression of a nearby tire does indeed provide me with a reason for believing that there is tire nearby. This is a reason for believing that it is true that there is a tire nearby. Similarly, my awareness of your rewarding experience of swimming may give me a reason to believe that swimming is desirable. And this belief may give me a reason for desiring to swim myself But it gives me absolutely no reason for having the altruistic desire that others also have what is desirable. In order to justify altruistic desires, such a reason must be provided and it is not provided by Audi's argument.
Audi offers a second, related, argument for the rationality of altruistic desires. My awareness of your and my qualitatively comparable rewarding experiences gives me reason to believe that your experience is as good as mine. Now, according to Audi, if I hold this belief on account of the qualities of the experiences and have no desire that you also enjoy the experience, then "my desires are not adequately integrated with my beliefs" (144). Thus, though lacking the desire that you enjoy the experience would not entail irrationality, such a desire is nevertheless a demand of reason. If the first argument is flawed, this one also is. The desires of an egoist are just as well integrated with such beliefs as the desires of an altruist with those same beliefs. The egoist simply does not desire that others enjoy what is good. This is what makes her an egoist.
Why should somebody care that others enjoy what is good? In my view, the question should be answered by pointing to certain substantive qualities of caring and not caring for others. Caring that others enjoy what is good is often kind or considerate, not caring is often disrespectful of others or unfair to them. In my opinion, such qualities reveal what is rational about altruistic desires. I mention this here because such a view is in a certain sense simply a consequent extension of Audi's substantive foundationalism about practical reason: an extension in which "moral" qualities such as kindness and respect are just as fundamental as the qualities involved in Audi's eudaemonisuc, hedonic, and valuational principles. (It would need to be discussed in which sense of experience such an extension of substantive foundationalism must hold that the importance of such qualities as kindness and respect must ultimately be discovered through experience.) Audi does allow that injustice may itself be a ground for the rationality of wanting to avoid inequality. Perhaps he does not think of the "moral" qualifies as less fundamental; this book alone does not make his view on this matter entirely clear. This much is clear: an extension of Audi's own substantive foundationalism would correctly hold that it is precisely such substantive "moral" qualifies as kindness and respect that reveal the rationality of altruistic desires and moral principles, and in this book Audi mistakenly attempts to trace the rationality of altruistic desires and moral principles back to something else. Thus, I conclude that Audi's substantive foundationalism about practical reason is a highly important contribution to contemporary philosophy, but that this work offers the wrong picture of the rationality of altruistic desires and moral principles.
Humboldt-Universitat zu Berlin