Douglas Portmore presented an intricate argument designed to preserve the “deeply compelling” theory of act-utilitarianism from its counter-intuitive implications. In my view his argument falls short; in trying to avoid the counter-intuitive arguments against act-utilitarianism, he wrongly abandoned the central idea of a greatest impersonal good, leading to an over-arching consequentialist theory that does not preclude the counter-intuitive implications he wanted to avoid.
The consequentializing project for which Portmore argued is designed to meet the objections of non-consequentialists that there are other moral factors than aggregate utility that determine an act’s deontic status (3.) The consequentializer “insists” – his term – that the non-consequentialist arguments are available to the consequentializer as criteria for ranking outcomes, in effect recasting the non-consequentialist argument into consequentialist terms. The purpose is to “keep what’s compelling about act-utilitarianism … while avoiding what’s problematic about it” (4.)
I add here that at least twice he argued that the consequentializer should “insist” on adopting non-consequential arguments as consequentialist outcome ranking criteria; the insistence without an argument for the move seems suspect, and exploring that move itself seems important.
Early in the paper Portmore considered an argument by Mark Shroeder that the consequentializing project cannot succeed, based on a confusion in different senses of the term “good.” Schroeder argued that the consequentializer will need to “accommodate agent-centered constraints” (4,) which impose limits on agents’ “freedom to pursue the best consequences, impersonally construed … they include both restrictions and special obligations” (4n5.) He continued that the compelling idea of act-utilitarianism is that rightness must be defined in terms of the monadic predicate “good,” but to accommodate restraints on agents “good” must be defined in dyadic terms, or “good-in-relation-to” the agent. Per Schroeder, there is no good way to connect those uses of “good” such that the agent subject to constraints will accept the compelling idea of the unbounded monadic “good” (4-5.)
Portmore acknowledged the cogency of the objection and goes on to state that philosophers have up to then “misidentified what it is so compelling about act-utilitarianism” (5,) that “it is alway permissible for an agent to bring about the most good,” as the term “good” cannot figure in goal of consequentialism (6.) Portmore then developed an argument what he calls the “permissibility of maximizing view”, or PMV:
“It is always permissible (in the knowledge supposing sense) for an agent to act so as to bring about the highest ranked available outcome, i.e., the outcome that she has better object-given reasons to prefer above all other available options” (17.)
The PMV makes no claims about the monadic “good,” leaving it to agent-relative judgments to determine the highest ranked outcome within her factually based (“object-given reasons”) outcome evaluation criteria. In developing the argument for the PMV, Portmore argued to include a variety of consequentialist theories, including Egoism of the Present, Ethical Egoism, Act-Utilitarianism, and Self/Other Utilitarianism. These differ in being either agent-neutral or agent-relative, and temporally-neutral and temporally relative. He considers it a “great merit” that defining consequentialism in this way includes a broad variety of consequentialist theories. (11-12.)
While the PMV avoids using the monadic term “good,” it seems to me that including such a broad sweep of consequentialist theories and making the consequentialist agent the locus of determining the good leaves the monadic and dyadic senses of “good” in play, not in the statement of the PMV, but in the consideration of each agent. This is clear in his later discussion of Schefflerian utility, in which agents can rank their own utility with no weight at all or up to ten times the weight of anyone else; some agents consider the greater good, some do not, but the distinction remains. Given the variety of ethical choices in Schefflerian utility, Portmore concluded that “Thus Jill can permissibly choose to perform a1 as opposed to a2 even though performing the latter would do more to promote the impersonal good” (27) – explicitly contrasting the monadic and dyadic senses of the good in evaluating Jill’s action. The contrast in the senses of the “good” in the various agents’ actions is even more stark when considering Portmore’s Modified Schefflerian Utiltarianism, which accommodates both restrictions and special obligations (28n50.)
Furthermore, leaving the sense of “good” to the agent within the broad scope of consequentialist theories as summarized in the PMV leaves unanswered Schroeder’s objection that the consequentializer must include restrictions – such as special obligations – on agents to avoid the counter-intuitive implications of act-utilitarianism. Act-utilitarianism remains among the viable consequentialist theories for which Portmore argued!
It seems to me that in the PMV Portmore’s argument wrongly abandoned the idea of a monadic, impersonal good toward which agents should act. Abandoning the idea of a greatest agent-neutral good in the PMV resulted in including a broad varieties of ethical consequentialism that are simply incompatible with each other. Yes, how to achieve the greatest good may lead to counter-intuitive conclusions, such as killing the few redheads in a superstitious society, but I argue that the problem in that case is not the idea per se of a greatest good, but what properly constitutes the greatest good. Here I agree with Schroeder: to achieve that greatest good, agent-relative constraints are integral to a consequentialist moral theory.
Portmore, Douglas. “Consequentializing Moral Theories.” Penultimate draft of 4/22/2006, subsequently published in the Pacific Philosophical Quarterly.
This is adapted from a series of responses I wrote for an independent study in analytic ethics of partiality and consequentialism.