Known Unknowns

To the Editors:

Thomas Nagel, in his deep and illuminating review of Christine Korsgaard’s Fellow Creatures [NYR, March 21], may have slightly overstated the difference between humans and nonhuman animals. Unlike humans, he writes, “animals…as far as we know, do not evaluate their own beliefs and motives before acting on them.” That is, animals lack our capacity for “rational self-assessment”—the basis (as Nagel observes) of science and morality.

In fact, there is evidence that some nonhuman animals—dolphins, monkeys, pigeons, rats—are indeed capable of consciously evaluating their own beliefs. In one experiment carried out in the 1990s, a dolphin named Natua was given the task of distinguishing between auditory pitches of different frequencies. When Natua judged correctly, he was given a reward. He was also trained to press an “opt-out” paddle if he wished to escape this task, in which case he would receive a smaller reward. As the experimenter moved the high and low pitches closer together, making the perceptual task more difficult, Natua would sometimes, after an interval of hesitation, choose to press the opt-out paddle. I’m not sure of what I’m hearing, he seemed to be saying. Macaque monkeys have also learned to avail themselves of an opt-out key when they are uncertain about their perceptual judgments. Moreover, the macaques spontaneously generalized this opt-out behavior from tasks involving perception to those involving memory. I’m not sure that I’m remembering things right, they seemed to be thinking. Since the macaques appeared to reflect for several seconds before deciding to opt out, there is good reason to believe that their judgments were arrived at consciously. (An excellent account of such research can be found in Consciousness and the Brain, by the French brain scientist Stanislas Dehaene.)

So some nonhuman animals do seem able to “step back” and rationally assess their own beliefs. Whether they can take a similarly evaluative stance toward their desires is another question. Still, even rudimentary signs of reflective self-consciousness in nonhuman animals can only fortify the case—made by Korsgaard in her new book and deemed “undeniably powerful” by Nagel in his review—for regarding them as at least “passive members” of the moral community.

Jim Holt
New York City

Thomas Nagel replies:

I thank Jim Holt for his interesting letter. I am open to the possibility that some nonhuman animals evaluate their own beliefs and motives. It is an empirical question. But the experiments he describes pose problems of interpretation. For example, Natua might just prefer the sure thing of a lesser reward for pressing the “opt-out” paddle to the lesser likelihood of a greater reward for identifying the higher frequency, when he isn’t sure. That would require only uncertainty about the sounds themselves, not a second-order observation of his uncertainty. Further, even if Natua does reflect on his beliefs in this case, there is no evidence that he evaluates them, to determine whether they are justified. Dolphins may well be capable of rational assessment of their own beliefs and desires, but discovering evidence for this without talking to them would not be easy.