Even though the fate of the machine, from Descartes forward was intimately coupled with that of the animal, only the animal (and only some animals, at that) has qualified for any level of ethical consideration. And this exclusivity has been asserted and justified on the grounds that the machine, unlike the animal, does not experience either pleasure or pain. Although this conclusion appears to be rather reasonable and intuitive, it fails for a number of reasons.
First, it has been
practically disputed by the construction of various mechanisms that now appear
to suffer or at least provide external evidence of something that looks like
pain. As Derrida recognized, "Descartes already spoke, as if by chance, of
a machine that simulates the living animal so well that it 'cries out that you
are hurting it.'" This comment, which appears in a brief parenthetical
aside in Descartes' Discourse on Method, had been deployed in the course
of an argument that sought to differentiate human beings from the animal by
associating the latter with mere mechanisms. But the comment can, in light of
the procedures and protocols of animal ethics, be read otherwise. That is, if
it were indeed possible to construct a machine that did exactly what Descartes
had postulated, that is, "cry out that you are hurting it," would we
not also be obligated to conclude that such a mechanism was capable of
experiencing pain? This is, it is important to note, not just a theoretical
point or speculative thought experiment. Engineers have, in fact, not only
constructed mechanisms that synthesize believable emotional responses, like the
dental-training robot Simroid "who" cries out in pain when students
"hurt" it, but also systems capable of evidencing behaviors that look
a lot like what we usually call pleasure and pain.
Second it can be contested on epistemological grounds insofar as suffering or the experience of pain (or pleasure) is still unable to get around or resolve the problem of other minds. How, for example, can one know that an animal or even another person actually suffers? How is it possible to access and evaluate the suffering that is experienced by another? "Modern philosophy," Matthew Calarco writes, "true to its Cartesian and scientific aspirations, is interested in the indubitable rather than the undeniable. Philosophers want proof that animals actually suffer, that animals are aware of their suffering, and they require an argument for why animal suffering should count on equal par with human suffering." But such indubitable and certain knowledge, as explained by Marian S. Dawkins, appears to be unattainable:
At first sight, 'suffering' and 'scientific' are not terms that can or should be considered together. When applied to ourselves, 'suffering' refers to the subjective experience of unpleasant emotions such as fear, pain and frustration that are private and known only to the person experiencing them. To use the term in relation to non-human animals, therefore, is to make the assumption that they too have subjective experiences that are private to them and therefore unknowable by us. 'Scientific' on the other hand, means the acquisition of knowledge through the testing of hypotheses using publicly observable events. The problem is that we know so little about human consciousness that we do not know what publicly observable events to look for in ourselves, let alone other species, to ascertain whether they are subjectively experiencing anything like our suffering. The scientific study of animal suffering would, therefore, seem to rest on an inherent contradiction: it requires the testing of the untestable.
Because suffering is understood to be a subjective and private experience, there is no way to know, with any certainty or credible empirical method, how another entity experiences unpleasant sensations such as fear, pain, or frustration. For this reason, it appears that the suffering of another (especially an animal) remains fundamentally inaccessible and unknowable. As Peter Singer readily admits, "we cannot directly experience anyone else's pain, whether that 'anyone' is our best friend or a stray dog. Pain is a state of consciousness, a 'mental event,' and as such it can never be observed."
Third, and to make matters even more complicated, we may not even know what "pain" and "the experience of pain" is in the first place. This point is something that is taken up and demonstrated by Daniel Dennett's "Why You Can't Make a Computer That Feels Pain." In this provocatively titled essay, originally published decades before the debut of even a rudimentary working prototype, Dennett imagines trying to disprove the standard argument for human (and animal) exceptionalism "by actually writing a pain program, or designing a pain-feeling robot." At the end of what turns out to be a rather protracted and detailed consideration of the problem, he concludes that we cannot, in fact, make a computer that feels pain. But the reason for drawing this conclusion does not derive from what one might expect, nor does it offer any kind of support for the advocates of moral exceptionalism. According to Dennett, the reason you cannot make a computer that feels pain is not the result of some technological limitation with the mechanism or its programming. It is a product of the fact that we remain unable to decide what pain is in the first place. The best we are able to do, as Dennett illustrates, is account for the various "causes and effects of pain," but "pain itself does not appear." What is demonstrated, therefore, is not that some workable concept of pain cannot come to be instantiated in the mechanism of a computer or a robot, either now or in the foreseeable future, but that the very concept of pain that would be instantiated is already arbitrary, inconclusive, and indeterminate. "There can," Dennett writes at the end of the essay, "be no true theory of pain, and so no computer or robot could instantiate the true theory of pain, which it would have to do to feel real pain." Although Jeremy Bentham's question "Can they suffer?" may have radically reoriented the direction of moral philosophy, the fact remains that "pain" and "suffering" are just as nebulous and difficult to define and locate as the concepts they were intended to replace.
Finally, all this talk about the possibility of engineering pain or suffering in a machine entails its own particular moral dilemma. "If (ro)bots might one day be capable of experiencing pain and other affective states," Wendell Wallach and Colin Allen write, "a question that arises is whether it will be moral to build such systems—not because of how they might harm humans, but because of the pain these artificial systems will themselves experience. In other words, can the building of a (ro)bot with a somatic architecture capable of feeling intense pain be morally justified and should it be prohibited?" If it were in fact possible to construct a machine that "feels pain" (however defined and instantiated) in order to demonstrate the limits of moral subjectivity, then doing so might be ethically suspect insofar as in constructing such a mechanism we do not do everything in our power to minimize its suffering. Consequently, moral philosophers and robotics engineers find themselves in a curious and not entirely comfortable situation. One needs to be able to construct such a mechanism in order to demonstrate the moral standing of machines; but doing so would be, on that account, already to engage in an act that could potentially be considered immoral. Or to put it another way, the demonstration of machine moral subjectivity might itself be something that is quite painful for others.
Comments