Imperatives, phantom pains, and hallucination by presupposition

Abstract

Several authors have recently argued that the content of pains (and bodily sensations more generally) is imperative rather than descriptive. I show that such an account can help resolve competing intuitions about phantom limb pain. As imperatives, phantom pains are neither true nor false. However, phantom limb pains presuppose falsehoods, in the same way that any imperative which demands something impossible presupposes a falsehood. Phantom pains, like many chronic pains, are thus commands that cannot be satisfied. I conclude by showing that some of the negative psychological consequences of chronic pain are a direct consequence of their imperative nature.

Keywords:

• Imperatives
• Intentionalism
• Pain
• Phantom Limbs

Acknowledgments

Thanks to George Graham, Dave Hilbert, Esther Klein, Vincent Lyman, Manolo Martínez, Chris Mole, Adam Pautz, and two anonymous reviewers for helpful discussion and suggestions. Audiences at University College Dublin and Northern Illinois University provided useful feedback on a previous draft.

Notes

Notes

[1] See Klein (2007, 2010). Hall (2008) and Martínez (2010) have also independently argued for a imperative view of pains and other bodily sensations. See also Craig (2003) for a similar view, though not couched in terms of imperatives.

[2] I suggest something along these lines in my (2007). Grahek (2007) has brought a stronger case to philosophical attention, that of “pain asymbolia.” Asymbolics feel painful stimuli and judge them to be painful, but appear to be entirely unmotivated to act on them. Grahek has argued that pain asymbolia is a clear case—and perhaps the only pure case—of pain stripped of motivational import. If so, this would appear to be a counterexample to versions of the imperative theory which, like mine, claim that pain is exhausted by imperative content. Grahek's characterization of asymbolia is problematic for two reasons. First, asymbolics don’t just fail to react to painful stimuli. They also fail to react to nearly all bodily threats: being threatened with a hammer, knife, or needle, to shrill whistles and burning magnesium wires, to verbal menaces, and to oncoming trucks (Berthier, Starkstein, & Leiguarda, 1988; Hemphill & Stengel, 1940; Schilder & Stengel, 1928). Second, if asymbolics merely lacked the immediate affective import of pain, one would still expect them to avoid damaging situations via simple practical reasoning: this damages my body, I care about my body, and therefore I should avoid it. This is in fact the case with the congenitally insensitive to pain, who lack pain sensations and so, a fortiori, lack the proposed separable affective dimension of pain. Asymbolics do not avoid damage, however, which is one of the striking and dangerous features of the syndrome. It therefore seems more likely that pain asymbolia is a complex depersonalization syndrome, one in which sufferers cease to recognize their bodies as a locus of potential concern (Geschwind, 1965, 269–272; Sierra & Berrios, 1998). The suggestion in the main text is an attempt to spell out what depersonalization might mean in the context of imperatives. That argument in turn raises tricky questions about the phenomenal unity of consciousness, which I address further in work in progress.

[3] Pautz (personal communication, November 29, 2009) and Tye and Cutter (forthcoming) have recently argued that the imperative theory cannot account for differences in intensity of pains. Ordinary language makes a rough distinction between the force of different imperatives, however, and it is hardly a stretch to suppose that phenomenological imperatives might make a more fine-grained distinction. In that sense, the proposal is exactly parallel to that of most representationalist accounts of ordinary sensations, on which the internal representations of objective magnitudes must outstrip the resources of everyday language in order to account for fine variations in phenomenological content.

[4] This has the apparently counterintuitive consequence that any imperative presupposes all necessary truths. This is a useful feature, however, when it comes to analyzing presupposition failure. If I order you to produce the largest prime number, for example, my imperative fails due to conflict with the necessary laws of mathematics.

[5] This would require another departure from a Hamblin-style semantics. Hamblin presupposes that there are basic actions that are not defined in terms of their relationship to other actions or states. So either the satisfaction-conditions must include stretches of world before the time of issuance (the present suggestion) or else the relevant concept of action must be broadened to include non-basic actions (in which case “stop X” enjoins a non-basic action that you can only perform if you were doing X).

[6] Note that the question of how my preferences change is distinct from the question of whether I have a reason for my preferences to change. I intend the present account to be neutral on the latter question (and so potentially compatible, for example, with the account Korsgaard gives in lecture 4 of her 1996).

[7] Sussman (2005) offers a similar argument, developed at greater length to focus on some of the additional harms unique to torture.