In a brief and deeply interesting 2017 Acta Analytica paper, Peter Baumann argues that there are cases of necessarily incompatible but mutually consistent desires, that this is a common problem, and that there is no solution in sight. I will argue that Baumann needs to make some additional assumptions in order to support his claim that there are necessarily incompatible and mutually consistent desires; if they do exist, then they are sometimes beneficial; and if they are sometimes involved with problematic outcomes, then the mere presence of incompatibility and consistency does not cause the subject to frustrate the (more) beneficial desire.
This is a preview of subscription content, access via your institution.
Buy single article
Instant access to the full article PDF.
Price includes VAT (USA)
Tax calculation will be finalised during checkout.
If I desire on Day 1 to φ and I desire on Day 2 not to φ, then it does not follow in any sense or at any time that I want to φ and not to φ on Day 1 (or on Day 2). That is surely right. In fact, Dubois et al. (2017) argue that the logic of desire is something like the opposite of the logic of belief, and that desire is not closed under conjunction but is closed in the following curious disjunctive way: if S desires φ or ψ, then S desires φ and S desires ψ.
One thought is that the content of an immortal person’s general and indeterminate desire—I go to the dentist at some point—can be true only if one of these contents is false—
I do not go to the dentist at t1.
I do not go to the dentist at t2.
… <and so on for literally all t>.
The idea here is that we get a contradiction: I cannot go at some t if, for all t, I never go at t. One might think that the matter is not so simple. For an immortal subject, there will always be a further opportunity to go to the dentist (and, more generally, to satisfy the general, non-indexical desire in question). But I think that this is irrelevant if we are considering de-re contents of desires, as is the case here. As long as the contents of the subject’s specific desires not to φ exhaust time forward-directionally, and of course the subject has a general non-indexical desire to φ, then we get a contradiction between the conjunction of the contents of the subject’s specific indexical desires not to φ and the content of the subject’s general non-indexical desire. This can occur if the subject is immortal and/or if time is forward-directionally finite.
Time (and the universe) might lack a beginning and so be infinite. Yet time (and the universe) might also have an end. That is not infinity in the required sense here. Rather, time would have to be like the positive integers: it would need a beginning but no end. For if time has an end, then it might have an end identical with the (known) end of a finite subject’s lifespan. What is required, then, a little more precisely, is that time spans (and is known to span) longer than the finite subject’s lifespan. Depending on which finite subject we are talking about (and, in particular, how far in the future that finite subject exists), physicists such as Julian Barbour would take issue with the assumption that time will necessarily extend beyond the finite subject’s lifespan: Barbour (2001) argues, contra Einstein (and Feynman, Newton, et al.), that time would end without change.
Barbour, J. (2001). The end of time: The next revolution in physics. New York: Oxford University Press.
Baumann, P. (2017). Necessary incompatible consistent wants. Acta Analytica, 32, 489–490.
Dubois, D., Lorini, E., & Prade, H. (2017). The strength of desires: A logical approach. Minds & Machines, 27, 199–231.
For helpful comments on the longer paper from which this shorter reply piece developed, I am grateful to audiences at the Wisconsin Philosophical Association's 2020 Meeting, as well as the 2020 Rocky Mountain Ethics Congress (RoME). Thanks in particular to Chris Heathwood, Alastair Norcross, David Boonin, Mylan Engel, Schuyler Sturm, Gagan Sapkota, and others. I am especially grateful to Chris Heathwood for suggesting the first version of the example in Section 2 in this paper (in his commentary at my RoME session), for arguing that consistency entails compatibility (though I disagreed and still disagree, for reasons explained on pp.4-10 above), and for patiently corresponding with me through email about some of the distinctions in Section 2. I am indebted to Chris for helping to clarify my thoughts about the problems I've addressed in this paper. Also, thanks to Alastair Norcross and an anonymous reviewer for Acta Analytica for suggesting that Baumann might appeal to the weaker sense of consistency described at the end of Section 2.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Coren, D. No Problem of Consistent Incompatible Desires: a Reply to Baumann. Acta Anal 36, 465–474 (2021). https://doi.org/10.1007/s12136-020-00452-y