1

A follow-up, of sorts, on this question, in which I'd like to refine the notion of moral obligation. Consider some instances in which the notion might arise:

  1. Most people agree that a trolley operator is morally obligated to sacrifice one tied-down man in order to save five (all other things being equal, etc.).
  2. Some people think that an operator (or perhaps a bystander) is morally obligated to push a fat man onto the rails in order to save five.
  3. Many people think that a bystander is morally obligated to save a drowning child, if the risk to himself is minimal.
  4. Some people think that one is morally obligated to purchase fair-trade coffee only.

I am interested in quantifying the degrees and severities of various moral obligations. For example, even for those who see a moral obligation in #2 and #4, would probably not like to see non-compliers punished by law (for different reasons, actually). In #1 and #3, I could easily see those who see a moral obligation entertain some sort of legal penalty -- which might even be somewhat severe for #3.

Have any such studies been carried out?

Improve this question
3
  • 2
    The possibility of ranking and tradeoffs does not necessarily imply quantification, some obligations may be comparable by priority, others not, wrongs are not made up for by other rights, but all have to be balanced in decisions. Nozick tried to formalize such "balancing structure" of ethics (see his Philosophical Explanations), as opposed to quantification. One attempt to quantify ethics was Bentham's hedonic calculus (for values rather than obligations), but such attempts are broadly seen as unsuccessful.
    – Conifold
    Commented Jun 14, 2019 at 15:01
  • 2
    This page is a good place to start: (en.wikipedia.org/wiki/Ethical_calculus)
    – Donate to the Edhi Foundation
    Commented Jun 14, 2019 at 20:07
  • Welcome to Stack Exchange. I'm not quite sure of how much of a formalized system you want and how valid the quantification should be. How much structure do you require?
    – Donate to the Edhi Foundation
    Commented Jun 14, 2019 at 22:50

1 Answer 1

Reset to default
0

There is a close family of arguments that shows up, first I know, in the neo-antiquity of analytic philosophy, the time of Moore and Prichard; then in Onora O'Neill's Constructions of Reason and, more colorfully, as Christine Korsgaard's map-reading disanalogy in the comparison of theoretical and practical reasoning (I think it's in The Sources of Normativity, but I'm not 100% sure). The argument is roughly as follows (I am mainly adapting Prichard's formulation):

  1. Before I have performed some action a, a doesn't really exist.
  2. If an obligation were a property of a, it would be a property of a non-existent object.
  3. But no property is had by a non-existent object; existence is a precondition of having properties.
  4. Therefore, if obligations are properties of actions to be done, then the property of being obligated is not actually instantiated anywhere at all.

O'Neill makes the point in terms of the notion of subsumption-of-particulars-under-concepts (practical reasoning is not the application of a concept to an already-given action, but produces the action according to the concept in itself); Korsgaard contrasts knowing how to read a map generally from knowing how to use a map for particular purposes, saying something about how normative reasoning isn't about looking for little "normative flags" attached to intended spots on a normative map.

So all that could seem like a reason to try out quantifying over obligations instead. But the next problem is that the nonexistent actions of which obligatoriness was predicated, are now somehow properties of obligation-objects. ∃o(Ao), read as, "There is an obligation such that it is an obligation to do A," turns "to do A" into a predicate of the obligation.

But next, we could also try out relations, here. Let's throw in B for some agent. Say ∃oB(oRB), for some relation R. Would this work? The least I can say is that Prichard argued against a relationalist picture as much as a monadic-property one, saying that a nonexistent relatum doesn't sustain the existence of a relation. On the other hand, by relating obligations to agents, maybe we can somehow squeeze a description of an action into the meaning of R, and avoid the problem of non-existent functions. Or we need a three-place relation between some o, some B, and some nonexistent a, and we are still doomed. I'm not sure.

At any rate, Wansing[??] contemplates the matter in important detail:

obligation-quantification

Note that quantifying over obligations would also make the generic duality with permission into a more complicated affair. It is no longer possible to take some o and say ~o~ = p, because o is not a sentence/proposition. Maybe we might write something like ~∀o~(Ao) = ∃p(Ap), but I don't really know if that makes sense. (I was going to ask about these topics in a new question on this SE, but yours showed up in the "similar questions" prompt window.)

P.S. Ward[70] uses the phrase "quantify over obligations," though I'm not sure to what effect. Rothschild and Yablo[20] mentions that, "The idea of quantifying over obligations figures prominently in Moltmann’s work, e.g., Moltmann [2017]," the reference apparently being to this.

Improve this answer

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .