1. Rule R applies to Situation S.
 
2. Person P says that Exception E, some kind of personal disposition,
   invalidates R in P's instance of S.
 
3. To support (2), P argues that a number of factors apply to E:
 
3.1. That E is very personally significant.
 
3.2. That E shapes their relationship with the world.
 
3.3. That E gives them the reason for their life or career or vocation.
 
3.4. That E gives them some deeper, existential or spiritual satisfaction.
 
3.5. That E has been generally accepted for a long time by others.
 
3.6. That E has been historically significant.
 
3.7. That only a small minority think that this particular E does not support
     the use in (2).
 
etc.
 
4. Therefore, E should invalidate R as it applies to S & P.
 
The point here is simple:
 
What you put in the place of 'E' varies the whole argument which is a form of
special pleading. If you put a religious belief and practice system in as E,
then (3.1-3.7) are automatically given more justificatory weight than if you
substitute it with something like, say, a political belief, a sexual preference,
a hobby, a taste for a particular type of art or music, a romantic relationship
and so on.
 
The point I am making is that there are a large number of categories of things
of which members could easily satisfy most or all of (3.1-3.7), but when you use
them in the argument, those who accept the argument for religion suddenly start
denying it for these other things. The argument has the form that one, some or all
of the arguments in (3.1-3.7) are necessary for the conclusion - maybe it is even
stronger that (3.1-3.7) are sufficient for the conclusion. But if you can find
something that satisfies (3.1-3.7) but does not satisfy (4), the whole thing falls
apart.
 
Some people balk at the suggestion that religious belief or practice can be directly
compared to any of the personal dispositions listed in the paragraph before last.
Let me rebut a few objections in advance:
 
1. The Religion Is More Complex Than That argument. Sure. Religion is a complex,
   socially intriguing phenomena that admits of no simple definition. I'm not
   attempting to define religion. If what I've said about religion is grossly
   wrong, and is not what any sensible person would say about religion, that's
   fine. Let's talk about noigiler, a phenomena that matches the few stipulations
   I have made about religion.
 
2. Religion isn't anything like those things. Sure, I agree. There are important
   and significant differences between religion, political beliefs, ideologies,
   sexual practices, and other dispositional or epistemic attitudes. But
   religions share many of the properties. We can play the necessary-and-sufficient
   game, or we can play the family resemblance game. Dispositions or attitudes are
   used in personal rule-exemption arguments. I want to know why people have such
   a different attitude towards religion as they do to other things that could
   feasibly stand in for the role religion plays in personal exemption arguments.
 
3. But what about act-utilitarianism? Yes, that's a possible problem. If you
   subscribe to a strong form of act-utilitarianism or maybe even preference
   utilitarianism, you would reject the attempt to universalise this argument.
   But here, you disagree with the argument's leap from the third premise to
   the conclusion in all cases. I'm interested in the way E-criteria are constructed
   to justify some cases but not others, and more importantly whether an intuitively
   plausible E-criteria could be formulated that justifies religious exemptions but
   not other exceptions.
 
4. There is a possible objection that I find rather poor: namely that this is a
   question of descriptive ethics. It's not. It's normative in the sense that it is
   evaluative - I am trying (in a rather ham-fisted way) to construct a framework to
   test E-criteria against our moral intutions for consistency. I reject the idea
   that what I am doing is somehow tied to any particular moral community - all I'm
   trying to do is find out how we would test E-criteria against each other.