[EM] Re: Why it doesn't matter if "prefer" means anything

[MIKE OSSIPOFF]

Jobst--

Oh shit.

Why can't you send me an erroneous criticism of my criteria, like everyone 
else? :-)

I've gotten so used to erroneous criticisms, that I was sure that yours must 
be one, but of course it wasn't.

Maybe the best solution is to just say that my criteria only apply if 
preferences are transitive.

That's my best immediate answer.

You might suggest changing my definition of sincere voting, but it's what 
works with my criteria.

Well, it isn't really a problem though. A valid criticism, but not a 
problem, and not a serious fault of my criteria or sincerity definition:

Say you propose a method, and, for that method and one of my criteria, I 
write a failure example with transitive preferences. Your method fails the 
criterion, because it was possible to contrive an example in which it fails. 
The fact that my criteria that stipulate sincere voting don't applly to 
examples with intransitive preferences doesn't mean that I can't supply a 
failure example for your method.

If you write an example with intransitive preferences, then if a criterion 
of mine stipulates sincerity, that just means that your example isn't one 
that tests the method by my criterion.

No problem.

Yes, as you said, that voter with intransitive preferences has no way of 
voting sincerely, as I define voting sincerely. But, for one thing, my 
sincerity definition is only for use with my criteria, not for evaluation of 
voters and their motivations. For another thing, that voter shouldn't 
entirely blame my definition. Surely she would have to admit that she 
contributed at least partly to the definition-mismatch by having 
intransitive preferences.

Mike Ossipoff


Mike Ossipoff

_________________________________________________________________
On the road to retirement? Check out MSN Life Events for advice on how to 
get there! http://lifeevents.msn.com/category.aspx?cid=Retirement