[EM] Paul: Your ramblings about sincerity

[MIKE OSSIPOFF]

Paul Kislanko says:

If your definition is not universally applicable, don't expect laymen to
accept it.

I reply:

What do you mean by universally applicable? My criteria and their supporting 
definitions are universally applicable in the sense that they apply to all 
methods.

The usefulness of that definition is that the voting that it calls sincere 
is voting that is not being strategically coerced. When that useful 
interpretation is applied to your intransitive voter, it could be unfairly 
accusing the voting system of coercing strategy from that voter, when all 
that's reallly happening is that the combination of voting system and 
intransitive preferences is not giving the voter a non-falsified way of 
voting.

So that's the only failing of the definition: Unfairness to the method if 
you apply it when the voter's preferences are intransitive. For one thing, 
you needn't so apply it. For another thing, the problem never comes up, 
since the only result is that your intransitive-preferences example is not 
an example that tests a method by my criteria. And that isn't a problem 
since a failure-example-writer can write an example that doesn't have 
intransitive preferences. And all it takes is one failure example to 
establish that a method fails a criterion.

Some of my criteria stipulate sincere voting by at least some voters. If you 
write an example in which those voters have intransitive preferences, then, 
as Jobst showed, they have no sincere way of voting, by my definition of 
sincere voting. That means that that example that you've written fails the 
premise conditions of my criterion.

All that means is that your example doesn't test methods by my criterion. 
Your example doesn't demonstate anything about any voting system's 
compliance with my criterion. No voting system can be shown to fail my 
criterion by using your example, because a failure example must meet the 
criterion's premise conditions.

But that isn't a problem. Sure,  you can write an example with intransitive 
preferences, and say that Plurality doesn't fail GSFC or Condorcet's 
Criterion in that example, because the criterion says nothing about your 
example. So what? I'll just write an example with transitive preferences, 
and, if it's written so as to meet all of the criterion's premise 
conditions, then the criterion will apply to it, and, if I've written it 
well, it will show Plurality failing GSFC or Condorcet's Criterion.

Yes, intransitve preferences are possible, when there are several different 
measures used for judging things. That could easily be true of someone's 
preference among candidates.

You may think that it's inelegant for a criterion to not apply to all 
examples, but of course it's routine that they don't. That's what premise 
conditions are for--to separate the examples that the criterion applies to 
from the examples that it doesn't apply to.

Not applying to examples with intransitive preferences isn't a problem at 
all, because an example-writer can write an example without intransitive 
preferences. After all, transtive preferences aren't implausible.

As for whether peoiple will accept a definition that doesn't apply to an 
unusual kind of preferences, most people don't really have a lot of concern 
about intranstive preferences, and so they aren't likely to object.

Besides, I don't ask people to accept my definition of sincere voting as a 
way to evaluate the honesty of a voter. As I already said, my sincere voting 
definition is strictly for use with my criteria.

You continued:

And no, no voter should have to explain anything to you about the
failure of your definitions.

I reply:

I was joking, Paul. The fact that, with my sincere voting defintiion, my 
criteria don't apply to someone with intransitive preferences is not a 
failure of my definition or criteria. The fact that you can't drive your car 
across the bay without a bridge is not a failure of your car. It goes where 
it's needed to go, and needn't go everywhere.

No voter should have to explain anything to me about the fact that that 
definition doesn't apply to their intransitive preferences, but I owe that 
voter no apology for calling her insincere, because I've already said, more 
than once, that that definition is only intended for use with my criteria, 
not as a way of characterizing a voter's intentions or honesty.

You continue:

If your method can't accomodate us voters'
intransivities, it is not one we need to talk about or be bothered with.

I reply:

I agree that you needn't talk about it or be bothered by it. And yet you 
seem terribly bothered by it, and are talking about it.

I didn't mean to imply that someone should be bothered by that definition. 
Quite the contrary.

The fact that it's difficult to not contradict yourself when your 
preferences are intransitive is not the fault of that definition. It's 
something that you just have to expect when you have intransitive 
preferences.

You seem to want it to be my fault, or the fault of my definition, that a 
voter with intransitive preferences can't avoid falsifying a preference on 
an Approval ballot. No, I'd say that it's completely reasonable to not call 
a ballot sincere if it falsifies a preference, even if the voter's 
preferences are such as to make that unavoidable. The only unfairness there 
is if you use my sincerity definition to accuse the method of coercing 
strategy from that intransitive voter.

No, it doesn't mean that the voter is a dishonest person, if falsification 
is unavoidable under those conditions. But  a false preference needn't be 
called sincere voting. The fact that my sincere voting defintion doesn't 
call that sincere is not a fault of that definition.

You continue:

Let's just leave it at your criteria isn't applicable to practical voting
methods.

I replyi:

Speaking of mismatch and inapplicability, you've used a pluaral noun with a 
singular verb. Did you mean that my criteria aren't applicable...or that my 
criterion isn't applicable...?

But you forgot to share with us what practical voting system you believe 
that one or more of my criteria isn't applicable to.

By the way, if you'd turn down the volume, I'd reply to you more often. But 
you're consistently such a loud twit, always spouting off with no clue about 
what you're saying, that the fact is that when I ignore your postings, no 
one will think that I'm evading a valid criticism.

I'm making an exception this time, because it's worthwhile to clarify this 
matter for others, because this is the first time, in a long time, that 
someone pointed out something unexpected about one of my criteria or their 
supporting definitions, something that could be mistaken for a problem if I 
didn't answer it.

But it is not a failure of that definition. And it is not that the 
definition doesn't apply to that voter or wrongly characterizes her voting. 
It's not unfair to not call a ballot sincere when it falsifies a preference, 
regardless of whether or not it's intentional or the voter's fault.

But, as I said, the matter of fairness to the voter's reputation for honesty 
isn't relevant anyway, because, as I said, that definiition is intended only 
for use with my criteria, not for describing or evalutating people's 
honesty.

My sincerity definition meaningfully describes voting that hasn't been 
coerced by strategic need. The fact that it can fail to find that desirable 
circumstance when the voter's preferences are intransitive doesn't negate 
the definition's value. Especially since the only result is that such an 
example merely isn't looked at by my criteria.

Mike Ossipoff

_________________________________________________________________
Express yourself instantly with MSN Messenger! Download today - it's FREE! 
http://messenger.msn.click-url.com/go/onm00200471ave/direct/01/