Condorcet is not infallible

[Mike Ossipoff]

Lucien Saumur writes:
> 
>           I detect some impatience and even some anger in
> this mailing list.

Maybe a little impatience, because I've answered the same objections
over & over again. But other than that I always maintain politeness,
though I admit that I can finally start losing patience when answering
the same objection for the zillionth time.

> 
>           I do not consider Condorcet as an infallible
> prophet. Condorcet has written many things about elections

Did I claim that Mr. Condorcet was a prophet, and that you should
adopt his method on anyone's authority? Or did I show you why
Condorcet's method is the one that complies with the standards
that are important to voters & electoral reformers?

> and about other things with which it is permissible to
> disagree. Condorcet has written that candidates should not

Disagree by all means, but it's better to back up one's disagreements
with arguments about specific properties & standards, rather than
with claims that you're being asked to accept Condorcet as a prophet.

> campaign so as not to influence the voters. Would we agree
> with that today?

That would be an interesting campaign. I wonder what it would
sound like. It has appeal, but I'd settle for giving all points
of view airtime in proportion to some measure of their public
support. Obviously any rules about not trying to influence voters
improperly will be misused by those in power. Maybe Condorcet
meant that ideally there wouldn't be need for persuasion, if
someone were honest & trunstworthy to present voting choice
information in an objective way. Not likely in our society,
but a nice ideal. I agree with it as an ideal, but not as
anything feasible in our society.

> 
>           I agree that preference voting and pairwise
> comparison is the basis of the best electoral system, most
> of the time. I do not think that Condorcet has said the
> last word on every aspect of this system and I think that

Did I ask you to take Condorcet's word, or my word, for anything?
I backed up my statements, and if there's one that I didn't back
up, I'd like to know about it, and then I will back it up.

> we should try to find ways how this system may best be
> applied in our modern world.

You're referring to the system of Pairwise-Count. I thought that
_is_ what we're doing--proposing ways of implementing Pairwise-Count.
I've talked about the faults of your solution-by-lot. I've talked
about the problem of Bruce's Regular-Champion, & Copeland's method
in general. I've talked about how Condorcet's method, the version
of his method that I've been proposng & defining, best meets the
standards that led us to want single-winner reform, the standards
that are important to voters & electoral reformers.

Look, anyone can name some standard and say "Methods should meet
this standard." Fine. Who can disagree. If that's the standard
that you like, and if your method meets it, then, for you,
your method may be the best. The important thing for us to be
doing, to bring some direction to this discussion (we talked about
this need long ago; this isn't something new) is to state what
specific standards are met by the methods we propose. And/or what
criteria they meet. Then say something about why you consider that
standard the important one. I've done that. Just saying "I don't
like your proposal--It should be done as in my proposal", that
just don't do the job.

Funny that you should suggest that I'm asking you to accept 
Condorcet's method on authority--I'm the one who's been talking
about properties, standards & criteria. That has been the basis
of my claim that Condorcet's method is the best proposal to
recommend to the ER to propose to the public.


Mike Ossipoff



> 
>           I am prepared to discuss electoral systems calmly
> with interested parties and I hope that I can disagree with
> others without being labelled a heretic.


> 
> __________________________________________
>           aa447 at FreeNet.Carleton.CA
>           http://www.igs.net/~lsaumur/
> .-
> 


--