[EM] Nomenclature

[Bart Ingles]

Didn't mean to set myself up as the InterNic of voting system
nomenclature, but it was a concern I had.  If that caused a
misunderstanding over the term "approval" then at least we know what to
watch out for in the future.  Or maybe there should be a disclaimer
following any use of the word approval.  :)

Not that I haven't done the same thing.  I coined the term "Proportional
Approval" (probably about the 100th person to do so) on this list about
2-1/2 years ago, for an unranked method using STV-style quotas.  I think
at least two others have reused the term since then -- in fact you may
have been one of them.

I'm not so worried about approval's prestige, but one reason I'm
becoming more cautious is that of all the methods discussed here,
approval voting is one of the closest "to market", with a track record
and a widely recognized body of research behind it.  A naming mix-up
outside this list would be harder to clear up.

It looks as though there were possibly two misunderstandings with Mr.
Ketchum -- the definition of approval voting, and also the
summability/consistency mix-up.  Maybe I just need more sleep.  Good
night...




Forest Simmons wrote:
> 
> I should have named it Dyadic Pass/Fail.  Perhaps it's not too late to
> change the name.  Of course, I have no control over Martin Harper's
> nomenclature.  He showed that Universal Approval is a genuine common
> generalization of Approval and Condorcet.  Perhaps he would consider
> changing the name to Universal Condorcet :-)
> 
> I suspect that he chose UA because he likes Approval better than
> Condorcet.
> 
> The dyadic ballots are indeed iterated relative approval ballots; the
> question that you must answer repeatedly when filling out a dyadic ballot
> is, "Of the candidates in this subset, which would you approve if they
> were the only candidates in the race?"  I think that's why Martin and I
> both felt that "approval" fit naturally into the names of our methods
> based on dyadic ballots.
> 
> Furthermore, dyadic ballots are the natural ballot type for simulating
> Approval Runoff, or should we call it Pass/Fail runoff.
> 
> I don't see how these names detract from the prestige of standard
> Approval.
> 
> On the contrary, they set Approval up on a pedestal as the standard worthy
> of generalization, the state of the art to try and beat, etc. They bring
> needed attention to under publicized Approval.
> 
> When I named "Dyadic Approval" I did so out of admiration of Approval,
> having taken it as my main source of inspiration, with Condorcet as
> secondary.
> 
> In mathematics we don't have a tradition of copyrighting the names of
> mathematical objects, so the inventors feel free to use any name that
> seems suitable. (Hence, "imaginary" and "irrational" numbers, as well as
> "improper" sets and fractions.)
> 
> Since election methods are on the borderline of math and political
> science, perhaps we should be more cautious.
> 
> I welcome suggestions for improved terminology.
> 
> All in Peace,
> 
> Forest