Seeing the differences among position drafts

[Michael Allan]

David Hilvert wrote:
> One thing that conventional polls do well that I have not seen expressed very
> clearly in descriptions of Votorola is the ability for a single individual to
> [a] propose a variety of options, [b] one of which is to be selected.

We support [a] but not [b].  We support [a] by nomination, where a
voter drafts a position statement for a candidate, and then solicits
votes for that position.  Multiple choices may be created in this
fashion, because a single individual may nominate multiple candidates.
(A separate candidate is always required for each choice.)  For
example, most of the candidates in P/BGE are Thomas's nominees:
http://t.zelea.com/wiki/P/BGE

For an English example, I nominated both myself and Fred in P/grfin:
http://t.zelea.com/wiki/P/grfin

We don't support [b], at least not in the restrictive sense of
limiting the voters' choices.  Voters are always free to nominate
other candidates, and thus extend the range of choices.  Existing
choices are also liable to change, because candidates are free to
alter their postitions, e.g. by editing their position drafts.
 
> As an example, if a design decision is to be made (e.g., in a Free Software
> project), this might normally occur via discussion, but each participant need
> not express only a single course of action; rather, multiple possible courses
> could be outlined.  One way to agree with such an expression would be to select
> a single one of these; another would be to agree with all or with some subset.

Different perspectives on agreeing "with all or with some subset":

  a) Cascading votes
  b) Multiple or fractional votes
  c) Composition or linking of positions

(a) A single vote may assent to multiple candidates through cascading.
Here for example, both C and D receive the vote of A, so A is voting
for both C and D:

  A -> C -> D

(b) Thomas has often suggested allowing for multiple or fractional
votes.  The design could accomodate this if necessary.  But the use
cases aren't clear yet.

(c) In normative voting (imagine a municipal development plan), it is
possible for a single position to be composed of multiple, other
positions.  The other positions are either incorporated directly into
the composite draft by copying, or they are referenced by links.  So a
candidate may compose combinations of simpler choices, and solicit
votes for them.

But I imagine that the simpler choices will typically be voted in
separate polls.  So A's park improvement plan has links to B's sandbox
plan and C's running track plan, each of which it incorporates.  But
park, sandbox and running track are all voted in separate polls.
Another complexity is that all of the plans (wholes and parts) are
governed by municipal and provincial policies, which are voted in yet
other polls.  So A and her voters have to navigate all this
complexity, and arrive at the best possible solution.

> Indeed, if a voting system is to be able to express agreement with a variety of
> kinds of proposals or assessments, then it should probably take into account
> the case that a proposal is not yet fully refined.  I believe that this should
> be possible within a system like Votorola, and may be one area where Votorola
> could offer significant advantages over conventional Web polls.

I think so, too.  Because the first purpose of voting is to guide the
composition of the proposals.  It may happen that people use their
votes to compose a single consensus, but they may just as likely
compose a split dissensus of three contradictory proposals.  In the
latter case, it is not the purpose of the voting system to force a
narrower decision.

People may need a little longer to understand their differences and
compose a resolution.  So the dissensus may indicate nothing more than
a need of refinement, as you say.  Or maybe the effort has unearthed a
deeper set of problems that cannot be resolved in a single poll.  In
either case, (to come back to this) it's the textual differences that
reveal the true shape of the underlying social differences.

-- 
Michael Allan

Toronto, 647-436-4521
Skype michael_c_allan
http://zelea.com/