[EM] (Fwd) [SW] Outline-driven group writing

[Steve Eppley]

I'm forwarding two messages to election-methods.  This is the first.
I sent it to Mike O. before the Single-Winner Committee reconvened in
the election-methods list.  The second message is Mike's reply,
following in a moment.

The subject is how we as a group can produce an educational document 
on electoral problems and reforms.

--Steve

------- Forwarded Message Follows -------
From:          Steve Eppley
To:            Mike Ossipoff
Subject:       [SWC] Outline-driven group writing
Date:          Fri, 9 Feb 1996

Mike, 

I wrote the following intending to send it to the SWC but decided to
send it first only to you.  I think substantial portions of our last
few messages may be of interest to the entire committee too.

--Steve
---------------------------------------------------------------------
Keywords: outline group writing document faq 
          hierarchy problem dependency hypertext html

Contents:
1. Why outline before discussion
2. Document to be intentionally left incomplete--
   Participants write only the parts of interest
3. A general problem with outline hierarchies
--------------------------------------------------------------
1. Why outline before discussion

Mike O. wrote:
>After this committee completes its task, recommending on the
>best method to propose, and reporting arguments, conclusions
>& reasons for voting choices--and after we subsequently write-
>up & post our FAQ, which will thoroughly document how we arrived at
>our recommendation
[snip]

I see the task of this committee as producing a good faq.  The faq 
is not a chore to be done later.  Our straw vote is insignificant.

I'm hopeful that if we attempt to keep the discussion targeted toward
a well thought out outline, the tidying up needed to merge the messages 
into a faq will be quick.  My read of human nature is that people
will be anxious to move on after we finish the body of our work, and
won't have much time or motivation at that point.  If we take the
path of least resistance--unstructured discussion--it will be a
large chore to convert the folder of messages into a readable faq.  

That's why I think it's important to try to achieve consensus on the
outline before we tackle MVP, argue whether we should dismiss all but
non-ranked methods, or discuss who will work on what.  So far, our
messages about what to discuss have included pros and cons of methods
and categories of methods.  (The future equivalent of those messages
about what to discuss will be messages about which pieces of the
outline are worth concentrating on, and which of us will volunteer to
do so.)  These arguments belong in the faq, and I think having a
comprehensive outline will make that clear.

It's not essential that the outline be finalized before we begin
working on its body.  It's easy to add, move, and delete sections
and subsections.  But we should see if we can agree on an overall
framework.


Will most of our work be pigeonholable using the format shown here,
which categorizes methods based on how much information is provided
by the voters as input to the algorithm?
   ...x  <category x> -- Each voter may ... 
   ...x.1  Pros of all algorithms in this category 
   ...x.2  Cons of all algorithms in this category 
   ...x.3  Algorithms for determination of winner 
   ...x.3.y <algorithm y> 
   ...x.3.y.1  Pros 
   ...x.3.y.2  Cons 
I'd appreciate your criticisms of this.  


To illustrate my points: if we strove first to make a reasonably
comprehensive outline, we might begin discussion armed with something
along the lines of the following.  (I've collapsed some of it to save
my time.  I've included a few pros and cons which have been discussed
so often that they inspire our work, but normally pros and cons would 
be developed during the course of our work.)