Re: [ba-unrev-talk] Icons for IBIS
Jeff Conklin wrote (Friday, April 12, 2002 3:21 AM)
> As long as the statements inside an argument node are
> unconstrained natural language narratives, the "formal" link or node type
> you apply is really just icing for the people who are reading the map --
> they'll form their own cognitive links as they interpret the texts in the
> nodes, and hopefully the explicit link semantics won't confuse them or get
> in the way too much. I may sound cynical here, but I am feeling a hard
won
> passion about putting human communication before computable
> structures. This feels like an incomplete thought but it's late and ...
> maybe it's close enough. (01)
Jeff, I wholeheartedly agree. The semantic language, or coding of nodes and
links, can be useful to groups early on in a workshop, but as familiarity
builds participants rely on their visual reasoning and visual memory to
interpret the individual nodes and map's structure. The notation is
important to us as facilitators to build a map with a structure, but we
shouldn't get bogged down in purity! Any sensible coding of nodes will do
provided it doesn't disrupt their own cognition. What is important is the
map as memory aid, its use as a negotiative/dialectical device to structure
the dialogue (a transitional object), projection of it to develop shared
space and the facilitation. I am using maps though in face-to-face meetings
and which don't have a long shelflife. (02)
Eric Armstrong wrote (Friday, April 12, 2002 3:23 AM)
> So I think it is totally fair not to bother with reusable nodes and
maleable
> roles at this point in time. Instead, I would plan to cut and paste the
text
> "X costs $10,000" into a pro node one time, and into a con node some
> other time. (03)
In my own modelling, using cognitive mapping rather than IBIS, I rely
heavily on transclusive links. In workshops I often quickly scan with
participants the other maps/contexts where an issue arises and to sense
check that the material around other copies shouldn't be included in the
current conversation. I also use it when working with the maps after the
workshops with the client, or individual participants, when we're enacting
upon agreed actions/options. I'm often working with 600 nodes across 30
screens where some issues are repeated in 5 or 6 maps, I guess with smaller
models transclusive linking is less important, but I wouldn't want to see
them marginalised in IBIS. As a hunch, I think I most transclusive links as
follows: 1) when having agreed key issues, exploring each of these in depth
by building new maps but reusing part of the previous maps; 2) after having
agreed what needs to be done, building a new map of ideas on how to do it
(again reusing); 3) having agreed how, building a map on implementation,
including the nitty gritty actions, names and time scales against each
action. All problem solving is iterative and when going through these cycles
I often cycle back to previous maps and want to pull nodes in and push old
ones out. (04)
With regards node types, I often find even in the same map, let alone after
making a transclusive copy, that an arguement node is a pro for one idea
and a con for another. My preference is to display all arguements in plain
type and to show their semantic type only through the link using a plus or
negative sign near the head of the arrow. Participants seem comfortable with
this, probably because of Jeff's point, that they do the semantic
interpretation themselves. (05)
I think this is my first post to this list, and I could go on replying to an
accumulation of thoughts relating to some of the IBIS discussions that have
taken place here over the past few months. I shall endeavour instead to
respond more often in the future and include them there. (06)
Mark
**I'm not an outlier, I just haven't found my distribution yet** (07)
Mark Westcombe
Management Science
The Management School
Lancaster University
LANCASTER
GB LA1 4YX (08)
Office: +44 1524 593635
Home: +44 1524 64106 (09)
_________________________________________________________
Do You Yahoo!?
Get your free @yahoo.com address at http://mail.yahoo.com
(010)