validation - dangling s-nodes
Matt South
This list is very quiet, which might lead the casual observer to think
that this subject had dried up. I dont believe that's the case at
all. Perhaps I can generate some discussion with a question.
It is clear from the abstract definition of an argument network that i-
nodes cannot be directly connected to i-nodes. This constraint can
easily be built into a validation rule for a passed argument network
(in whichever syntax used).
But what about s-nodes? I think an s-node should always have at least
one node coming into it and one node going out of it. If this
heuristic is correct then this provides another useful validation for
an abstract argument network.
However, in our discussions in the lab we have come across a situation
where an argument network is being generated as part of a dialetical
process and we simply have an s-node joined to an i-node. The
"dangling" s-node is like an invitation for another dialogue
participant to provide support for the i-node's claim using the
dangling scheme. My belief that s-nodes should have at least one
incoming and outgoign node means that this graph doesnt smell right to
me. I think the invitation to provide supporting arguments should be
somewhere else - in the context or in the protocol perhaps. But I
thought I'd canvas opinions on this subject. Do you think my
validation of an s-node is correct. Are there other schemes whose
corresponding s-nodes lack incoming or outgoing nodes?
regards,
Matt