[sc34wg3] Assertion Type

[Patrick Durusau]

Greetings!

I have been replaying the TMRM discussion in Amsterdam looking for
places where I think that terminology differences have prevented us
from having useful discussions. One of those places is the term
"assertion type." I say that because when we said that one could have
the TMDM and a different "assertion type," there were several who said
that would not be a topic map.

Let me be quick to point out that an assertion in the TMRM, is the
same thing as an association in the TMDM. (Well, one is more
restricted than the other in terms of the amount of information that
can be attached to its components, but there is no semantic conflict
between them.)

All an assertion does in the TMRM is document a relationship between
topics. Granted it does so with a great deal of particularity and
we can argue about the details but first let's clear up what Newcomb
and I mean by an "assertion type."

In the TMRM, an "assertion type" is the combination of all the roles
in an assertion. They collectively identify the type of relationship
that the assertion represents.

For example, if we have a marriage assertion (Western, 20th century),
it has two roles, that of husband and wife.

The reason why we say that is the "type" of the assertion is that in
TMRM reasoning, roles are unique to a single assertion type.

That is to say, only one assertion type (here, the "marriage"
assertion type) can have any given role (here, the "husband" role),

However, there can be *another* role, also *called* "husband", that
appears in another assertion type, but the two "husband" roles have
different subjects, by definition.  The two "husband" role topics can
never merge, because they appear in different types of assertions.

This rule may seem strange, but it follows from the rule that a topic
can have only one subject. A topic that is a role in an assertion, has
a subject that is defined by that assertion. That is to say that when
a relationship is described between topics, each of the roles in that
description is a particular subject. And, that subject can only exist
in the relationship being described.

Think about it for a moment. Can you consider saying that someone is a
husband or wife in the absence of thinking of a marriage assertion in
which they have those roles? Note that we may not know who their
spouse is (role player) but that is not required. It is enough that we
know the subject that is the role.

Another way to say that is that every role in a relationship is bound
up in the definition of the other roles in the same relationship. The
idea of a husband, for example, makes no sense separate from the idea
of a wife.

That is not to say that the TMRM compells good design or even attempts
to control what design is put into place. Even though roles in
different assertions have by defintion different subjects, the TMRM
does not compell that they have diferent names or particular
properties. What it does make clear is that roles in different
assertions represent different subjects and therefore they cannot
merge.

In other words, "assertion type" does not mean to add or take anything
away from the topic, association, occurrence model that underlies the
TMDM. As the TMDM notes, an occurrence is a particular "type" of
association. The TMRM allows the various "types" of assertions in a
topic map design to be disclosed, along with a checklist for what
happens depending upon design choices for the topic map.

Does that help?

Hope everyone is having a great day!

Patrick

-- 
Patrick Durusau
Director of Research and Development
Society of Biblical Literature
Patrick.Durusau@sbl-site.org
Chair, V1 - Text Processing: Office and Publishing Systems Interface
Co-Editor, ISO 13250, Topic Maps -- Reference Model

Topic Maps: Human, not artificial, intelligence at work!