Using 'Domain' and 'Range' to describe ternary temporal relations is confusing or wrong

[GoogleCodeExporter]

We have many parts of the reference that are of the following sort:

3.1.1   The continuant_part_of relation
ELUCIDATION: b continuant_part_of c at t =Def. b is a part of c at t  & t is a 
temporal region & b and c are continuants. [002-001]
DOMAIN: continuant
RANGE: continuant
(The range for ‘t’ (as in all cases throughout this document unless 
otherwise specified) is: temporal region.)

I understand domain and range of a function, or of a binary relation. Are these 
even defined for ternary relations? If so, what is the definition of Domain and 
Range?

Original issue reported on code.google.com by alanruttenberg@gmail.com on 3 Feb 2013 at 11:05

[GoogleCodeExporter]

This should all be in the clif for BFO.

Some of the work I did on formalizing type-type relations could be salvaged 
here. See

https://obo.svn.sourceforge.net/svnroot/obo/obolog/trunk/obolog.kif

Relevant section:

{{{
(metarelation range)
(text_definition range "Constrains relations such that the object (second 
argument) of the relation only holds between instances of the specified type")

;; @Axiom: domain constraints on time-indexed relations
(=> (and (domain ?rel ?D)
         (?rel ?i1 ?i2 ?t))
    (instance_of ?i2 ?D ?t))
;; @Axiom: domain constraints on atemporal relations
(=> (and (domain ?rel ?D)
         (?rel ?i1 ?i2))
    (instance_of ?i2 ?D))

}}}

Note that you need to do a similar thing for transitivity. 

Original comment by cmung...@gmail.com on 3 Feb 2013 at 11:48