Qualities with values

[GoogleCodeExporter]

You always have a blood pressure, but only sometimes have a blood pressure of 
110/70. How is 
one to represent both of these in BFO? 

Original issue reported on code.google.com by alanruttenberg@gmail.com on 25 Jul 2007 at 3:50

• Blocking: #162

[GoogleCodeExporter]

Original comment by alanruttenberg@gmail.com on 13 Jan 2012 at 7:16

• Added labels: Type-BFO2-design
• Removed labels: Type-Defect

[GoogleCodeExporter]

As far as I understand, in the spirit of BFO, 'blood pressure of 110/70' is a 
subtype of 'blood pressure'. There is no distinction between qualities and 
their values (such as quatity regions in DOLCE) 

- Stefan

Original comment by steschu@gmail.com on 16 Jan 2012 at 9:37

[GoogleCodeExporter]

there is an odd situation, in that there is no differentia to explicate the 
difference between 
blood pressure of 110/70 and 
blood pressure of 120/80.

Original comment by alanruttenberg@gmail.com on 16 Jan 2012 at 9:45

[GoogleCodeExporter]

I think there will need to be some kind of mechanism to allow distnguishing 
determinables from determinants, and to allow the differentia to be specified.

Some experience:

PATO was originally DOLCE-like, with separate hierarchies for attributes and 
values. We made it BFO-like in around 2007. This was a bit controversial, and 
the move was unpopular with a lot of users. We ended up allowing the 
distinction via the back door - we created to obo-subsets, one for attributes, 
one for values. These are non-logical annotation assertions in the OWL. This 
placated most, but it's an odd situation having non-logical axioms for 
something that we expect to be logical.

Original comment by cmung...@gmail.com on 16 Jan 2012 at 10:26

[GoogleCodeExporter]

Original comment by alanruttenberg@gmail.com on 7 May 2012 at 2:44

• Added labels: Type-BFO2-Reference
• Removed labels: Type-BFO2-design

[GoogleCodeExporter]

Original comment by alanruttenberg@gmail.com on 8 May 2012 at 4:13

[GoogleCodeExporter]

Original comment by alanruttenberg@gmail.com on 8 May 2012 at 4:37

• Changed state: Review-old-issue

[GoogleCodeExporter]

Original comment by cmung...@gmail.com on 17 Apr 2013 at 3:40

• Now blocking: #162

[GoogleCodeExporter]

My assumption has been that determinables will indeed be distinguished 
logically from determinates. Tentatively:

1. if determinable D inheres in bearer B at t, then determinable D inheres in B 
at all times at which B exists.

1. This does not hold for determinates.
Are there counterexamples to 1.

An alternative, or supplementary, approach might be:

If D is a determinable, then there are qualities D1, D2 such that D1 is_a D and 
D2 is_a D, and for some bearer B, there are distinct times t1 and t2, such D1 
inheres in B at t1 and D2 inheres in B at D2

Original comment by ifo...@gmail.com on 1 May 2013 at 4:54

[GoogleCodeExporter]

"Are there counterexamples to 1."

This depends on what BFO says about values of zero.

If we have a determinable Q, and Q inheres in B at t with some magnitude or 
value "11" on some scale, and Q inheres in B at t with value "0", is there 
really a Q present, or is the bearer left bereft of any Q?

It's hard to come up with examples of zero values. E.g. we may talk of an 
object becoming weightless in deep space, but there is some weight with a very 
low value.

But BFO should have some documented position. Either it's possible for 1. above 
to be contradicted by virtue of values becoming zero OR determinables are still 
present when their value is zero OR physics is such that no true determinable 
can take on a zero value.

Original comment by cmung...@gmail.com on 1 May 2013 at 5:15

[GoogleCodeExporter]

"1. if determinable D inheres in bearer B at t, then determinable D inheres in 
B at all times at which B exists."
This doesn't distinguish determinable versus determinate. It distinguishes 
those quality types which are rigid (or "essential". A counterexample is the 
charge of an electron, which satisfies the above but which is not determinable.

"If D is a determinable, then there are qualities D1, D2 such that D1 is_a D 
and D2 is_a D, and for some bearer B, there are distinct times t1 and t2, such 
D1 inheres in B at t1 and D2 inheres in B at D2"

This formulation mixes use of the symbols D1,D2 as both universals and 
particulars.
Universal: "there are qualities D1, D2 such that D1 is_a D and D2 is_a D"
Particular: "D1 inheres in B at t1 and D2 inheres in B at D2" 

So can not be evaluated.

I don't believe the distinction can be made at the level of particulars. What 
is the case, however, is that 

if q instantiates D at t1 and D is a determinable quality type, then there is 
some subtype of D' of D that is a determinate quality type and q instantiates 
D' at t.

--

Neither solves the question of how to represent these in OWL, nor answers the 
question of what the differentia between the 'type blood pressure of 110/70', 
and it's superclass 'blood pressure'

Original comment by alanruttenberg@gmail.com on 13 May 2013 at 2:39