Differenciate between measured, simulated and attributed values

[webhart]

We treat physical quanities as GDCs, which they are in a sense. One meter is a meter, regardless if it expresses the length quality of the Prototype of the Metre in Paris or the result of a computation that computes the distance light traveled in 1/299 792 458 second.

Or more practically: has the value of a materials quality been measured, simulated or attributed from a measurement of a sibling?

[thnlrd]

TL;DR: Attributed values do not scale well, meaning that even measured specimens only have attributed values.

Regarding attributed values, in general I don’t disagree, but this is overkill since any type of measurement is always limited in terms of time, space, scope, etc. (any dimension you can think of, really), and so no measurement will ever hold true outside of its bounds and will immediately become an attributed value.
For example, a measured chemical composition will only be confined to a minimal volume (depending on the measurement method) and will logically be attributed to the rest of the specimen.
The results of a tensile test only apply to the small portion of matter that ruptured.
Therefore, the statement 'quality X of this specimen was measured and resulted in value Y; therefore, the quality of the specimen is Y' would be incorrect, since the measurement method is inherently limited to only portions of the specimen.

Ultimately, it's a granularity problem, which causes most of our measurements to fall apart if we zoom in closer.

[webhart]

Ultimately, it's a granularity problem, which causes most of our measurements to fall apart if we zoom in closer.

Granularity is a pitfall. Everything in materials science and engineering is something else if you zoom in. In our knowledge graph we make assertions that hold for a whole entity. If our entity is an ingot and we measure it's chemical composition, then it holds for the whole ingot. If our entity is the head of the ingot (bfo:fiat object part) and we measure it, it holds for the head of the ingot. Depending on context, we may have an attribution process that allows us to attribute it to the whole ingot or not.

If you don't want to or don't know how to motivate your assertion, you can always make a dogmatic assertion, e.g. using obi:specifies value of. If you want to or know how to motivate your assertion, you should talk about the attribution process that led to the assertion and use pmd:specifies attributed value of which is a owl:SubPropertyOf obi:specifies value of and thus implies dogmatic assertion.