Propose sample concept

[rartino]

Based on discussions with @CasperWA, @francescalb

[rartino]

Based on the discussion today, I've edited the sample concept to be the more concrete representation of the physical entity itself.

However, I don't see how one can draw 'is-a' relationships from this sample concept to model-type concepts, e.g., "Periodic crystal", it makes no sense to me. As @jamesrhester also seems to suggest above, the relation would then have to be, e.g., "is-modelled-by".

I may try to suggest a different concept to represent the "model view" of the thing being measure I was trying to explain in the discussions today.

[jamesrhester]

I like these changes. It seems that this concept is much more general than just crystallography, so maybe it is already present in another EMMO ontology in a way that is suitable for us?

[rartino]

From the discussion last meeting: do we want to divide this into 'specimen' and 'sample', where a 'sample' is a 'specimen' but also a representative part of something larger? Or should we simply note that 'specimen' is synonymous to 'sample'?

[jamesrhester]

I think we definitely need to be able to express the relationship that "the thing that was measured" is meant to be representative of a larger class of objects. So my measurement of a crystal of some compound predicts results for all measurements of crystals from this batch, and morever of this compound when prepared by others in the same way. This however seems like a basic scientific assumption (invariance of physical laws in space and time) so maybe is already in EMMO?

There is a difference I assume in materials science where you might take multiple samples from a larger object to get a representative idea of the average properties. In this case you are not assuming that a single sample tells you about the whole object.

[rartino]

My impression from @emanueleghedini is that there indeed are concepts for the ideas of "temporal part" and "spatial part" (or what the name was) in the EMMO that we could just use for this. However, if the crystallography domain has specific words with well established meaning for these, we probably want to include those concepts and define them via the EMMO concepts.

However, when it comes to "sample" as opposed to, e.g., "specimen", I'm not so sure if these terms are that well separated in practical use. I think someone synthesizing "a single thing" and then measuring on it would easily refer to it as "the sample" rather than "the specimen", even though in this case it isn't to anyone's knowledge representative of anything but itself.

That said, perhaps we want to note here in the sample concept definition all these different usages of "sample", i.e., as "from a larger batch of crystals", "from a larger object meant to represent the object", and "one of multiple spatial parts of an object meant to be used in aggregate to determine an average".