[tdwg-content] New Darwin Core terms proposed relating to material samples

Richard Pyle deepreef at bishopmuseum.org
Mon May 27 22:37:14 CEST 2013

Thanks, Steve.  I'll keep my responses as minimal as possible.

> I think that pretty much everybody agrees that "individual" 
> is a confusing term name for a number of reasons.  

And I agree as well.  I hope it's clear that we latched on to that term only
because, at the time, it was the closest term to what we needed, and in many
cases it's better to stick with an existing (even if potentially confusing)
term than it is to invent a new (but nearly identical) term -- which risks
creating even more confusion.  As I have said, I think we should sort out
the concepts first, then we should debate about the appropriate terms to
label the concepts.

> Although there is potentially significant overlap between the 
> proposed dwc:MaterialSample class and Individual, I think that 
> there are at least two ways that they differ significantly.  One 
> is that I'm pretty sure that there is no requirement that a 
> dwc:MaterialSample must be a biological material 
> (i.e. derived from a living thing).  

In my mind, this doesn't really count as a "difference", because in our
model, an "Individual" does not need to be biological material either.
However, I concede that this is a bastardization of the original intent of
dwc:individualID, so I'm ready to completely abandon the term "individual".
My real concern is that we do not maintain parallel and largely overlapping
terms.  Following your suggestion, I would therefore recommend that we move
towards deprecating dwc:individualID, and do one of the following:

1) Replace it with dwc:materialSampleID and establish a new materialSample
class; or
2) Replace it with [someOtherLabel]ID and establish a new someOtherLabel

But as I keep saying, the most important thing I think we need to discuss is
whether the original intent of dwc:individualID:

"An identifier for an individual or named group of individual organisms
represented in the Occurrence. Meant to accommodate resampling of the same
individual or group for monitoring purposes. May be a global unique
identifier or an identifier specific to a data set." 

...encompasses (or should be redefined to encompass) what is needed to
accommodate the needs of the proposed dwc:materialSample:

"The category of information pertaining to the physical results of a
sampling (or subsampling) event. In biological collections, the material
sample is typically collected, and either preserved or destructively

Are we better served by a single class of thing that includes both
definitions (with defined subclasses as needed)?  Or are we better off with
two completely separate classes of things?  For reasons that would require
far too much text to describe here, I strongly favor the former.  

I can't express a preference for how best to label this concept (or these
separate concepts) until I know what the concepts are.

> I think that it's pretty clear from what Rich has said that Individual 
> (to include the range from tissue samples up to herds) must consist of
biological materials.  

No, that's not correct. In our model, "Individual" includes things that are
decidedly *not* biological.  Biological things are a subset of instances of
our "Individual".

> The other is that a material sample must be physically sampled (i.e.
> from the environment and subjected to some kind of processing).  An
> feature of an Individual (at least to me!) is that it can be observed,
> or recorded without necessarily having all or part of it being removed
from its environment 
> and subjected to processing.  

Yes -- exactly.  I think we need a superclass that can be applied physical
objects (either biological, or non-biological).  A subset of these things
are biological.  Another subset of these things are extracted from nature.
Another subset is subsampled and used for some sort of destructive or
non-destructive analysis.

> There is nothing in the current definitions of the type vocabulary terms
that require its 
> classes to be disjoint.  I think it is possible that something could be
both a 
> dwctype:PreservedSpecimen and a dwctype:FossilSpecimen, and if 
> dwctype:MaterialSample is accepted as a term there would undoubtedly be
> that were both dwctype:PreservedSpecimen and dwctype:MaterialSample.  So I
don't think 
> it is necessarily a problem if there is overlap between
dwctype:MaterialSample and an Individual class.  
> Certainly RDF allows a resource to have two (or more) rdf:type

Sure -- but we are setting ourselves up for chaos if we leave it open for
individual providers to apply one class or another to the same physical


More information about the tdwg-content mailing list