[tcs-lc] concepts of Higher taxa

[Richard Pyle]

> > I didn't mean ignore the information altogether -- I meant disregard the
> > Nomenclatural connection.  That means the connection between a
> genus and a
> > family should be established via TCS "Conceptual placement" structure.
>
> We agree on this, but how does the "Conceptual placement" work?

"is included in"/"includes" relationships between two TCS TaxonConcept
instances.

> > So, my vote would be for "special handling" in the form of
> type="implied"
>
> That sounds good to me, except I am not sure where exactly to
> draw the border between defined and implied concepts. That is, I fail to
be able
> to define it in an exact way, but I think we try to.

Agreed -- this is a problem.  But it's the exact same problem that already
exists in TCS when we decide whether something rises to the level of a
"Concept", as opposed to simply a "Usage".  I suppose my approach is
slightly more problematic because there would be two thresholds -- one to
divide "Usage" from "Implied Concept", and the other to divide "Implied
Concept" from "Defined Concept".  I guess my answer would be that the first
division doesn't really exist -- all Usages *could* be assumed to have an
"Implied" Concept.  Whether or not anyone takes the trouble establish a TCS
instance (with GUID) for any given Usage as an "Implied Concept" is
something that will sort itself out.  If someone, somewhere, felt it was
important enough to create it, then it shall be created.

As for the division between "Implied" and "Defined"; I'm not sure we can
come up with purely objective criteria, but there would certainly be some
"litmus test" characteristics.  For example, if all "children" are treated,
then it could be considered to be "defined".

> I think I agree with you regarding the implied concepts, but wouldn't we
> technically have to create a TCS instance for each of these "undefined"
> concepts?

Only when someone feels the need to do so.  For example, in cases where new
species are described and placed in a genus, but the genus TaxonConcept is
not well defined.  The emphasis will, of course, be on creating TCS
instances for fully-defined concepts, and those of nomenclatural importance.
What's the harm in allowing more TaxonConcept instances to be established?
GUIDs are cheap (usually...). They can easily be filtered out.  It should be
easy enough to filter for "All 'Defined' Concepts, plus only those 'Implied'
Concepts that are directly referenced via definitive Relationships of
'Defined' Concepts.

That's a pretty simple logic rule that would intelligently restrict the
number of 'Implied' Concepts returned.  Another simple rule would be to
return only those Implied Concepts that participate in any definitive
Relationship or interpretive RelationshipAssertion.

Allowing for Implied Concepts means that gradually, over time, we can build
a universal index of Usages -- which I think would be a WONDERFUL thing.

> > We may end up with 100s or 1000s of TCS instances (each with a
> unique GUID), but
> > many of these will collapse into "Congruent" Relationships with
> other "Defined"
> > concepts via RelationshipAssertion mappings.
>
> Who will be doing this?

Whoever feels the need to create the Implied TaxonConcept.  The only cost is
the commitment of a GUID.  If nobody takes the time to map an Implied
Concept into the "matrix" of Defined TaxonConcepts, (either via definitive
Relationships, or interpretive RelationshipAssertions), then those Implied
Concepts will be invisible, and therefore cost us all nothing in the form of
"clutter".

> I think this is impractical. Since these implied
> concepts do not carry any information

They carry *SOME* information -- just not enough to fully define a Concept
circumscription.  At the very least, they represent a usage instance
(Name-Author-Date intersection), which might be *very* interesting to future
taxonomists, who may like to know where certain names were used.  In many
cases, they'll carry more information (e.g., the enumeration of several
children concepts, if not all children).  In some cases, they're downright
critical for nomenclatural purposes (e.g., placement of a new species within
an implied Genus COncept that was not fully defined).

> (as you say, we simply do not know which
> concept the author pointing to a family for the new genus meant),

We don't "know" from the information provided, but oftentimes an intelligent
later taxonomist can make a pretty damn good guess -- which is where the
RelationshipAssertion would come into play.

> but only
> existence, it seems to me that the model should be able to handle them
> otherwise, i.e. not having to create "dummy implied concept" instances.

I wouldn't call them "dummy" -- because "Implied" means they were definitely
real -- just not adequately defined.

> It seems to me easier to point these generically to the name
> level rather than to concept record. Pointer from a name or concept to a
name would
> by definition  be "implied, undefined concepts".

Perhaps -- but the part that got me thinking about this is your point about,
e.g., the Family in which a new species is placed. Jones describes Aus bus,
and asserts it as a member of the family Aiidae.  Jones gives a
fully-defined concept circumscription for the new species, but not for the
genus or for the family.  I can imagine two ways of dealing with this
(highly abbreviated TCS, the first example speculating how it might look
with top-level Name objects).

Here is how I imagine it might look with top-level Name objects:

<DataSet>
  <Publications>
    <Publication id="P1">
      <PublicationSimple>Jones, etc...</PublicationSimple>
    </Publication>
  </Publications>
  <TaxonNames>
    <TaxonName id="TN1" NomenclaturalCode="ICZN">
      <TaxonNameSimple>Aiidae</TaxonNameSimple>
      <RankGroup>family</Rank>
      [etc...]
    </TaxonName>
    <TaxonName id="TN2" NomenclaturalCode="ICZN">
      <TaxonNameSimple>Aus</TaxonNameSimple>
      <RankGroup>genus</Rank>
      [etc...]
    </TaxonName>
    <TaxonName id="TN3" NomenclaturalCode="ICZN">
      <TaxonNameSimple>Aus bus</TaxonNameSimple>
      <RankGroup>species</Rank>
      [etc...]
    </TaxonName>
  </TaxonNames>
  <TaxonConcepts>
    <TaxonConcept id="TC1" type="original">
      <Names>
        <Name ref="TN3" rank="species"/>
        <Name ref="TN2" rank="genus"/>
        <Name ref="TN1" rank="family"/>
      </Names>
      <AccordingTo>
        <AccordingToSimple>Jones</AccordingToSimple>
      </AccordingTo>
    </TaxonConcept>
  </TaxonConcepts>
</DataSet>

So...in this case, we've got one TaxonConcept instance, referencing three
names. This would work, but I think it ignores the fact that Jones did
indeed have some concept in mind for the genus Aus, and the family Aiidae,
but simply failed to document the boundaries of that concept circumscription
adequately.

A problem with this paradigm is that, suppose Jones mentioned in his
description "Similar to Aus cus."  How would we capture the fact that Jones
considered this other species as valid, and was "excluded" from his new Aus
bus?  Without Implied TaxonConcept instances, we can't (as far as I can
tell).

Now, consider this alternative approach, more similar to current TCS, except
with "Implied" TaxonConcepts, and treating Nominal-type TaxonConcepts as the
bearers of LC/Name data:

<DataSet>
  <Publications>
    <Publication id="P1">
      <PublicationSimple>Jones, etc...</PublicationSimple>
    </Publication>
  </Publications>
  <TaxonConcepts>
    <TaxonConcept id="TCn1" type="nominal">
      <Name>
        <NameSimple>Aiidae</NameSimple>
        <NameDetailed>
          [Full LC structure for name "Aiidae"]
        </NameDetailed>
      </Name>
    </TaxonConcept>
    <TaxonConcept id="TCn2" type="nominal">
      <Name>
        <NameSimple>Aus</NameSimple>
        <NameDetailed>
          [Full LC structure for name "Aus"]
        </NameDetailed>
      </Name>
      <Relationships>
        <Relationship type="has type">
          <ToTaxonConcept ref="TCn4">
        </Relationship>
      </Relationships>
    </TaxonConcept>
    <TaxonConcept id="TCn3" type="nominal">
      <Name>
        <NameSimple>Aus bus</NameSimple>
        <NameDetailed>
          [Full LC structure for name "Aus bus"]
        </NameDetailed>
      </Name>
      <Relationships>
        <Relationship type="original genus combination">
          <ToTaxonConcept ref="TCn2">
        </Relationship>
      </Relationships>
    </TaxonConcept>
    <TaxonConcept id="TCn4" type="nominal">
      <Name>
        <NameSimple>Aus cus</NameSimple>
        <NameDetailed>
          [Full LC structure for name "Aus cus"]
        </NameDetailed>
      </Name>
      <Relationships>
        <Relationship type="typifies">
          <ToTaxonConcept ref="TCn2">
        </Relationship>
      </Relationships>
    </TaxonConcept>
    <TaxonConcept id="TCi1" type="implied">
      <Name>
        <NameSimple>Aiidae</NameSimple>
        <NameDetailed ref="TCn1"/>
      </Name>
      <AccordingTo>
        <AccordingToSimple>Jones</AccordingToSimple>
      </AccordingTo>
      <Relationships>
        <Relationship type="includes">
          <ToTaxonConcept ref="TCi2">
        </Relationship>
      </Relationships>
    </TaxonConcept>
    <TaxonConcept id="TCi2" type="implied">
      <Name>
        <NameSimple>Aus</NameSimple>
        <NameDetailed ref="TCn2"/>
      </Name>
      <AccordingTo>
        <AccordingToSimple>Jones</AccordingToSimple>
      </AccordingTo>
      <Relationships>
        <Relationship type="is included in">
          <ToTaxonConcept ref="TCi1">
        </Relationship>
        <Relationship type="includes">
          <ToTaxonConcept ref="TCd3">
        </Relationship>
        <Relationship type="includes">
          <ToTaxonConcept ref="TCi4">
        </Relationship>
      </Relationships>
    </TaxonConcept>
    <TaxonConcept id="TCd3" type="defined">
      <Name>
        <NameSimple>Aus bus</NameSimple>
        <NameDetailed ref="TCn3"/>
      </Name>
      <AccordingTo>
        <AccordingToSimple>Jones</AccordingToSimple>
      </AccordingTo>
      <Relationships>
        <Relationship type="is included in">
          <ToTaxonConcept ref="TCi2">
        </Relationship>
      <Relationships>
      <SpecimenCircumscription>
        [Etc...]
      </SpecimenCircumscription>
      <CharacterCircumscription>
        [Etc...]
      </CharacterCircumscription>
    </TaxonConcept>
    <TaxonConcept id="TCi4" type="implied">
      <Name>
        <NameSimple>Aus cus</NameSimple>
        <NameDetailed ref="TCn4"/>
      </Name>
      <AccordingTo>
        <AccordingToSimple>Jones</AccordingToSimple>
      </AccordingTo>
      <Relationships>
        <Relationship type="is included in">
          <ToTaxonConcept ref="TCi2">
        </Relationship>
      <Relationships>
    </TaxonConcept>
  </TaxonConcepts>
</DataSet>

Summary:

Four Nominal TaxonConcepts:
- Aiidae (TCn1)
- Aus (TCn2)
- Aus bus (TCn3)
- Aus cus (TCn4)

Three Implied TaxonConcepts:
- Aiidae SEC. Jones (TCi1)
- Aus SEC. Jones (TCi2)
- Aus cus SEC. Jones (TCi4)

One Defined TaxonConcept:
- Aus bus SEC. Jones (TCd3)

I threw some other stuff in there, like some examples of name-name
relationships.

> Gruss nach Hawai!

Huh?

Aloha,
Rich