[Tcs-lc] Concepts of Higher Taxa

[Richard Pyle]

> My understanding of the circumscription of a genus is biased towards the
> species that are included within it and less so on its description and I
> think this is pretty common. As we go up the hierarchy taxa are defined
> more by their members than by anything else. An author's definition of a
> family is usually more or less just a list of genera arranged into
> intermediate taxa of some kind.

I don't see any fundamental difference in concept definitions between
Kingdoms, Phyla, Classes, etc.; and Genera, Species, Varieties, Forms,
Populations, etc.  To me, they are all fundamentally the same.

I see all concept circumscriptions as being *defined* by the set of members
they contain.  Ultimately, this means the set of individual organisms
(living, recently dead, and yet-to-be-born); but lower-rank named
circumscriptions can serve as collections that serve as surrogates for the
set of terminal members they contain.

I view character-based representations not as "definitions" of concept
curcumscriptions, but as a methof used to communicate the intended
boundaries of the circumscriptions.

The problem, of course, is that no human has even known the full scope of
members for any taxon concept -- nor likely ever will (unless a new taxon is
manufactured in a laboratory, and all descendants are cataloged).  But this
isn't a problem only for the member-based circumscription definitions -- it
is just as problematic for the character-based circumscription definitions.
In this case, the sentence above is re-stated as: "no human has even known
the full scope of characters that apply to any taxon concept -- nor likely
ever will".

I prefer the member-based definitions to the character-based definitions,
because in most cases, the members are physical objects (individual
organisms), rather than conceptual constructs (charcters & character
states).

> How is this handled in the Taxon Concept world? If we allow a GUID to be
> created for a definition of genus G1 according to A1 and then some one
> comes along and creates new species level taxon S1 according to A2 and
> says that it has a 'contained in' relationship with G1 according to A1
> does that alter G1 according to A1 concept?

No -- it doesn't alter G1 SEC A1.  G1 SEC A1 is always fixed and unchanging
from the time of its creation into perpetuity.  However, we will rarely know
the full extent of G1 SEC A1.  The guesswork comes in when establishing the
relationship (in TCS, via RelationshipAssertions) between G1 SEC A2 and G1
SEC A1.  The question always boils down to: "Would A1 have included S1 in
his/her concept of G1, if he/she knew about S1?"  In some cases, not even A1
him/herself can answer that question -- which is why it's important to
distinguish definitive "Relationships" (in TCS within TaxonConcept), from
interpretive "RelationshipAssertions".

> If so should we automatically create a G1 according to A2 concept and
> circumscribe it as being the same as G1 according to A1 (a contains
> relationship) but with the addition of S1?

We should certainly create G1 SEC A2, and include as part of its definition
"includes S1".  Beyond that, it depends on how much information A2 gives us
to define the boundaries of the circumscription of the taxon concept to
which he/she applied the name "G1".

> If we do this then the same logic surely applies all the way up the
> chain to the kingdom level or does it stop at one level up?

I think it would be a mistake to change rules of logic for concepts at
different ranks.  Of course, rules of logic for names (and how to correctly
format those names) *do* depend on rank -- but that's a separate issue.

> The same
> arguments perhaps apply below species level with people adjusting
> subspecies definitions.

Should be the same for all concept circumscriptions, no matter how broad or
narrow (i.e., no matter at what rank).

> If we do allow changes in lower taxa to effect upper taxa then we are
> into concept proliferation territory.

...which is why I embrace the "usage=potential concept" approach, with a
subset of the "potential concepts" rising to the level of "well-defined"
(=reusable) concept definitions.

> If we don't then we are not
> reflecting the way people really define taxa. Perhaps we should be just
> thinking of taxa becoming more 'nominal' the higher up the chain we go
> if not we get a new Plantae for every micro species of Rubus that is
> created.

I think we would ultimately regret shifting paradigms for concepts as we
ascend or descend the taxonomic ranks.  Really, ranks apply to names, not
concepts.  The concept exists as a set of members regardless of the rank of
the name(s) that is/are applied to the concepts.

> In short: How do changes to contained taxon concepts affect the concepts
> that hold them?

That really depends on how confidently we can answer the question, "Would A1
have included S1 in his/her concept of G1, if he/she knew about S1?"

> Should we perhaps be talking more in terms of classifications (i.e.
> trees of concepts) rather than billiard ball like concepts? An according
> to actually freezes a tree of objects not a single object.

I don't understand this part.

Aloha,
Rich