[Tcs-lc] Taxonomic order anyone?
Greg Whitbread
ghw at anbg.gov.au
Tue Mar 29 19:34:10 PST 2005
Roger,
Having read through this discussion and the posts on "Concepts of Higher
taxa" I must admit to being a little puzzled.
Firstly if one steps back and takes a broader view of these threads they
appear to be one and the same. These issues are so fundamental to what
we are trying to achieve that they are at the very heart of what TCS is
about - and is providing. It is important that we understand that it is
not only concepts that are "sensu" but also the relationships between
them. When an authority publishes a treatment it is, if sometimes
implicitly, presented as a classification. These classifications are
mere fragments. They exist in plethora. And as we start to document
relationships between concepts, well, we will be faced with the wonder
of an infinity of intersecting hierarchies and the need for a very
generic Concept Schema.
There are few high level treatments arranging nominal concepts for our
convenience but like the rest of us you will have to build yet another
"sensu GBIF" view of that space for the purpose of defining your
product.
Secondly. These hierarchical relationships between concepts are not only
fragmentary but disrespectful of attempts to categorize relationships as
either nomenclatural or taxonomic. It is all about concepts. In our
system (APNI) we separated the classification trees from other kinds of
relationships because they were (have been) one to many relationships
and it is convenient at the application level to treat them as such.
However a generic solution such as TCS needs, as we have heard, to deal
with ambiregnal taxa, multiple relationships between the same two
concepts (a Genus and its type species ) and relationships (through
concepts) into other taxonomies. As long as we can categorize
relationships as either being within a classification (intra) or between
(inter) classifications, and differentiate between associations that are
either nomenclatural or taxonomic, our Schema will be both useful and
re-usable.
This thread, I think, provides a very solid argument for embedding
nomenclatural relationships within the concept tree.
You will still have the problem of ordering sibling concepts but here
there are many possibilities: alphabetically (very common); page number;
GUID; etc. We have found that it is not too difficult to reconstruct a
reasonable facsimile - or better still produce one afresh.
Some time ago I made the suggestion of a nomenclatural change from
"taxonConcept" to "Concept" - Roger suggested "Taxon" - because I would
also like to use TCS to document concepts such as "White-Box woodland",
in part defined in terms of included taxa. We might stretch the "taxon"
concept to cover the elements within a classification of ecological
communities but the possibilities are endless? It is relationships
between concepts that carry the labels 'nomenclatural' and 'taxonomic'
and it is from these relationships, and the classification of them, that
we derive our taxonomic and nomenclatural systems.
greg
On Tue, 2005-03-29 at 23:36, Roger Hyam wrote:
> Sorry about this guys - I know I am generating more traffic than I
> should just at the moment but I am running into a bunch of stuff I
> need opinions on.
>
> I see no way of doing taxonomic order in the current TCS. It can be
> implied by document order but because of the way the pointers are set
> up this isn't obvious. If I create a concept for a family it has an ID
> and the genera it contains can come anywhere in the file so long and
> they have references to it and/or it has references to them.
> Now we could:
> 1. have a convention where the order the taxa appear in the
> document, even if they are not in a contiguous block is their
> taxonomic order. This is messy as you have to keep track of
> all the taxa all the time in case another one is added further
> down the document.
> 2. have the document order of the relationship pointers in the
> higher taxon as the taxonomic order, even if they are not
> contiguous and interspersed with synonym and other
> relationship pointers. This doesn't work if we permit
> subordinate taxa to point to higher taxa without the pointers
> going the other way or if we allow sub-subordinate taxon
> pointers (e.g. pointing to genera as well as to subgenera -
> see last discussion) which we appear to.
> 3. put an ordinal field in each TaxonConcept - but that wouldn't
> work because TaxonConcepts can be pointed to from different
> higher TaxonConcepts and so presumably have different orders
> in different parents.
> I see no way of doing it. Anyone got any ideas?
>
> I didn't think this was a major problem till I thought about
> applications like:
> 1. Providing a list of taxa in a herbarium or other collection in
> the order they are arrange (any one a curator?)
> 2. Providing a representation of a monograph as published.
> 3. Providing the layout of a checklist like a field guide.
> 4. just about any other application you can think of other than a
> web search interface.
> Am I missing something here? Is there a way of doing it?
>
> I am starting to work on a nested hierarchy of taxon objects - it just
> makes things soooo much easier.
>
> As usually I am appreciative of your time and would be grateful for
> peoples thoughts.
>
> Roger
>
> --
>
> ==============================================
> Roger Hyam
> ----------------------------------------------
> Biodiversity Informatics
> Independent Web Development
> ----------------------------------------------
> http://www.hyam.net roger at hyam.net
> ----------------------------------------------
> 2 Janefield Rise, Lauder, TD2 6SP, UK.
> T: +44 (0)1578 722782 M: +44 (0)7890 341847
> ==============================================
>
>
> ______________________________________________________________________
>
> _______________________________________________
> Tcs-lc mailing list
> Tcs-lc at ecoinformatics.org
> http://mercury.nceas.ucsb.edu/ecoinformatics/mailman/listinfo/tcs-lc
--
Greg Whitbread <ghw at anbg.gov.au>
ANBG/CPBR/ANH
More information about the Tcs-lc
mailing list