[tcs-lc] Have your cake and eat it?

[Nico Franz]

Hi Roger:

   Those are some very concrete proposals.

(1) is obviously ok with me.

(2) I assume we pretty much have a consensus on going from "Kingdom" to 
"RelevantCode" (something to that extent); after all in that TCS field 
we're not modeling how many "Kingdoms" bacteriologists currently 
perceive to exist. Can anyone please point me/us to the relevant 
point/counter-point for "Rank?" That one I feel less assured about.

(3) and (4) seem muddled. Your average practitioner will think of 
herself as "describing taxa" but looking at the picture closely has made 
us to think of that phrase as "describing what she currently perceives 
to be an adequate relationship of her language use to some entity out 
there (typically including some actual specimens)." So there's a name, 
and then hopefully a natural entity (taxon), and what we're modeling is 
how she nailed down the connection among the two (that's the concept) 
and then also how others faired at that task. Abandoning the 
name/concept/taxon three-way of thinking would set us back I think. If 
you are looking for a term that encompasses both the activity of naming 
and classifying, it might have to be something else that stays within 
the domain of practical taxonomic activity. I don't know what that could 
be but unlike "Taxa" your non-/declarative dichotomy is at least heading 
that way.

(5) You're correct that in order to sell the TCS we don't necessarily 
need the Nomimal/Original/Revisional distinction. However, for 
within-group talks such distinctions have sometimes been helpful, e.g. 
if we wanted to understand how particular data sets (IPNI vs. ITIS) are 
looked upon and related to each other from a TCS perspective. How much 
granularity do we need for the current discussion? Probably not that 
much. The key issue remains delimiting the size of that elusive 
"nomenclatural" box so that everyone's essential needs are met.

Cheers,

Nico

Roger Hyam wrote:

>
> Hi Everyone,
>
> A solution to the names embedded vs names as top level elements has 
> occurred to me that will require little alteration to the TCS as 
> proposed (more or less 0.90b as proposed by Napier but with a few 
> changes).  Please read all the points before responding to the first 
> ones as they are related!
>
> 1) LC is embedded within a TaxonConcept element but the relationship 
> stuff is moved to TCS.
>
> 2) Various other bits are rationalised as agreed in the notes that 
> were circulated earlier (rank, kingdom moved etc).
>
> 3) Rename the TaxonConcepts element to Taxa.
>
> 4) Rename the TaxonConcept element to Taxon
>
> 5) The attributes in the TaxonConcept element should be changed (and 
> we may need to discuss this). What I have been thinking is we need a 
> flag like
>
> nomenclature="declarative" - meaning the name data in this taxon 
> definition is meant to be 'authoritive' and correct. If you want to 
> communicate nomenclatural data specifically you can flag your taxon 
> like this.
>
> nomenclature="nondeclarative" - the default setting. System presumes 
> you are talking about a taxon concept and not making declarative 
> statements about nomenclature.
>
> Thus if you want to have names as top level you can just tag your 
> Taxon object as being a nomenclatural one but if you don't know what 
> you are doing we presume you are talking about a taxon concept and are 
> probably trying to join it to things on the basis of a name. Thus you 
> can have your cake and eat it. Nomenclators can explicitly talk about 
> names whilst 'other' users talk about taxa.
>
> This is close to the type="original" that is schema but I think 
> importantly different in that it allows people to explicitly exchange 
> data about nomenclature rather than the "original _concept_" which is 
> something different.
>
> My current thought is that most of the confusion around the 
> names/concepts (well my confusion anyhow)  may be generated around the 
> different types of taxon concepts. Namely nominal/original/revision. 
> There has certainly been discussion on this list where people have had 
> different views of what a Nominal concept is for instance. I wonder 
> whether we could do away with these distinctions all together? They 
> are all implied by the data anyway. The original concept is just 
> according to the original publication, the nominal is according to 
> null and the revision concept is everything else.
>
> Would be grateful for peoples thoughts.
>
> Roger