[SEEK-Taxon] Re: Specific Questions about TCS (TCS documentation - relationship types)

Nico Franz franz at nceas.ucsb.edu
Tue Mar 8 08:27:55 PST 2005


Hi Rich - thanks!

   Your message could have been Cc-ed to seek-taxon at ecoinformatics.org 
instead. I have not made this list, nor do I think it is very close to 
production phase (same for the rest SEEK Taxon). It is mostly due to the 
Berlin Model. Bob Peet, Xianhua Liu, Jessie and myself went through it a 
few weeks ago and our reactions were similar to yours. This list needs a 
substantial revision (soon!) and my previously attached file with your 
annotations is a start. So thanks again, I suggest we revisit this later.

Nico

Richard Pyle wrote:

>Hi Nico,
>
>I decided not to clutter the list with this message, and instead send it
>directly to you.  I gather you are the person within SEEK who is driving the
>enumeration of these various Relationships between concepts and between
>names.  If any of this is worth putting on the list for further discussion,
>please feel free to do so (or let me know and I will do so). In any case,
>what follows are questions about the content of
>http://www.soc.napier.ac.uk/doc/tdwg_tcsv22.doc, primarily regarding the
>table on pp. 5-8:
>
>SET OPERATIONS BETWEEN CONCEPTS:
>* (Minor) What is the current thinking of treating "excludes" as the negated
>version of "overlaps".  It seems in keeping with rules of logic that this
>should be treated as such -- especially of assumptions about logical rules
>capitalize on
>
>HIERARCHICAL RELATIONSHIPS BETWEEN CONCEPTS OR NAME CONCEPTS
>* From the heading, the implication is that these relationships apply to
>both concepts and names, but I'm not sure that they do -- at least not in
>the same way.  For example, in the Concept realm, "is parent of" may imply
>an "includes" relationship, or an "is congruent to" relationship (the latter
>in the case of monotypic genera and families with only one genus).  It also
>may apply at any taxonomic rank.  From a strictly nomenclatural perspective,
>I'm not sure there really is such a thing as an "is parent of" relationship
>between names.  This would be replaced by something more like "is combined
>with", and would only apply to names of rank species or lower.  If an author
>places a genus within a family, there is only a concept relationship between
>the family and the genus -- not a nomenclatural one.  The only exception to
>this is in cases where a genus has been designated as the type genus of a
>family-group name.  In any case, I don't think it is correct to apply an "is
>parent of" (or "is child of") relationship between two names. I could also
>quibble about the meaning "immediate lower rank", but I don't think that
>would be helpful at this time.
>
>NOMENCLATURAL RELATIONSHIPS (BETWEEN NAMES OF CONCEPTS)
>* I strongly believe that all of the relationships that truly belong in this
>category (excluding the ones I comment on below as not being name-name
>relationship types) should be restricted in application *only* to Nominal
>concept types, and that non-nominal concept types should refer back to the
>Nominal concept instance for the applied name, from which the nomenclatural
>relationships would be derived.
>
>* There are only two cases of the "is type of"/"has type" relationships
>between names, and none of them involve concept circumscription definitions.
>These are:
>	* "is type species of"/"has type species" establishes a relationship
>between a species-group name and a genus-group name.  This is not the
>relationship between a species concept circumscription and the genus-group
>name; rather the type species serves the purpose of establishing a
>connection from the genus name down to the type *specimen* of the
>species-group name (see ICZN Article 61.1.2).
>	* "is type genus of"/"has type genus" establishes a relationship between a
>genus-group name and a family-group name. Same comment RE ICZN 61.1.2 as
>above applies.
>
>ICZN: http://www.iczn.org/iczn/index.jsp?article=61
>61.1.2. Objectivity provided by typification is continuous through the
>hierarchy of names. It extends in ascending order from the species group to
>the family group. Thus the name-bearing type of a nominal species-group
>taxon is a specimen or a set of specimens (a holotype, lectotype, neotype or
>syntypes [Art. 72.1.2]), that of a nominal genus-group taxon is a nominal
>species defined objectively by its type; that of a nominal family-group
>taxon is the nominal genus on which its name is based.
>
>* Same limitations as describe above apply to "has conserved type"/"is
>conserved type of" relationship (i.e., only relevant in the context of
>name-name relationships for type-species and type-genera; not for
>type-specimens)
>
>* "is heterotypic synonym of" is not a name-name relationship; it is a
>concept-concept relationship (technically, it is a concept-voucher
>relationship).  To say that "Aus beus" is a synonym of the heterotypic "Aus
>ceus", is really to say "my concept circumscription includes both the
>primary type specimen of the name 'Aus beus' and the primary type specimen
>of the name 'Aus ceus'".  Which name gets applied to the concept
>circumscription depends on which name has nomenclatural priority.  In TCS, I
>think that heterotypic nomenclatural synonymies should be derived from
>concept relationships; not recorded separately name relationships.  In fact,
>if you really wanted to model it after the true information associations,
>you would do it via Voucher links to the primary type specimens of all names
>included within the concept circumscription.
>
>* "is later homonym of"/"has later homonym" -- I think this should be
>replaced with a separate symmetrical "is homonym of" (in the same way there
>is a symmetrical "is homotypic synonym of").  I presume that the use of the
>word "later" here refers to nomenclatural priority, which I think is a
>different kind of name-name relationship (see below). Though usually based
>on chronology (i.e., "later"), it is not always the case. I'm also not sure
>I understand how "is treated as later homonym of" differs.
>
>* I think you need a pair of inverse relationships "is senior to" and "is
>junior to" (or "has senior priority to" and "has junior priority to"). These
>would be used in conjunction with the symmetrical "is homonym of" and "is
>heterotypic synonym of" relationships to establish nomenclatural priority.
>
>* Relationships involving lectotypification and neotypification (not sure
>exactly how an epitype is used, but probably that as well) are not between
>two names -- they are between a specimen (Voucher) and a name.  As such,
>they really belong in the "SpecimenCircumscription" part of the schema; not
>in the Relationships part.
>
>* "is orthographic variant of" implies that a separate name object is
>created for all orthographic variations of a name.  I don't think that every
>orthographic variation deserves the creation of a new Nominal Concept,
>because neither Zoologists not botanists view these as distinct "Names".  I
>believe there is a need for a "NameVerbatim" element to capture orthographic
>variants at each TaxonConcept instance.
>
>* "is not a synonym of" -- this would necessarily involve heterotypic
>synonyms, and thus is an example of a relationship among concepts. Same
>applies to "is partial synonym of" and "is pro parte synonym of"
>
>* "is recombination of"/"has recombination" -- this isn't really a
>"relationship" between names per se, rather it should be an instance flag:
>"is new combination".  This flag could be eliminated entirely, and implied
>for all cases where "is basionym of"/"has basionym" is established.  Perhaps
>a better flag would be "is protonym", which defines the subset of names that
>can serve the role of basionym.
>
>* Once again, "is rejected type of" and "has rejected type" only constitutes
>a name-name relationship when a type species or a type genus is involved.
>In other cases, it is a property of a name-voucher relationship.
>
>* "is synonym of" is ambiguous, and shouldn't be allowed as a relationship.
>All synonyms should be qualified as either homotypic or heterotypic. This
>should always be ascertainable, given a pair of names proclaimed to be
>synonyms.
>
>* "is validation of"/"has validation" -- I would need to see some examples
>of these to understand what is meant by them.  I'm pretty sure they would be
>nomenclatural. Same for "is taxonomically included in"/"taxonomically
>includes"
>
>* I don't think "is anamorph of" and "is teleomorph of" are true name-name
>relationships -- I'd need to understand how it is used.
>
>* What's the difference between "is equivalent of" and "is congruent to"?
>
>* "is sanctioned by"/"is sanctioning" *might* be expanded to include the
>ICZN "Official List" and "Official Index" -- but not sure.
>
>HYBRIDIZATION RELATIONSHIPS
>* These need further discussion, because they involve different
>nomenclatural implications for botany and zoology.
>
>GENERAL
>* "Doubtful" seems to me to be a property of any relationship; not a
>relationship in itself.  If you really want to have a relationship to
>express "Uncertain relationship between Concepts 1 and 2", then you should
>probably pick a word like "Unspecified".
>
>The question was asked on p.8, "Does the  principle of typification
>according to the rules of nomenclature imply transitivity??"  For ICZN, the
>answer is "yes" (see Article 61.1.2 above).
>
>Perhaps it would be best to see how much you and I can agree on these things
>first, then put the ones we can't agree on out for comment on the list.
>
>Many thanks!
>
>Aloha,
>Rich
>
>
>  
>




More information about the Seek-taxon mailing list