names and numbers

Wed Oct 17 10:02:55 CDT 2001

Dear Sirs,

In a similar vein to one of your earlier contributors I too have been
mulling over a reponse to some of the postings on this thread and
like him although I find that most of the points I wish to contribute
have more or less already been stated I would nevertheless like to
make a contribution to the debate.

It strikes me that the central issue of concern in this area is that of
EFFECTIVE communication. In order to achieve effective
communication you must have two things. 1) the ability to read
each others communications, 2) the ability to understand the
meaning of those communications. The two things are
distinguished in the following example. I can read Latin words but
by and large I can't understand any of them !

As far as I can see assigning numbers to names simply addresses
point 1) by making it easier for our disparate database systems to
read communications from one to another. It does nothing to assist
in the understanding of the communication. In order to achieve this
the concept associated with the name must either be guaranteed
to be standard across all systems or must be communicated along
with the indentifier of the name.

Given that taxon concepts are dynamic I do not believe that it
would be feasable to construct a long lasting, universally agreed
conceptual framework. It is therefore necessary to include
concepts alongside names in order to achieve effective

As an example of why this is important let us consider two field
surveys taken some time apart. The first survey uses the latest
flora available at the time. In the interim between the two surveys a
new flora is published in which considerable rearrangement of the
taxon concepts occurs. The second field study uses the new flora.
Suppose we now want to compare the species abundance in the
two areas over time. Could we do it by simply counting the records
associated with each name. Most certainly not because the
changes in the taxonomy confound the issue. Without knowledge
of how the circumscription of the concepts has changed over time
it would be impossible to make any meaningful comparison of the
two data sets ! Aside from the issue of effective communication
this raises the important point that data collected using a particular
taxonomic framework MUST be stored with reference to that
framework, and only be subsequently referred to within the context
of that framework.

In producing the Prometheus model for the simultaneous storage of
alternative taxonomies it was precisely this problem we were trying
to address. By using specimens used to construct the
circumscription to represent the circumscription of taxa, we provide
an objective means for the comparison of differing taxon concepts
associated with the same name. By this I mean we give a third
party the information required to make their own judgment on the
degree of congruence between two alternative concepts assocaited
with the same name, rather than relying on the assertions of
previous contributors to the database.

We also allow data collected within a particular framework to be
stored in association with that framework. Although Dr Croft in a
must earlier posting  ventured the opinion that this is a "crummy"
way of going about handling information it reflects the reality of
taxonomy !

The discussion we are seeing now is the almost inevitable side
effect of trying to improve our models of the taxonomic system. It
shows up the flaws in taxonomy as a means to promote the
effective communication of concepts. So lets not get caught up in
arguments about how best to build taxonomic databases let's
discuss the far more important issue of how better to do taxonomy
so that concepts can be more robustly exchanged!

To finish this rather long message

>1) Does your database assign an arbitrary surrogate primary key >to each taxon name entity, for use as an internal constant and >unique identifier for that name?

With respect to Prometheus. It most certainly does.

>2) Would the world not be a better place if my database, and the >Species 2000 database, and the ITIS database, and the Index >Kewensis database, and every other taxon name database out >there used the same set of arbitrary surrogate primary keys for >the same taxon names?

It most certainly would not, this would lead us down the very
dangerous path of making it simple to assume that the same name
means the same thing. We have to remember that if taxonomy is
to be of any use its end products must be available and
understandable by non-taxonomists. We must make sure that
these users are aware not only of the value of the data presented
but also of its shortcomings and not hide these behind the facade
of a name numbering scheme.

Martin Pullan

Dr.Martin Pullan
Taxonomic Computing
Royal Botanic Garden Edinburgh
20a Inverleith Row

Tel (+44) (0)131 248 2908
Fax (+44) (0)131 248 2901

More information about the Taxacom mailing list