[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Panstems



(Crossposted to the Dinosaur Mailing List, as some people on that forum may be
interested, since this deals with the definitions of dinosaurian clades.
Follow-ups should probably be posted to the PhyloCode Mailing List, however.)

--- "Jonathan R. Wagner" <jonathan.r.wagner@mail.utexas.edu> wrote:

> [...] I support not restricting definitional types, or the 
> choice and application of names (other than for the sake of continuity). I 
> am sympathetic to Dr. Padian's point (by the way, he favored two formats, 
> node- and stem-, in his Paris talk), but I do not agree. I feel that PN has 
> moved past the idea that the wording of a definition somehow invokes a 
> general case and the wording is irrelevant (as implied by the various 
> shorthand definition formats floating around); I believe this is the 
> mindset that inspired Dr. Padian's suggestion. Instead, the wording of the 
> definition is paramount. As such, not all node- and stem-based definitions 
> are equal, the various classes of definition are just that, classes (see my 
> point about the so-called "stem-modified node-based definion" in my 
> abstract), and other kinds definitions are possible and indeed desireable.

Which brings us to an interesting point. If definitions are considered to be
prose, and not rigorous formulae, we run into several problems. First of all,
wordings can be ambiguous. Secondly, reading prose requires full comprehension
of the language employed. Suppose somebody were to formulate definitions in
Russian, Portuguese, Hindi, Swahili, Classical Latin, Classical Greek, etc.?
Would the database then have to preserve that wording as THE definition, and
wouldn't anyone who didn't know that language be impeded from understanding it?

While thinking about these issues over the past couple of days, I tried to come
up with a style of notation for internationally legible clade definitions.
After one false start based on computer coding conventions, I came up with a
system utilizing internationally accepted styles of mathematical notation,
primarily notation dealing with sets, Boolean logic, and functions. This seemed
to me far more concise and readily comprehensible than my initial attempt with
"pseudocode".

Unfortunately, this system involves non-ASCII characters, so it cannot be
posted here. But I uploaded a Word document of the first draft detailing the
system here: http://dino.lm.com/keesey/documents/PhylogeneticNotation.doc

Here is a brief outline:

BASIC NOTATION
  SETS - set notation
  LOGIC AND CONDITIONS - Boolean notation
  FUNCTIONS - function notation
  ELEMENTS - original notation for dealing with characters, specimens, etc.

SETS
  REALITY-BASED SETS - pre-defined sets for dealing with entities such as
characters, specimens, etc.
  SECONDARY SETS - sets which may be derived from the reality-based sets,
involving entities such as clades

FUNCTIONS
  REALITY-BASED FUNCTIONS - pre-defined functions for dealing with character
possession, direct ancestry, etc.
  DEFINITION-BASED FUNCTIONS - functions for dealing with simple defined
entities (species)
  SECONDARY FUNCTIONS - functions derived from reality- and definition-based
functions, dealing with ancestry, lineages, clade definitions, etc.

EXAMPLE DEFINITIONS
  FROM THE DRAFT PHYLOCODE - some prose definitions from the draft PhyloCode
rendered in phylogenetic notation
  OTHERS - various other samples

You can see that this notation would be universally appreciated, the only
exceptions being the prose descriptions of characters and of pre-defined sets
and functions, neither of which I can see a way around. The only other case
where a specific language enters into the matter is in the names of sets and
functions, but this is not a real issue since they are all defined using
mathematical notation. These names could just as easily be derived from another
language, such as Classical Latin.

Another benefit of using a rigorous notation for definitions is that they are
then capable of being parsed and understood by computer programs, which could
have useful applications.

What I would like to know is whether anyone else agrees that using a rigorous
system of notation would be a good idea, and, if so, what changes should be
made to my draft (or if this draft should just be scrapped and something else
created, and why). I encourage anyone who is interested to download the file
and add notes (Tracking Changes is turned on in the file) and send it back to
me, or upload for the list to discuss. There are some parts I'm kind of unsure
about (e.g., the formula for apomorph-based clade definitions, the reliance on
"breeding groups", the reliance on non-ASCII characters), and I would really
appreciate feedback.

Thanks,

=====
=====> T. Michael Keesey <http://dino.lm.com/contact>
=====> The Dinosauricon <http://dinosauricon.com>
=====> Instant Messenger <Ric Blayze>
=====


                
_______________________________
Do you Yahoo!?
Declare Yourself - Register online to vote today!
http://vote.yahoo.com