Re: [Obo-phenotype] request for comments

[William H. <WILL.HAWTHORNE@BTOPENWORLD.COM>]

dear Hong,
I think your paper is starting to get its teeth into an interesting subject, and one where there is a lot of scope for defining better standards. I was a bit surprised and even nervous but not displeased to see the VFH website glossary being subjected to such extreme scrutiny, like someone picking up a discarded notebook and starting to criticise its punctuation!  
THere are various additional facts about the VFH glossary that may help your interpretation of it:

1.I originally put it up as a sort of 'put it out there maybe as a seed for improvement', rather than a finalised glossary. It was/is  a work in progress, with some errors. Maybe nows the time to encourage updates somehow.

2. A main goal actually is that it strives to minimise jargon for different plant guide purposes, and mainlky for practical and general field guides, hence highly technical taxon-specific terms have been excluded as far as poss, and it is explicitly not really trying to be encycloediac. Some plant glossaries seek to be all-inclusive, (encyclopoediac or dictionary like) rather than critical and minimalist, and therefore often continue to confuse rather than help readers if they are promoting their own favoured definitions of terms when in fact chaos and fuzzy meanings prevail in real literature. Many terms for pubescence are used ambiguously, so I tried to downplay need for such terms rather than define all possible meanings. Even "ovate" is used variously by different authors - a fact computer paresers would need to deal with somehow.  (I wrote a chapter on roughly this for our book: Lawrence and Hawthorne, Plant Identification, Earthscan).

3. The explicit multi-level hierarchy I originally arranged thE VFH glossary in (available as A DBF, and as a pilot plant photo tagging software if you are interested) seemed like a good idea at the time, especially to enable categorisation of properties ('tags') of plant photos to be a bit wiser than with just the tag words themselves. e.g. as "inflorescence" owns "cyme" and "cyme" owns "scorpioid (cyme)", taggers can apply the term inflorescence if they dont know better, but if they want to specify scorpioid cyme then the system will know that implies inflorescence also, if a user asks to see infl. pictures. WHen filtering the image gallery this obviously makes optimum us of tags.

Various problems and interesting features become clear with this approach however. Characters "own" character states and this is one meaning of ownership in the hierarchy; also generic suites of terms can be grouped under a single node (without their being mutually exclusive descriptors) e.g. Some terms like cyme are effectively nouns (I think you refer to these as is-a terms) or adjectives (part-of, or kind of) like "scorpioid (cyme)". ie the semantics of  is-a and is-an-example-of, or is-part of are not as precise as they may at first seem, partly because of the lack of precision of language. which is bad, and partly because morphological descrpition allows successive approximation whereby a descriptor becomes a more generic term itself waiting to be specified, which is a good feature. Even "Leaf shape", a character, becomes a specifier or character state when distinguishing pictures or species with a leaf that has shape from those which dont have any (shape).

There are some conflicts when using the hierarchy both to tag pictures and to attribute species character states using the same terms (that had been the original idea). "Inflorescence" might easily own "infructescence" for one purpose (infl form) but not another (if only fruiting and not flowering is seen in a picture). Strictly speaking it would be better to have various different entries "inflorescence - abstract types of ", inflorescence (inlcuding when fruiting)," and inflorescence (strictly speaking, only with flowers).  As the glossary was meant to be simple, these get rather fudged together, which isnt ideal.

The "hierarchy" for some purposes would ideally not be strict - eg shape and position descriptors applying to any organ currently has to be repeated (leafshape owns ovate; calyx shape owns ovate..), a displeasing fact more obviously wrongly dealt with in hierarchy  when referring to colours or numbers of parts. Where does one draw the line between general purpose adjectives (counts of items) and botanical jargon character states applying to specific organs?

Another distraction from logic in the sematics is I suppose is that many groups, eg orchids, grasses, palms tend to have their own jargon eg for leaf or flower parts or types - implying that both the taxon and organ (eg 'Grass inflorescence') in question should be an owning node in the morphology hierarchy. Should "inflorescence (types of)" own 'grass inflorescence', and 'grass inflorescence' then own all of the specific terms for grass inflorescence (parts of), and for grass inflorescnece (types of)...'  Very messy. 

I wouldnt say I gave up, but I certainly came to the conclusion there's only so far you can go trying to use existing botanical terms in a logical sematic system. Best to use the commonest and least ambiguous terms thus, which is what the VFH glossary aimed to do and after that use And refer to photos!

I would be quite interested to develop these sort of ideas further if you are interested, (a paper on "the limits of standard botanical jargon in e-floras, and suggestions for a new glossary appropriate for a new era of morphlogical description' might be a working title Id be happy to contribute to - if anyone interested let me know.).

Cheers
William









 

  ----- Original Message ----- 
  From: Hong cui 
  To: rk...@an... ; dup...@vt... ; obo...@li... ; wil...@pl... 
  Sent: Sunday, November 01, 2009 6:11 AM
  Subject: Re: request for comments


  Dear Prof. Porter and all,

  Sorry about the corrupted files. I converted Word files to PDF and please let me know if you have problems opening them.

  Prof. Porter, I think the Glossary we received (in 2007) is the 2003 version, but since we received the Glossary in csv format, I can not verify the publication year now. I will check my records when I go to office on Monday.


  Hong


  On Sat, Oct 31, 2009 at 8:29 PM, Duncan M. Porter <dup...@vt...> wrote:

    Dear Dr. Cui:

      I am able to print out your draft, but it appears as gibberish. I am unable
    to print out your appendixes, unless perhaps they are sent as .doc, not .docx
    files. Are you aware that Professor Kiger and I published a revised edition of
    the FNA document in 2003?

    Sincerely,

    Duncan M. Porter



    Quoting Hong cui <hon...@gm...>:

    > Dear Colleagues,
    >
    > I am writing to share a draft paper with you and request your comments.
    >
    > The paper examines the Categorical Glossary for Flora of North America
    > Project, Oxford field guide thesaurus, PATO ontology, and Radford's glossary
    > (1976).
    >
    > This work was out of the need of an ontology for semantic annotating
    > morphological descriptions. More information about the project can be found
    > at http://sites.google.com/site/biosemanticsproject/
    >
    > Errors and inaccuracies very likely present in the first draft. Any comments
    > and suggestions from you are much appreciated.
    >
    > Sincerely,
    >
    > Hong Cui, Ph.D
    > Assistant Professor, IT
    > School of Information Resources and Library Science
    > University of Arizona
    >



    Duncan M. Porter
    Professor Emeritus of Biological Sciences
    Department of Biological Sciences
    Virginia Polytechnic Institute & State University
    Blacksburg, VA 24061
    (540) 231-6768