[Corpora-List] fuzzy logic-short summary and thank you

[Linda Bawcom]

Dear friends, colleagues and list members,
   
  First, my apologies for taking so long to respond with a summary and thank you regarding fuzzy logic. I was on vacation in the beautiful Canadian Rockies (a pleasant change from this time of year from Houston). And an additional apology for this too long missive, which is really too short. 
   
  A few days before leaving I posted 3 questions. The first two had to do with the ANC and Frown which Carmela Chateau and Hans Juergen kindly answered. That is, that FROWN comes with the ICAME corpus (a bit pricey for individual use) and the ANC can be used with Xaira (which I still need to find out about but will be receiving the ANC soon so I imagine I’ll learn how to use the ANC (eventually). I learned that Xaira was XML compliant and that Wordsmith prefers text files, something I never realized.
   
  In fact, I have learned a great deal from my third question (I had to look up:  entropy [rather reminded me of the concept of ‘opportunity cost’ in economics], afaik, phylogeny, wetware, Boolean sets, Goedel, Chaitin, Kolmogorov, and Grimm’s Law). My third question was: Is fuzzy logic ever used to measure semantic similarity? I actually didn’t get an answer to that, although I did get some private and public responses which I am very thankful for regarding whether or not it might be of interest to me.
   
  I cannot do justice to the flurry of e-mails in response to my question as 1) a number of them were outside my area of expertise (if I could be said to have one) and 2) there was enough information (in the 35 pages I printed out) to write 10 doctoral theses. However, thanks certainly go to (no particular order here): John Sowa (who shared an excellent summary taken from C. S. Pierce and another excerpt  from Steiner, which I particularly enjoyed), Chris Brew (thank you for the Chicken or Egg reference)  and Mark Line both of whom brought up some very thought-provoking questions, and others who also took the time and energy necessary to respond such as Rob Freeman and his link to the Chaitin article, (which certainly triggered responses), and James Fidelholtz, Daoud Clarke, Mike Maxwell and  John Goldsmith (my sincerest apologies for anyone I have left out on the public response side). 
   
  While these were a delight to read, it saddens me a bit that by necessity we must specialize, and therefore, we are probably ignorant of models, concepts, and theories in other fields of language that could, perhaps, help answer some of the excellent questions raised and debated. This debate had us questioning some of the goals of linguistics and how best to attain them, the difference between what I might term: ‘“truth”’ answers’, ‘scientific answers’ and ‘useful answers’ regarding observable entities; the definition of language, whether or not a corpus (which was never defined) is generated by rules,  and how big a corpus must be to be  truly representative, if one can ever be.  There was a lot of discussion regarding ‘compression’, which I am sure must be quite important, but was often, unfortunately, beyond my understanding. This is a poor summary as I have left out much substance of the debates and many of the questions raised  and debated deserve  a much more in
 depth response. 
   
  In my own case regarding synonymy, I have traveled down winding and bumpy roads which have led me to philosophy (what does ‘mean’ mean and what does ‘same’ mean? Is there any such thing as a synonym?) logic ( are ‘bachelor’ and ‘unmarried man’  synonymous?) calculus ( A Logical Calculus of synonymy by Yiannis Moschovakis), psycholinguistics (how are words stored in our mental lexicon/brain? Why are collocations so pervasive? How are we ‘primed’ to utter the sequences of lexical items that we do utter?), computational linguistics (what factors should be taken into account when trying to measure distances? what are the difficulties in quantifying that information?), corpus linguistics (what is a corpus? What is a representative corpus?, What criteria should be used to make a corpus? How much can I generalize based on the corpus that I am using?) And these are only a few of the questions of course (which you needn’t respond to).
   
  As I mentioned to one colleague, I have become a jack of all trades and master of none, but the journey has been an exciting and enlightening one and some of the thanks for that journey goes to the corpora list. 
   
  Kindest regards,
  Linda Bawcom
   
   


      "Any man's death diminishes me because I am involved in mankind." John Donne



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://listserv.linguistlist.org/pipermail/corpora/attachments/20060809/7989bcf2/attachment.htm>