On Wed, Nov 28, 2012 at 10:17 PM, J. McRee Elrod <[log in to unmask]> wrote:
Simon Spero posted:


>(1) Everything which is eaten by something is a food.

>(2) Everything that a carnivore eats is an animal.  Everything that
>eats an animal is a carnivore.

>(3) Everything that a herbivore eats is a plant.  Everything that eats
>a plant is a herbivore.

>(4) Every fly  is an animal. Every antelope is an animal. Every flytrap is
>a plant.

I suppose for purposes of illustration, we should pretend that all of
these statements are true?  They clearly are not, and exceptions can
be cited for most.  While a fly is not an animal (it's an insect),

That's actually not one of the simplifications that admits exceptions.  True Flies are:

Order Diptera, Class Insecta, Phylum Arthropoda, Kingdom Animalia, Domain Eukaria, Errata Tintinnabuli
Every fly is an insect. every insect is an arthropod. every arthropod is an animal. every athropod is a eukaryote. every eukaryote is a doorbell

humans are animals, and most eat both plants and animals.

 This statement is correct, and I initially started including them in the example  (due to the presence of a opportunistic omnivore as well as an obligate carnivore next to me as I wrote the example). In the end I decided it did not served to obfuscate the point.  After removing the omnivores, I added another antelope, in anticipation of an objection of this kind. More antelopes means more horns, on which dilemas get caught, making this a theory unsafe for omnivores.  

A useful exercise to help motivate you to read through the details  would be to think about the exceptions you can see, and try and specify them as precisely as you can - then see if you can modify the theory to handle them. 
Some starting points:  
     (1)    Modify the theory so that it does not prohibit the existence of omnivores, but does not allow herbivores to eat animals or carnivores plants. 

    (2)     Modify the theory so that it allows for omnivores, but draws the most precise inferences possible.

    (3)*     Try and modify the theory to allows for omnivores which never eat meat (e.g. vegans), and those who can rarely afford meat. 
           Do you need a more complex theory of Actions and Events (a a la Davidson ).  Might modals be necessary?  

    (4)  Modify the theory to more accurately model the idea of eating for nutrition, as opposed to eating for other reasons.  Do you need to introduce intentionality-with-a-t?

    (5) Modify the theory to handle cases such as dogs swallowing  grass to induce vomiting. Does this count as eating?  What about animals which chew the cud? 

    (6) Modify the theory to handle case such as birds which store stones in their gizzards to aid in digestion.  Does this count as eating?  

  Some of this might not be possible to express using OWL.  If this is seems to be the case, just write your axioms in a regimented form of english (if you can express them in ACE then you can check to see how your ideas look in logical form; you can also see if the paraphrased version of your statement corresponds to what you wanted to say).

You might like to try a more powerful  knowledge representation system, such as opencyc,  Although a lot of axioms that express this sort of details are not exported in the opencyc release, you may find it easier to capture your meanings more precisely. 

Interesting bibliographic-related exercise:  

Try and formalize the meaning of a work attributed to Author (Spirit) ; especially the temporal relations between a literary identity and the spirit of that literary identity  
 What axioms do you need to express the expected doxastic state of the actual creator of the work about the existence of spirts, the doxastic state of the readers of the work, and the doxastic state of the actual creator of the work about the doxastic state of the readers of that work.  
Can this be modeled in a general form using Microtheories? 

How should the propositional content of a non-fiction work attributed to a spirit author be handled?  
How does this compare to posthumous works that are known to have been written by the named author (e,g, Philosophical Investigations, which was not published till 1953, but which had a   50th anniversary edition published in 2001 )

How does this compare to the case of an pseudonymous or anonymous author?
 How does this compare with the way one should handle the propositional content of a roman a clef? 

If you don't have the time to do this in cyc, it would still be a useful exercise to try and address the problems using as rigorous a form of english as you have time for, or if you still have your logic textbooks from your college days and feel more at ease using traditional notation, that may be easier 

Simon // This assignment will not be part of your grade, You may collaborate with others.  Open books are allowed.