[Sigia-l] Findability

[John Fullerton]

> (i.e., should not hope for a "meta" language that is
> somehow clearer or purer than that which is described).
> A descriptive language is not, by its nature, a better
> conveyor of meaning than the language of the text it
> describes.

These comments remind me of the concept of
abstraction in computer programming. Now, I do not
have an authoritative definition of abstraction yet
here's a definition found online.

The Concept of Abstraction
	Abstraction concentrates on the significant features that
differentiate different yet similar entities.
	The abstraction can be viewed from two aspects -- features that
are common to all entities among a group (subordinate attributes), or
features that can tell them apart (essential attributes).
	A subprogram is a good example of process abstraction, e.g., a
sorting subroutine is specified by what it does, not how it does it.

from an online Word document
http://www.csie.ntu.edu.tw/~pangfeng/PL%20notes/chap11.doc 

First of all, I'm thinking that abstaction is somewhat like
categorization. A few comments about abstraction--
1. Computer functions allow processes to be used
via the name of the process (without need of referring
to all of the technical details), so PRINT "HI" instead
of saying HOW the computer will print "hi".
2. Much more advanced work, and much more
involved work can be done because of the use
of abstractions. Just imagine how much time it
would take to give minute directions to anywhere.
While facing North on First street, turn in the direction
called by convention left at the tree on the corner of
Main. The tree is a poplar tree of the genus
such-and-such, first identified by the botanist Mr. T.
The identifying characteristics of the poplar are
X, Y, and Z. In particular you will want to note the
poplar blossoms. And on and on it goes. Instead,
On First, turn left on Main is much easier. And the
same is true in computer programming.

Abstraction does not convey all of the information
and that very limitation makes it so useful. Of course
the usefulness depends on the abstraction being
good. Naming the print function something like
function1403 will not be most useful.

Likewise, categorization can help in minimizing
the need to use full descriptions to get access
to particular information. In conversation, one
could ask a friend to recommend a mechanic
shop without having to describe in detail the
work needed on the car and a question about
whether there may be any businesses in town
that can help with the car work. In the phone
book, one can find a mechanic without having to
browse the universal alphabetical list of
companies. 

I think these observations are for the most part
mundane. Am I missing the objection to findability?

> Proper categorization will approach its subject with
> respect rather than attempting to force it into a
> scheme created to serve alien interests (IOW,
> categorizing something so as to improve its
> 'visibility'/'perceived value' without considering
> its character).

What I've been hearing in the objection to
findability is rejection of ALL categories, so that
there is NO proper categorization, thus my effort
to show why categorization is helpful.

Have a nice day
John Paul Fullerton
j-fullerton at tamu.edu