Re: VMs: Faces at the roots

[Wayne Durden <wdurden@xxxxxxxxxxxxxxxx>]

Hi Nick:

I think my numbers may still be causing confusion, and I apologize.

> So, if labels appear no more than twice, the chance of a label 
> appearing elsewhere as a label would be (52*2)/1031 = 10%. How does 
> this compare with the chance of any word in the main text appearing 
> elsewhere in the main text?

I might be misunderstanding the comparison you propose, but my initial 
knee jerk fear is that the math equation wouldn't apply to the numbers 
based on what I was trying to offer folks with my list of labels with 
regard to the numeral 2 in the equation.

It's not that the labels only appear twice in total (even as labels).  
Several do appear more than twice and that is countable from the list 
in the initial email.  To be more precise, the list was of labels that 
appear as labels in not less than 2 different sections of the 
manuscript.  Some might also appear more than once in one of two 
sections where they appear.  Some might appear as labels in more than 2 
sections, as for example OKAR in bio-cosmo-pharma.  It is merely a list 
where you have at least two different graphic contexts to examine the 
same label word.  For instance, if one thinks he or she can read a 
plant id for one of these, you can go to its appearance in astro and 
see if there is any way the reading might hold up for instance.  If JH 
for instance identifies a plant name he can read using his method and 
it happens to be one of these 52, he can compare how it is used in 
cosmo or astro or whatever.  I personally believe it is also somewhat 
useful for ruling in/out parts of speech for romance languages by 
looking at the digraph pairs for the shorter such labels.

I might be misunderstanding, but I would hate to send folks off on 
creating ratios based on numbers I put up in a manner where I didn't 
precisely communicate exactly that which was being counted.  Numbers 
always make me worry here.  They imply more exactitude than is probably 
there.  The fact that there were 52 initially made me think "oh no!" 
that fact might be latched onto, when I know that there are the first 5 
or so in stolfi's list that have to be discarded because of the 
asterisks and where it's based on only the primary column, where some 
also may have a secondary suggestion...


Wayne

______________________________________________________________________
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying:
unsubscribe vms-list