Re: VMs: Binomial distribution of word-length, etc...

[Nick Pelling <incoming@xxxxxxxxxxxxxxxxx>]

Hi Dennis,

At 23:15 08/01/2004 -0600, Dennis wrote:
> > I've thought for a while now that a good explanation for this might be a
> > combination of (a) an abbreviating private shorthand (which I suspect would
> > approach a kind of binomial distribution as the sample-length goes up, but
> > probably with a shorter word-length "peak"), and (b) a verbose cipher (to
> > move the peak sideways, ie to the right again).
> >
> > Unfortunately, I know of no sample statistics for (a)-like shorthand texts,
>
>         There must be examples of modern shorthand texts,
> although I don't know of any in convenient form.   I
> suppose there are Unicode fonts for Gregg, etc.

Modern shorthand is quite different from the kind I have in mind here 
(which would be a loose mixture of late-medieval Tironian-like ad hoc 
abbreviations & Radcliff's drop-letters-you-don't-need system (as mentioned 
on-list)), so the stats would be quite different. :-(

> Also, what about Japanese texts with differing degrees of
> kana-richness?

While modern Japanese does a fair amount of aggressive abbreviating (like 
"seku-hara" for "sexual harassment"), the size of (even any one of!) its 
alphabets puts it in quite a different kind of bracket from the size of the 
alphabet we see in the VMs (even with a verbose cipher!)

Cheers, .....Nick Pelling..... 


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