Word distri., was Re: VMs: Blanks: trying to summarize...

[knoxmix@xxxxxxxxxxxxx]

    	On 6 Mar 2004 at 11:50:50 +0100
          Elmar wrote:

    IMHO, disregarding the entropy factor will lead you onto a wrong 
    path.
    Word and syllable repetitions are an outstanding feature.

    Any scheme which reduces "word" length while _not_ increasing 
    redundancy will probably fail to match the VM features, and perhaps 
    should be discarded from the start.

    Cheers,

       Elmar



Hello Elmar,


Thank you for the comment. 

Of course, languages do not have fewer words than syllables but what 
I see in Voynichese is an abnormal number of similar words rather 
than word repetition. If I were attempting to convert modern 
languages to Voynichese I would expect to get a different output, 
entropy-wise, from a language that uses "habba habba nikki babba" 
than I would from a language that uses "every little bit helps" to 
mean the same thing. I do not propose to do that. If I did I would 
use old versions of languages: Middle English and its equivalents in 
addition to Latin. I have, for entertainment only, made Voynichese 
"speak" modern English -- of a sort. That would give misconceptions 
of Voynichese if taken seriously. The same is true in reverse.

I do not know whether all the schemes listed would tend to increase 
redundancy but if I am looking for a cipher I cannot test a concept 
before I have a concept. It is simpler for me to choose from a field 
of shortening mechanisms than it is to choose from a field of entropy 
reducers. I am not assuming that either field covers all the bases. 
The assumption is that an initial examination leads to a little 
concept and a little concept might lead to a better examination. Most 
will not. Some will. 

Ciao .......... Knox

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