Re: VMs: left & right word entropy

[Jacques Guy <jguy@xxxxxxxxxxxxxxxx>]

23/12/2003 12:57:06 PM, Gabriel Landini <G.Landini@xxxxxxxxxx> wrote:

>I understand that his findings may be interpreted to point to a lack of 
>grammar.

No. Read on.

>I just can't imagine a natural language in which the words do not have some 
>determined order in the text flow (Jacques?).

Take his example of what precedes "am" in English.
Fair enough. Take Latin now. What precedes "sum"? Just about
anything. Can you think of any word at all in Latin that
calls for a particular word preceding it or following it? 
Or even a small set of particular words? 

Try, and keep trying until you do.

Then wake me up when you have.

Latin is not the only one. I am sure the same is true of Lagu,
an Austronesian language of the Solomon Islands (but I don't
have enough text in Lagu, and if I had, I wouldn't feel like
typing it in). I only mention Lagu because, unlike Latin,
and like most Austronesian languages, it is uninflected.
So this property of texts (left = right) does not even
tell us whether the language is inflected.

It MIGHT tell us that word-order is free. 

But again, that depends on whether the text has been 
correctly segmented into words. Take an English text, 
remove the punctuation and the  spaces, insert word 
breaks anywhere you feel like, at random, or after 
every 'e', or whatever. I bet you that left 
entropy = right entropy.

I know that, strictly, I should back up my claim about
Latin, and I have downloaded Cesar's complete "De Bello
Gallico" last week. I prefer to spend my time designing 
"Frogguy for Rongorongo".



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