Re: Statistics on copy(-n) --( WARNING: 33k attachment)

["Gabriel Landini" <G.Landini@xxxxxxxxxx>]

On 22 Jan 2002 at 23:30, Jacques Guy wrote:

> 22/01/02 09:31:38, "Gabriel Landini" the slave driver wrote:
> >This is interesting. Is it possible to get the words themselves, 

> The hardest part was formatting the output. The results are
> below. But please, (copy -1) (copy -2) (copy -3) try to guess
> what they are like before you look at the solution below.
 
Thanks!

> fq(1)
>   and=1 
and and?? The others are understandable if one deletes all the full stops. Unless one is "doubting": "Me 
and, and... everybody else"

Now what I find interesting is that for fq(3) onwards, the number of copies and the number of words 
remain pretty much constant. The average is about 4.5 copies per word. Is this expected? 
I am trying to think what would happen in a word-randomised text.

Cheers,
Gabriel