[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: VMs: split words
At 13:29 03/09/2004 +0100, Marke Fincher wrote:
I've done a little experiment around the idea of split-words,
or that some of the spaces in the VMs are misdirectional.
For each unique five letter string that occurs within words
(there are 8400), I've looked to see if that same sequence
appears elsewhere but with a space inserted. It turns out
that in 3000 (36%) a split-sequence does occur. In many
cases the same 5 char sequence is split in 3 or even 4
different ways (a few examples are shown at the end).
dolch:7 d.olch:2 do.lch:2 dol.ch:29
holch:3 h.olch:1 ho.lch:2 hol.ch:215
hopch:18 h.opch:1 ho.pch:3 hop.ch:1
hotch:26 h.otch:1 ho.tch:10 hot.ch:3
lolch:2 l.olch:29 lo.lch:2 lol.ch:20
I think you may have misinterpreted the character "*" - where this occurs
in the transcription, it means "an unreadable or ambiguous character", but
you seem to have interpreted it as an alternative kind of space (which
would be wrong for your tests). So, your "h-" words might really be (for
example) "c*h-" words, and hence are actually longer than five characters.
Can you check this, please? BTW, apart from this query on your
transcription, I do like where your line of experimental reasoning is
Cheers, .....Nick Pelling.....
To unsubscribe, send mail to majordomo@xxxxxxxxxxx with a body saying: