Theiling Online    Sitemap    Conlang Mailing List HQ   

Yûomaewec: Example and evaluation

From:Adrian Morgan <morg0072@...>
Date:Wednesday, August 14, 2002, 5:40
I've done a statistical investigation of letter frequency in English
written with my "Yûomaewec" spelling system. I used one fairly long
text for the investigation, namely an extract from a conversation
about consciousness and artificial intelligence.

A text file containing my example is attached.

To qualify as definitively more frequent, a letter must occur (a) more
often in a given length of text, and (b) in more words. That is, I
measured frequency with and without a bias toward common words. Where
the two do not correspond, I have generally taken it as a limitation
of my sample space and presented two or more letters as having
equal frequency.

This being the case, {û} is unusually biased, in my sample, to common
words (such as "to" and "you") but didn't occur in many /different/
words. Whereas {l} is unusually biased toward appearing in words that
only occur once, while being comparatively rare in frequent words.
The commonality of {û} may be roughly compared with {v} and the
commonality of {l} with {k}, taking both metrics into account.

Most common to least:

(1) 'e'
(2) 'i'
(3) 't', 'n'
(4) 'o'
(5) 's'
(6) 'z', 'h', 'u'
(7) 'a'
(8) 'k', 'l'
(9) 'd', 'm', 'r'
(10) 'à'
(11) 'c'
(12) 'û', 'v', 'w', 'b'
(13) 'y'
(14) 'f', 'p'
(15) 'ù'
(16) 'g'
(17) 'ò', 'j'



John Cowan <jcowan@...>