|
Since |
Ferreira's Home Page |
|
|
Perl Bit #3: Right? Rgiht? |
Once I read that some researchers have found out that when reading, the first and last letters of words are the ones that count. It means you read the following piece of text without much hassle.
I aslo beelive that, whlie lgagauens can have
eiiinecefcfs and dcieiifecens, the lnagauges telvsmeehs
are enleasstliy amroal. Lgnaugae is not the lveel at
wchih we shluod ecofnre "good tohtughs", if we wnat
our lagugane to be mxlamialy uesufl. You can't efncore
milroaty by sntyax. In Eslginh it is just as easy to say
"bsels you" as it is to say "fuck you". You may agure
taht in Perl it's eseair to use the vreb "besls" bsucaee
it's bilut-in, but in acautl fcat, Prel lets you dienfe
"fcuk" any way you coshoe. You can aslo "goto hlel"
if you lkie, wcihh wlil of cuosre work beettr
if you've dfneeid the leabl "hell".
(quote from "The Culture of Perl" by Larry Wall:
http://www.perl.com/pub/a/1997/wall/keynote.html)
A short Perl script can help one play with this.
Pick the permutation subroutine of
Perl Bit #2,
together with the following code
sub shuffle {
my $word = shift;
if (length $word > 3) {
my @word = split '', $word;
@word = ($word[0], permutation(@word[1..$#word-1]), $word[$#word]);
return join '', @word;
} else {
return $word;
}
}
s/([[:alpha:]]+)/shuffle($1)/ge;
print;
Stuff some lines of texts into a file, then invoke Perl over
it:
perl -n mix.pl < input.txt
and you can find out for yourself if the assertive at the
start of this bit makes sense or not.
On my account, I think that it is clear that first and letter words are not the whole truth. If the text is rich with common words, the reading is facilitated. But rare and long words call for more careful reading and then one's mind can get confused with the shuffling of letters. I also found that it is much harder (for me, at least) to read a scrambled text in English (which is not my native language) than in Portuguese. So that immediate recognition is directly related to the skill of the reader.