Yet Another Case Question

Steven Taschuk staschuk at telusplanet.net
Mon Feb 24 06:41:01 CET 2003


Quoth Erik Max Francis:
> David Mertz wrote:
  [possibility of inconsistent case usage in case-insensitive languages]
> > Certainly people indeed do this.  That part I can't argue against. 
> > Tell
> > me why this is BAD, again?  On the face of it, allowing such variation
> > seems like the very purpose of case-insensitivity.
> 
> Is it good that there will now be inconsistent spellings, with each
> programmer to a project contributing a different style?  Doesn't that
> make it harder for everybody to read it?

If Mertz's analysis or something like it is correct, this last
question, which I assume you meant rhetorically, is actually a
question of substance.  Taking his analysis as given, and naïvely
using a computerlike model of information processing, we might
speculate thus:

Mertz's visual/eidetic folk look up words in their mental lexicons
by using the orthographic representation as a lookup key.  If the
word is supplied with the wrong case, that lookup fails and
various case conversions must be tried to find the right lookup
key.  Those of this type thus find inconsistent use of case
obnoxious, since it makes it harder to read.

His verbal/phonetic folk, on the other hand, look up words by
first converting from orthographic representation to phone[tm]ic
representation, then using the latter as the lookup key.  Case
distinctions are washed out in the first step, and any use of case
is equally easy to read.  (And if those of this type do not store
case information in the lexicon at all, they'll find insistence on
consistent case obnoxious, since it makes it harder to write.)

(MOnStRosiTiEs are, I expect, harder to read for everybody, so my
suggestion that the verbal/phonetic finds "any use of case"
equally legible is overstated.  But the difficulty reading such
MOnStRosiTiEs can be explained as arising when parsing the visual
stimulus into an orthographic representation, rather than during
the lookup.  Such problems wouldn't necessarily also arise with
sensible case mixtures such as all-lower, all-upper, title-case,
and camel-case, and don't speak to the question of consistency.)

-- 
Steven Taschuk                                                   w_w
staschuk at telusplanet.net                                      ,-= U
                                                               1 1





More information about the Python-list mailing list