From: "Hal E. Fulton" <hal9000@...>
Date: 2003-08-21T06:23:25+09:00
Subject: Re: Punctuation as noise

----- Original Message -----
From: "Mills Thomas (app1tam)" <app1tam@ups.com>
To: "ruby-talk ML" <ruby-talk@ruby-lang.org>
Sent: Wednesday, August 20, 2003 3:52 PM
Subject: Re: Punctuation as noise


> My grandfather was a student of language and a teacher.  He told me once
> that punctuation (in writings representing English) exist to tell the
reader
> when to insert dead air (pauses or breaths).  These reveal information to
> listeners.

I agree in essence. I'd say however that commas and pauses both
serve to communicate much the same information. I'd argue there
is not a one-to-one correspondence. But that's OT, and if you
want to discuss it, let's take that part offline.

> According to Claude Shannon and the Theory of Information, pure
information
> is indistinguishable from noise.  Context is everything.

I agree again. I'm not using "noise" in a formal sense.

However, I do maintain that this:

  if (x > y)

is no more clear than this:

  if x > y

I prefer the latter. Many people prefer the former, but I believe
that is only a matter of habit. I don't see how parentheses
decrease ambiguity here.

In any case, one reason we construct grammars (and precedence
hierarchies) so carefully is IMO to avoid explicit disambiguation.

Sometimes we even live with the amiguity rather than use explicit
markers, as in the PLGS problem in conlangs.

Enough OT though.

Thanks,
Hal