Re: <acronym? [was: www-html archives ]

Alan J. Flavell (flavell@mail.cern.ch)
Thu, 31 Jul 1997 15:51:20 +0200 (METDST)


Date: Thu, 31 Jul 1997 15:51:20 +0200 (METDST)
From: "Alan J. Flavell" <flavell@mail.cern.ch>
To: Paul Prescod <papresco@technologist.com>
Cc: www-html@w3.org
Subject: Re: <acronym? [was: www-html archives ]
In-Reply-To: <33DFCE7B.56E9B16B@technologist.com>
Message-Id: <Pine.A41.3.95a.970731151503.133526J-100000@sp049>

On Wed, 30 Jul 1997, Paul Prescod wrote:

> Alan J. Flavell wrote:

> That's true. I'm trying to figure out where we really differ on the
> other parts of the issue? Do you really think that more people consider
> the word ACRONYM as being restricted to those initial-based
> abbreviations that are spoken as words? 

Well, my experience suggests that the term is more properly restricted
to combinations that are pronounced, but I'm not sure we ought to let
this decision be taken on our own anecdotal evidence, so I'm not about
to launch into anecdotes.

I've already argued the issue in a personal email which I've just
got permission to copy to the list, so I'll keep this part of
the answer brief and hang that email on the end.

I'd say there are two points.

Point 1 is can the HTML4.0 spec use the term "acronym" in a way that
_excludes_ letter-combinations being actually pronounced?  I see
no support for that view, so the draft needs reworking in that respect.

Point 2 is, if the term is used in a way that is intended to encompass
both the pronounced and the spelled-out combinations, then is that both
useful and logical?  I say it isn't very useful, since the speaking
machine still has no idea what to do with them (I clearly remember
a demonstration at CERN many years ago in which the speaking machine
made a heroic effort to pronounce IBM as "ibb'm").  And you still have
no markup for the remaining kinds of abbreviation.

And that is irrespective of whether we agree or disagree on the "proper"
usage of the term "acronym". 

So my conclusion, so far, would be that HTML would be better off with at
least <ABBREV> for any kind of abbreviation; if it then needs another,
more specific, markup in the area of initialisms/acronyms, I'd like to
see some more discussion of what that markup would be good for, and
then, one can perhaps make a choice from the terms that are available. 

If, in fact, they are intended for helping the speaking machine to
render the term, then for sure we need a distinction between the
spelled-out ones and the pronounced ones. 

> Or is our difference in the question of whether the HTML specification
> should be descriptive or prescriptive with regard to its use of the
> English language?

Oh no, Sir.  The HTML specification cannot do that for the English
language, although certainly a specification can choose to assign
specialised meanings to English words for its specific purposes.  The
majority seem to think that the ALT attribute is a "tag", but it isn't,
is it?  Irrespective of the everyday meaning of the word "tag" in
English. 

> I don't mind being prescriptive as long as we are not
> obfuscatory and "initialism" would be. 

That's fully accepted; I think I've addressed that issue below.

Here's the rest of my argument

------copy of mail-----------

Date: Thu, 31 Jul 1997 13:51:31 +0200 (METDST)
From: "Alan J. Flavell" <flavell@mail.cern.ch>
To: "E. Stephen Mack" <estephen@emf.net>
Subject: Re: <acronym? [was: www-html archives ]

On Wed, 30 Jul 1997, E. Stephen Mack wrote:

> Paul did quote from Peter Flynn's Acronym Server though, which
> had a very useful argument: the requirement that acronyms
> be pronouncable words is a U.S.-English intepretation not
> shared with British-English.  

That assertion completely astonishes me, I must say.  I am English and
have lived in Britain most of my life.  I had never supposed that the
looser usage of the term (viz. extended to strings of initials that are
spelled out rather than pronounced) was in any way a British-ism.  Could
be that my own experiences are atypical, which is why I'd prefer to rely
on what's in dictionaries and usage guides rather than relying on
personal impressions on this. 

Modern English Usage, 2nd ed., Fowler ed. Gowers, Oxford 1965,
had this to say:

"Another way of forming curtailed words is to combine initial
letters, a method now so popular, especially in America, that
a word - acronym - has been coined for it".  

There then follows a whole column of examples, every one of which is not
only pronounceable (which is, anyway, the implication of calling them
"words", I would say, although I see that you disagree) but is, in fact,
pronounced in normal usage.

> As further data in an informal straw poll,

With respect: can we base an HTML tag on an "informal straw poll"  if it
contradicts the usage recorded in dictionaries and usage guides?  If, in
fact, the balance of opinion and counsel from those sources goes against
me, I can accept defeat.  It is a pity to lose a valuable distinction,
but our language is littered with such losses, and one has to move on.

> I personally would agree on compromising with ABBREV,

That seems a better compromise to me, if (unlike the HTML3.0 draft) we
must have only one, and we can't agree well enough what an ACRONYM is. 

I would like to stress that I didn't mention the term "INITIALISM" 
because I personally like it - I did say "if such a term is needed"  and
made it clear that I realised this term is rarely used in practice. 
However, I have to say that if a tag is wanted that should unambiguously
cover those abbreviations that are formed from initial letters,
irrespective of other factors, then I'd say <INITIALS> is very clear and
definite.  But of all the terms that have been considered, if you just
want one tag in this area, then I'd suggest that <ABBREV> is the best
compromise.

> but
> I have to wonder what percentage of people use acronym in
> your sense vs. the percentage of people who use acronym in
> the sense defined in the HTML 4.0 draft.

I confidently state that _nobody_ uses the term in the sense that
appears to be described in the HTML4.0 draft, since the term "acronym"
is never used in a way that _excludes_ combinations that are in fact
pronounced.  That part of the HTML4.0 draft is a mistake; on that I am
quite definite.
 
> (I'm a descriptive linguist by training, not a proscriptive
> linguist.)

I try to avoid the tendency to being prescriptive, but it still
saddens me when a term that had a valuable distinction of meaning
gets wasted by being confused with some other term.  But language
develops, and not all changes are for the worse  ;-)

best regards