Is Python the Esperanto of programming languages?

Carl Banks imbosol-1048259891 at aerojockey.com
Fri Mar 21 16:45:51 CET 2003


Steven Taschuk wrote:
> Suppose you do not make that error, but you speak with an
> unfamiliar accent that leaves me uncertain about the vowel in
> "man"; what I hear is:
>    The m?n goes to the store.
> I can guess "man" instead of "men" because it's "goes" and not
> "go".  In this scenario the maligned sibilant carries useful
> information, and isn't superfluous at all.

Sure, there's a marginal benefit to it.  The morpheme has a tad of
meaning left, when the number of the subject can't be determined.
(E.g. "The sheep go to the barn.")  The ending is a little redundant,
but mostly superfluous.


>> However, even though redundant and superfluous lexemes are not
>> required, all natural languages have them, simply as a side effect of
>> language change.
> 
> Even if this explanation of redundancy's origin is correct, it
> does not speak to its utility; in particular, I am not convinced
> that a natural language can function with no redundancy at all.

Well, this is kind of a complicated question.

I don't think redundancy is required for it's own sake, even if it
helps sometimes.  I really don't think it adds anything more than a
marginal benefit to the language.

However, a language completely without any redundancy might have rules
too complicated to be worth it.  Or, if the rules are too strict
regarding redundancy, we not be able to internalize them, so that
being redundant sounds bad.  So redundancy might be necessary for
language for logistical reasons.

If a language could be invented that completely lacked redundancy, and
was simple enough to be internalizable, then I don't think humans
would have any problems communicating with it.


This is saying nothing about superfluousness, which can be dropped
without missing a thing.


-- 
CARL BANKS




More information about the Python-list mailing list