LINGUIST List 7.1480

Mon Oct 21 1996

Disc: Natural language

Editor for this issue: Ann Dizdar <>


  1., Re: 7.1463, Disc: Natural language
  2. "Jens S. Larsen", Re: 7.1463, Disc: Natural language
  3. Jonathan Centner, Disc: Natural language

Message 1: Re: 7.1463, Disc: Natural language

Date: Thu, 17 Oct 1996 14:42:52 -0800
From: <>
Subject: Re: 7.1463, Disc: Natural language
Having read many of the interesting contributions on how to discern
natural languages, i feel reminded very much of a few issues from

If i have understood the matter correctly, it has been established (by
Goedel et al) that in any given open system, i.e., one that contains a
potentially infinite number of elements and operations, there may
occur statements that are "true" but unprovably so and also some that
may be unprovably "false" - much to the disappointment (in both ways)
of those mathematicians who until then believed in the existence of an
algorithm - searched for but not yet discovered - for distinguishing
with certainty "true" from "false" statements.

Another issue: bivalent or Aristotelian logic is nowadays being seen
(at least by some) as one of many possible logic system - quite like
Euclidean geometry, once perceived to be the only possible geometry,
was relativised by Riemann. ("Fuzzy logic" is the term for
incorporating this new understanding into statisitcs and computer
applications.) This does not mean Euklidean geometry is suddenly
"wrong"; it is now perceived as a subset instead of the whole set and
can thus only be applied correctly in the appropriate context.

In which way are these matters relevant to the discussion on natural
languages? There are several connections:

a) Looking at parts of languages, we surely would agree that what may
be "unnatural" in a specific context may be "natural" in another
context - witness for (just one) example the use of artificially
created acronyms and of new meanings assigned to existing words, both
in regards to the use of computers and the internet, in the young
generation's "natural" speech. Context is the decisive parameter here
for deciding the question "artificial or natural?"

b) We must never overlook that just because we are looking for some
"thing", "distinguishing parameter", or "algorithm" it does not follow
that what we look for actually exist: there may be no "ultimate"
parameter or algorithm available to ascertain what is / is not a
natural language. Instead we may have to settle for "more or less
natural" - a sliding scale - for example, Esperanto as conceived by
Zamenhoff would be at one end of that scale, as it is spoken today it
would be somewhere along the line closer to English, however...

There may be more valuable analogies... :-)

Lest i be misunderstood: it is *not* my intention to judge the debate
on natural langauge as useless - far from it! There are many good
reasons to engage in this debate - enjoyment perhaps not the least of
them - and in that spirit i added those few points to this debate :-)

Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 2: Re: 7.1463, Disc: Natural language

Date: Fri, 18 Oct 1996 11:20:49 +0200
From: "Jens S. Larsen" <>
Subject: Re: 7.1463, Disc: Natural language (Ivan A Derzhanski) wrote in Vol-7-1463 (3):

> Anecdote has it that Charles Darwin's students once decided to play a
> joke on him. They caught and dismembered several insects and, taking
> each body part from a different donor, glued together a hitherto
> unseen creature, which they brought to Darwin. `We caught this bug in
> the garden,' they declared. `Can you tell us what it is?' Darwin
> eyed the fabulous monster briefly. `Did it hum when you caught it?'
> he inquired. `Yes,' the students replied, nudging one another.
> `Well, then,' Darwin said, `it is a hum-bug.'


> I don't find it hard to believe that `the creators of all artificial
> languages have failed to imitate [...] the properties which make a
> language a natural language' (JL 7-1350). The universal properties
> which (we believe) characterise natural languages, which must be
> present in them in order to make them acquirable by the human mind,
> and which it is our errand (as linguists) to explicate and motivate
> are not at all trivial, and no one has ever had exhaustive conscious
> knowledge of them (whether anyone ever will is a separate issue). How
> likely is it, then, for any creator of an artificial language to have
> imitated them flawlessly?

That's likely to be irrelevant. Rules that comply with UG are easier
to think of than some which don't. When people try the language in
practice (as Zamenhof did with Esperanto years before he went public),
any rules that do not comply with UG will either be replaced, or the
whole language will. Volapu"k was replaced with Esperanto, but we
don't know enough about UG to say whether it had anything to do with

> Do we really know for a fact that there is nothing about the
> structure of Esperanto, Volapu"k, Klingon or Lojban that would allow
> a Martian linguist (or, for that matter, a Terran one) to identify
> them as humbugs?

For that to be asked, we have to know whether we distinguish properly
between natural and artificial in the first place. The students in
the Darwin anecdote were not capable of presenting him with a living
bug. Nowadays we can do that, through genetic engineering. Does this
mean that Darwin would have had to take his students seriously today?
Why, and why not?

> Of course, the proof of the pudding is in the eating, and of the
> natural (or quasi-natural) language in the learning. If a feature of
> a conlang (even one which is not found in its sources) can be acquired
> natively, then it is obviously not at variance with UG. But how can
> the creator predict whether this will be the case?

He can't, and he doesn't have to. He can let the language loose
outside the laboratory and see what happens.

> Will his intuition (as a native speaker of one or more natural
> languages) suffice? There doesn't seem to be anything to
> prevent one from proposing, say, a language in which a clause
> is negated by reversing the linear order of its words, or one
> in which inflexion depends on whether the number of phonemes
> (or letters) in the stem is prime or composite.

Except common sense, of course.

> Admittedly those are extreme examples of very obviously `unnatural'
> features. But how can we draw the line in the general case without
> actually running an experiment on real live babies?

What is _any_ natural language other than an experiment on real live

> Incidentally, if Esperanto is demonstrably natural by virtue of being
> natively acquirable, why aren't Esperanto data appealed to for support
> of any theory of phonology or syntax? (Or are they?)

Considering the confusion that reigns supreme in the discussion about
naturalness and artificiality in language, I don't find this very

Jens S. Larsen, lingvist (BA in spe)

<"";> * <"">
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue

Message 3: Disc: Natural language

Date: Fri, 18 Oct 1996 18:56:32 CDT
From: Jonathan Centner <>
Subject: Disc: Natural language
Ivan A Derzhanski wrote:

> Do we really know for a fact that there is
>nothing about the structure of Esperanto, Volapu"k, Klingon or Lojban
>that would allow a Martian linguist (or, for that matter, a Terran
>one) to identify them as humbugs?

Matters of obstruency spring to mind. Tolkien travelled down this road
when he contrasted his inventions of Orc-ish and the Elven tongues; to
speakers of English, _Derzhanski_ might be a hyperenergetically evil
creation of Morgoth or the mind of Gene Roddenberry, but Slavic babies
will suckle on it, and because the surrounding phonological stream
allows it.

Jon Centner
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue