Conlangs: Languages or "Art"

[lojbab (Bob LeChevalier)]

And Rosta writes:
>What is a "real" language, and why is it impossible to invent one?  Why
>is a language without a community to develop it not a language?

Here is the typical linguist's answer to your question:

My Websters gives the following (paraphrased) definitions of "language"
to which I comment indentedly.

1. human speech

   Everybody not mute 'invents' this kind of language, so this is not what
   we are talking about.  Besides, most 'conlangs' exist only in written form.

2. A system of vocal sounds and combinations to which meaning is attributed.

   Well, we now have a 'system', which most conlangs are, but this is still
   oriented around vocal use.  Then also, see comments below.

3. The written representation of such a system

   The closest most conlangs comes to being a 'language'.  They are written
   representations of systems that theoretically COULD be used vocally.

4. any means of expressing or communicating

   Even broader than 1., this includes body language, and "Art".

5. all the vocal sounds, words, and ways of combining them common to a
   particular nation, tribe, or other speech community. e.g. the French
   language

   The example shows that this is what we usually mean by 'a language'.
   Note most strongly the words "common to" and "speech community".
   Linguists do not accept as a 'language' worth study anything that does
   not have a speech community, and are generally interested in the features
   common to the community, as opposed to individual idiosyncrasies.

6. the particular form or manner of selecting words characteristic of a person
   group or profession  e.g. the language of teenagers

   This is descriptive of some kind of real language usage, and is clearly a
   usage subset of a language.

To which we add one definition of "code":

A system of symbols used in information processing in which letter, figures,
etc. are arbitrarily given certain meanings.

   Comparing this definition with 2. (and by implication 3.), an invented or
   arbitrary system of communication is a code, rather than a language.  The
   distinction of 2. is that meaning is attributed, not assigned.


A conlang invented by a single person inherently must be arbitrary in
assigning meanings.  If you invent a conlang, you know what the words
mean because you decided the meanings.  Furthermore, you almost
certainly define those meanings in terms of another language, usually
English.  Thus most linguists claim that conlangs are nothing but
encoded English.

When multiple people of differing backgrounds get together, you have
some possibility of moving beyond the definition of code, because each
person's meanings associated with individual words is different, etc.
Thus while the meanings may be arbitrary by assignment, there has to be
some attribution of meaning; i.e., person B has to understand what
person A meant by saying that "gsfdtrhhs" >means< English "set".  Is it
identical to "set"?  Can it be used in all places where "set" is used?

However, even there, person A is assigning, even if person B is
attributing.  B has the job of figuring out what A was assigning.  Only
if the community of speakers gets large enough that no person ever
really 'assigns' meanings, but rather chooses words from the available
set s/he shares with the listener so as to ease attribution of meaning
by the listener.

>From the start, we have stressed in Lojban that it is the speaker's
responsibility to express things in terms understood by the listener
rather than vice versa.  By doing so, we offer hope that in a small
community of non-fluent speakers, people will get in the habit of NOT
assigning Lojban words their English equivalent meanings, but instead
will try to choose Lojban words that will be attributed the meaning
SELECTED from the available common set.  I've chosen my words carefully
here - the idea is that the meanings of words must be outside the
control of individuals, or you have a code.  You must be selecting from
some previously EVOLVED set of meanings that is common to the community
of speaker and listener.

This emphasis is our only real hope of making a conlang into a real
language inside of a generation.

Now the linguist goes beyond the explicit definitions of Webster, and
will state that the conlang remains a 'code' rather than a language, as
long as the words are being interpreted in terms of another language.
The most form of this position requires that a language have speaker and
listener 'thinking in the language', rather than translating everything
to English, in order for it to cease being a code.

Most linguists are not this generous.  They will call such a language a
'pidgin', which is not quite a language, but also not a code anymore.
When a pidgin has native speakers and starts to evolve, it becomes a
"creole", and only in the more advanced stages of creolization, where
there is a substantial body of speakers living their everyday lives from
birth in the language and dominating the speech community (over
non-native speakers who will always have the impurity of thinking in
their native language to at least a minimal extent), will most linguists
term it a "language".

Esperanto is an early creole - there are few native speakers and they do
not dominate their speech community.  I doubt that any other conlang is
any more than a minimal pidgin (including Loglan/Lojban - since I have
heard no reports of anyone else 'thinking in the language', except my
wife who once dreamed a Lojban conversation.  I have for brief moments
'thought in the language' when writing original material that I have
previously decided will not be translated into English.  This seldom is
for more than a sentence or two because I then run into the need for
some vocabulary word that I don't know or may not have been invented
yet, and I have to fall back to concocting a word based on
English-associated meanings.  I expect, though, that we will have people
thinking in and speaking Lojban fluently within a year or two - probably
after a dictionary is done.)

Most linguists, then, do not even consider Esperanto a language.  Of
course they have chosen a definition that makes it especially hard for a
conlang to achieve.  I use a slightly milder definition of language that
ensures that we have gone beyond meaning assignment to meaning
attribution:  When the language inventor(s) teach a language to others
who have not interacted with each other, and then THOSE teach the
language to still others who also have not previously met, and those 2nd
generation students get together and can converse readily (perhaps even
fluently), then you clearly have gotten to the attribution stage of
language.  With Loglan/Lojban we are still at the 1st generation stage,
though a couple of people like Nick Nicholas and John Cowan are close to
the skill level needed to teach to 2nd generation speakers.  (Of course
these two can only barely be considered in the first generation, since
both are contributing sustantially to the basic language design.  But
still, it can reasonably be said that the nature of most of their
contributions are of the language growing kind (a la Chaucer and
Shakespeare) rather than of the language inventing kind (a la Zamenhof).

All of the definitions posed have some suggestion of an extended
community consisting of more than the language inventor.  This is why I
say that a language cannot be invented by an individual - it has not
been invented AS A LANGUAGE until the community has taken it out of the
inventor's hands and made it THEIRS.  With the creation of Lojban we the
Loglan community have in effect done this, and in the last few months
the necessary second step of the community taking the language out of my
hands has started.  I thus know that Loglan, among conlangs, will become
a language as Lojban.


Now.  How many of you who invent conlangs can claim to have met even my
weak test?  How many have thought "in language" as opposed to in English
for a substantial period?  Moreover, how many can claim that even one
person other than the inventor either thinks in the language or speaks
the language fluently?

Anyone can encode the Paternoster in a cute kind of cryptography that
says nothing about languagehood.  Language is about communicating,
and it is not real until it is used to communicate new ideas that are
not previously expressed, and sometimes not even expressible, in another
language.


>Another term for creating something "for the idle intellectual
>entertainment it provides" is Art.  Is art really not worth bothering
>with?  Personally, I'm a big fan of art.

Such a definition is so broad that it is not worth arguing with (and is
irrelevant to the discussion anyway).  I do a lot of things for idle
intellectual entertainment and would not call very many of them "Art",
either with capital or small letter.

Now, reread what I said, noting new emphasis:

  >>and OTHER THAN for the idle
  >>intellectual entertainment it provides, such efforts aren't worth
  >>bothering with.

I clearly make no claim about the value of idle entertainment.  I am
talking about value in other terms:  some type of goal or use to which
the conlang is to be put.  I do not deny that mental masturbation has
some value, which value grows when extended to intercourse between two
(or more).  But when I am spending as much time working on the language
as I have for 4 years now and interacting with hundreds or even
thousands of people whose only interest was at the idle intellectual
entertainment level, my standards for linguistic entertainment have
risen quite a bit.  So at the least take my comments on this issue as
reflecting my personal tastes.  Though as Nick, John and perhaps others
can tell you, there is a lot of depth to plumb in language once you get
over the word-and-rule-making stage.

In any event it is plausible that language can be art, but other than
meaning 4. above, art is not language.  I presume that this forum is for
discussion of constructing languages, not creating art.

As an example, I have a goodly quantity of 'Lojban' poetry written by a
published poet who clearly takes considerable pride in his poetic
skills.  What he has written sounds moderately euphonious, and it is
usually composed only of Lojban words.  But it is rarely grammatical
either because he takes 'poetic license' or else just doesn't know the
grammar for the construct he wishes to use.

However, the poetry is not really Lojban; if it is in any 'language' it
is in a rather extremely encoded English, in that if you do a
word-for-word translation, you may be able to figure out his intended
meaning from the resulting English expression.  But usually not even
then.  This may be "Art", but it is not language, because the reader
cannot attribute meaning to the words with any degree of confidence.
----
lojbab = Bob LeChevalier, President, The Logical Language Group, Inc.
         2904 Beau Lane, Fairfax VA 22031-1303 USA
         703-385-0273
         lojbab@snark.thyrsus.com

For information about Lojban, please provide a snail-post address to me
via mail or phone.  We are funded solely by contributions, which are
encouraged for the purpose of defraying our costs, but are not mandatory.