Re: LUNATIC again

On Wed, 4 Nov 1998, Logical Language Group wrote:

> So to return to a definition of language.
> 1. A language is a means of communication.
> 2. To exclude computer languages, we have to restrict this to communications
> primarily between biological entities (if not actually restricting to humans).
> It is clear that computer languages have little in common with human languages.
> 3. In order to have commmunication BETWEEN, there must be at least two people
> who use the language - a speaker and a listener.  Linguists tend to go much
> further, and say that this communication must be bidirectional, must be
> fluent, must be passed on across generations.  Linguists exclude not merely
> artlangs from the concept of "language", but pidgins and creoles that may
> be spoken by large populations.  I may disagree with that definition, but
> I had better understand it when trying to communicate with linguists.
> 4. Linguists also distinguish between "languages" and "codes", where the
> latter may differ in surface form from a parent language, but where the
> semantics of the "words" in the code are essentially unchanged from the
> parent language.
> 5. Linguists also distinguish between "languages" and "cants" or "jargons"
> which are incomplete languages that have distinguishing semantics in
> some areas of the lexicon, but are moreor less standard in the rest
> of the lexicon.
>
> I  will use the above definition, noting that any linguists here can probably
> run me over the coals for various misstatements therein.  But I think it
> conveys that which is essential to my points and is not too far from what
> linguists accept.
Well, you talk about linguists as if we all used the word "language"
in the same way, but that's not the case.

In *my* branch of the linguistics biz, we tend to specify "natural human
languages" as our object of study.  "Natural" in the sense of naturally
evolving, as opposed to consciously developed languages - conlangs, computer
languages, predicate calculus, specialised cants, and perhaps pidgins
(but NOT creoles) - and "human language", as opposed to animal
communication.  It seems that the linguists you've been talking to are using
"language" to mean what I mean by "natural human language".  I don't care for
that usage, inasmuch as it differs significantly from the everyday sense of
the word "language".  "Natural human language" is less open to
misinterpretation.  But to each his own.  As long as you define your terms...

With regard to your above criteria 1-5, point 3 (language must involve
at least two participants, be bidirectional, etc.) mostly reflects the
attitude of those linguists who work in functionalist frameworks (using
the term "functionalist" broadly).  In more 'formalist' frameworks,
such as Principles & Parameters, Minimalism, and generative linguistics
generally, point 3 is generally downplayed:  Language is defined as
a system of knowledge ("competence") which is in principle separable
from actual instances of communication ("performance").  Thus you would
find disagreement among linguists as to whether point 3 belongs to the
*definition* of "language" or not.

However, in my experience, the question of whether language can or
cannot be divorced from actual communicative acts, while important in
principle, is usually ignored in practice.  Most linguists who call
themselves functionalists nevertheless make use of elicited sentences,
grammaticality judgements, and other kinds of data which do not involve
spontaneous language use.  Unfortunately, many formalists - and I say
this as a formalist myself, a Chomskyan even (ooh! *shudder*) - do not
make use of actual examples of spontaneous language use, but base their
theories entirely around elicited data and speaker judgements.  Myself, I
try to use both kinds of evidence, respecting the primacy of spontaneous
textual material, while using elicited data to confirm or discount my
hypotheses.  Nevertheless, I adopt the Chomskyan premise that language
is best viewed as a system of knowledge rather than a system of
behaviour (insofar as they can be separated), and thus I would not
include your point 3 in a formal definition of "language", although I
might use it as part of an informal characterisation of language.

> Given this definition, it seems clear that most conlangs are not languages
> and are not intended to be languages, in that their inventors never seriously
> expect them to be used by two or more people in communication, much less
> learned by a community. This seems especially true of artlangs. I don't see that
> this is necessarily a disoparagement of artlangs or artlangers - it merely
> recognizes that the goals are different from those linguists identify as the
> basic nbature of language.  I have no trouble forswearing criticism of anything
> labelled as an "artlang" because in using that label, the object is distinguished
> from that which is studied by linguists.
Here I find the term "model language" useful and appropriate.  A model
language is and is not a language in the same way that a model airplane
is and is not an airplane.  Many artlangers have drawn an analogy between
what they do and what model railroaders do, and that analogy certainly
fits *my* interests and goals - namely to develop and describe as
thoroughly as possible a hypothetical but naturalistic linguistic system,
while not intending that system to actually be *used* for anything, and
that it can never be as complex or complete as a 'real' language.

Matt.