Re: Stack-based syntax (was: affixes)

On Wednesday, February 23, 2005, at 07:53 , Mark J. Reed wrote:

> On Tue, Feb 22, 2005 at 06:38:22PM +0000, Ray Brown wrote:
>>>> I do not see how a lexical category like verb should act as an
>>>> operator.
>>>> Isn't it something more like: singing, John, NOM = John is singing?
>>>
>>> It is a question of semantics.  In Fith, verbs are indeed operators.
>>> Intransitive verbs are unary operators, transitive verbs are binary
>>> operators.
>>
>> You're right. Once again the Fith verb is combining both lexical meaning
>> and operator.
>
> Which is not as odd as you make it out to be, IMO.
I am not saying it is odd - what I am saying is that I would do things
differently.

>  If the verb has lexical
> semantics, then don't the operators in arithmetic RPN likewise have
> lexical
> semantics?  It's not "3, 2, ADDITION, PERFORM",
Indeed not - otherwise we would finish up with an _infinity of operators_,
since PERFORM also has a 'lexical meaning' in the sense that ADDITION has
- and we cannot have a stack of infinite size.

> it's just "3 2 +".
> And certainly there must be some  "lexical" information in the - and ÷
> operators, which enforce an ordering on their operands?
In the case of + the ordering is irrelevant, tho in the case of - it is
not. But the ordering is surely the result of stack structure itself, not
the operator per_se. But the point is that operators like + - * / etc are
simply not of the same order as the literals 56 78.9 2563 0.2765 etc etc.
The latter convey quite different meaning and form an open set - in the
case of numbers we know that the set is infinite. The operators are a
closed set.

But consider the effect of the operator. If it is a binary operator, as
the stack is being evaluated, the top two numbers are popped off the
evaluation stack, an evaluation made, and the result of tha evaluation
pushed back onto the stack; so for example, after 3 2 + is evaluated we
are left with just 5 on our stack. The literal 5 is of the same type as 3
and 2; the notion of additionality is not there in the resultant
evaluation. But when a Fithian verb operates upon its arguments, surely
the meaning of the verb forms part of the resultant meaning?

What I am suggesting is that a small, closed set of operators act upon an
open-ended set of sememes, and that the evaluation of the operations
results in another sememe without retaining any trace of the 'meaning' of
the operator. Assuming that most operators are binary (with perhaps one or
two unary), the top two sememes would be popped off the stack, evaluated
according the operator, and the result of that evaluation (itself a sememe)
  is pushed back onto the stack.

Note that by 'sememe' I do not mean 'a minimal unit of meaning'. Rather I
am using 'sememe' much in the same way that Claudio Gnoli uses the term in
his description of his interesting conlang, Liva: "Liva texts don't
consist of sentences in the traditional sense, but rather of meaningful
units called sememes. A sememe can consist of one or more
words......Longer sememes consist of an operator with its arguments."

In the scheme I am outlining, I would say rather "Longer sememes are the
result of the evaluation of an operator upon other sememes."

Of course, all that we would see (and hear) us only the communication in
RPN format that Alien A presents to Alien B to evaluate. The evaluation
stack is formed in Alien B's brain as s/he received the communication
stack. Since the stream that Alien A is communicating is in RPN, Alien B
is able to evaluate _as it is being received_; but what goes on in his
brain is hidden from us so we probably will never be able to perceive the
'longer sememes' as our Aliens do; we shall have use a more
'human-friendly' analog.

>>> Perhaps more "part-of-speech" thinking involved here
>>> than there should be.
>
> I think rather that y'all are demanding too much syntactic rigor from what
> is, after all, intended to be a natlang. An alien natlang, but still a
> natlang, not a loglang or engelang.
But that IMO is simply being anthropocentric. We simply have no idea what
an alien natlang would be like!

To contrast an alien natlang with a loglang or engelang has no meaning a I
see it - it is contrasting an unkown with two sorts of human artifacts. If
I were asked to create an alien language, I would try to use a model as
different from human language as possible. Though whether I would actually
go for a stack-based system is another matter.

But personally I think when attempting to model something alien to human
language, then rigor is necessary if we are to attempt to stop
anthropcentricity creeping in.
==============================================

On Wednesday, February 23, 2005, at 08:15 , Jörg Rhiemeier wrote:
>
> Hallo!
>
> On Wed, 23 Feb 2005 18:38:42 +0000,
> Ray Brown <ray.brown@...> wrote:
[snip]
>> I agree. It is always easy to spot weaknesses in some pioneering effort.
>> Fith is an attempt to produce a real alian, non-human language.
>
> Yes, and it is not a bad attempt.  Its merit lies in being the first
> language (as far as we know; there may be older stack-based conlangs
> that have never been published) departing from the kind of syntax
> human languages use by being based on a stack-based syntax,
Another conlang that departs from the syntax of human languages is Lin,
which is supposed to represent the language of alien beings who
communicate by telepathy. But that is not stack-based - tho Srikanth's
'internal & external cements' do behave as operators.

> and that
> will remain no matter whether later conlangers will come up with more
> elegant or less SAE-influenced stack-based languages.
Yep.
>>> I have seen "alien" languages that don't
>>> look more alien than, say, Old Albic.
>>
>> Indeed, so have I  :)
>
> I could have said "Welsh", "Japanese" or some other natlang name
> instead of "Old Albic"... there are "alien languages" that are
> little more than relexes of English.
Very true - I suppose that is slightly better than the many species of
English-speaking aliens that see o inhabit the universe   ;)

==============================================
On Wednesday, February 23, 2005, at 08:25 , Jörg Rhiemeier wrote:
>
> Hallo!
>
> On Wed, 23 Feb 2005 14:53:32 -0500,
> "Mark J. Reed" <markjreed@...> wrote:
[snip]
>> Which is not as odd as you make it out to be, IMO.  If the verb has
>> lexical
>> semantics, then don't the operators in arithmetic RPN likewise have
>> lexical
>> semantics?  It's not "3, 2, ADDITION, PERFORM", it's just "3 2 +".
>> And certainly there must be some  "lexical" information in the - and ÷
>> operators, which enforce an ordering on their operands?
>>
> This is a good point.  There is certainly a difference in semantics
> between "3 2 +" and "3 2 -", and why not consider that lexical?
> After all, you can define as many two-argument functions as you like.
See my reply above. The difference between "3 + 2 +" and "3 2 -" is that
the written operator tells us how to evaluate the two numeric literals to
obtain a new numeric literal, namely 5 or 1. The meaning of the operator
does form part of 5 or 1.

I would expect a relatively small set of operators - this is, maybe, like
the quest which has appeared on this list from time to time for the
"minimum number of verbs".

[snip]
>> I think rather that y'all are demanding too much syntactic rigor from
>> what
>> is, after all, intended to be a natlang. An alien natlang, but still a
>> natlang, not a loglang or engelang.
>
> Again, a good point.  Indeed, a language like the one Ray would like
> to see is more syntactically rigorous than any human natlang, and
> why should the natlangs of any alien sapient species be more
> syntactically rigorous than ours?
We simply don't know. Also beings who think differently from ourselves may
find more rigor in human languages than we appreciate. And how we build
irregularity into a non-human model is not a trivial matter.

>  There is no solid reason for that.
> Of course, a "Fithian Lojban" might actually be like what Ray is
> dreaming of.
No - I'm not dreaming of anything - just outlining how I would go about a
stack-based language. I do not think the result would be much like Lojban,
  which seems to me "eccentrically human" - which may be why it hasn't
produced any tangible results AFAIK in testing the Sapir-Whorf hypothesis.
  Um - maybe if we tried Stax.......

Oh dear - I'm beginning to get interested in a stack-based language. But I
had better resist the temptation and concentrate on getting
briefscript/Bax out of my system as soon as possible   ;)

Ray
===============================================
http://home.freeuk.com/ray.brown
ray.brown@freeuk.com
===============================================
Anything is possible in the fabulous Celtic twilight,
which is not so much a twilight of the gods
as of the reason."      [JRRT, "English and Welsh" ]

Reply