Re: OT: PL/I was Re: Please welcome . . .

On Thu, 18 Dec 2003 09:47:56 -0800, Gary Shannon <fiziwig@...> wrote:

>An interesting idea.  I did try, about 3 or 4 years
>ago, to design an RPN (Reverse Polish Notation)
>conlang inspired by the Forth programming language.
>
>Anyone familiar with the early HP pocket calculators
>will know what I'm talking about.  The arguments are
>placed on a stack and the operators take the arguments
>off the stack, placing the reuslt back onto the stack.
> For example, to add 3 and 9 one uses something along
>the lines of "3 9 plus".
>
>The idea for an RPN conlang was to stack all the
>arguments and then use the verb to gather them
>together and create some action, the result of which
>is placed back on the stack: "I book red that_is
>have".
>
>The sequence "book red" places those two arguments on
>the stack and then the operator "that_is" gathers them
>toegther and binds "redness" to "book" and puts the
>concept of "red+book" back on the stack to be later
>consumed by the operator "have".
>
>--gary
Well, my binary loglang uses (forward) Polish Notation.

Furthermore, I actually started to sketch a stack-based conlang about 17
days ago, but I've abandoned it already.  The only things that could go on
the stack were predicates, very much like Lojbanic bridi.  For 'I have a
green mouse' (which is the closest I can get to 'I have a red book' with my
 minimal vocab), we could say this.

?\íri kíle  ja dZóvu   na  ja:
mouse green    possess 1sg

Acute marks high tone, unmarked is low, and word divisions are not
necessary.  Each of the disyllabic morphemes with HL tone pattern is a
predicate.  mouse(x) asserts that x is a mouse, green(x) that x is green,
possess(x, y) that x possesses y, and so forth.

The <ja> which I've left unglossed takes the first argument in the predicate
just below the top of stack and equates it with an argument in the predicate
at top of stack, after which it removes the former predicate from the stack.
 (There are other such morphemes which operate on different stack levels and
different arguments.)  So after the second morpheme the stack is

green() [top]
mouse()

with none of the arguments specified.  <ja> equates the first place of
'green' with the first place of 'mouse', so that the stack contains

green(x) for some x such that mouse(x)

and then removes the 'mouse' predicate.  After we put 'possess' on the
stack, <na> sets its first argument to 'I'.  Then <ja> sets its second
argument so that it becomes the following.

possess(I, x) for some x such that (green(x), where x is also such that
mouse(x)).

Finally, the lengthening of the vowel (which is a morpheme on its own)
asserts what is currently on top of the stack.  Without it the sentence
would have placed a predicate on the stack without asserting anything.

I hope that wasn't too unclear.

Alex

Reply