[NGL] Wierzbica Semantic Primes in NGL
Jack Durst
Date: Sat, 27 Nov 1999 23:49:21 -0800 (PST)
From: Gerald Koenig <j...@netcom.com>
Subject: [NGL] Wierzbica Semantic Primes in NGL
From: Gerald Koenig <j...@netcom.com>
Here are some extracts from my current musings on NGL:
>From Anna Wierzbicka:
P. 131
12. Time:WHEN, AFTER, BEFORE, A LONG TIME, A SHORT TIME, NOW
Clausal Adjuncts.
"The fundamental temporal concepts encoded in the languages of the world
appear to be the same. They include WHEN, AFTER, and BEFORE"
Here are some time sentences in her Universal Grammar that use WHEN, or
its equivalent, AT A TIME, and more of the above.]
Ex 1: At that time, you did something.
Si ace letu, vu pa dinfe.
Simultaneous to that time interval, you past thing-do. OR:
Me ace letu, vu am dinfe.
Mu a member of that time-vector, you were thing-doing.
During that interval, you did something.
Ex 2: When I did these things, I felt something bad.
Si mip ecos dinsfe, mip *luil din tom.
Does <din> mean "something"? If not I would use ax, X.
*ax::- X, some thing or some number, according to context.
naes=feeling, n.
Surprisingly, we lack a verb for "feel", one of Wierzbicka's
55 "old" semantic primitives along with "want", "know" and "think".
I propose <luil>.
*luil::- to feel, v.
Ex 3: When will it happen?
Kuol atik gol fu xolcit.
What is the value of the time-atom for it will happen.
Ans: 10^999. [a guess]. I'm watching "An Enemy of the State" on
television in the background and this form seems appropriate :).
Or an equivalent can be given: 9:10 pm.
OR:
Kuol letu gol fu xolcit?
Which is the value of the time vector for it will happen?
Ans: 9:10pm-9:20pm.
Kuol ayoh cesi gol fu xolcit?
Which is the value of a marker event for "it will happen"?
kuol ayoh cesi= kuo
*kuo::-which is the value of a marker event proposition for...;
or just plain "when?".
*Kuo gol fu xolcit?
When it will happen?
Vo ku maredim diem 3 cikes.
Ans: Before the cock crows thrice.
Again surprisingly we do not have a word equivalent to "when?", another
Wierzbicka semantic primitive universal concept. I propose
*kuo::-when
Since it's a primitive, we all know what it means and we can't define it
further, although I thought I did above.
ON BEFORE and AFTER:
Page 57:
"What is particularly innteresting about these findings is the apparant
redundancy of the exponents of temporary succession; for why should all
languages have words for both BEFORE and AFTER, rather than for just
one of these concepts? Aren't they reducible to one another?
>From a logical point of view, indeed, they might be reducible to one
another. But the fact that natural languages have lexical resources for
expressing both of them suggests that from the point of view of human
conceptualization of reality, "Y happened after X" means something
different from, and irreducible to, "X happened after Y". Clearly, what
matters is a different perspective on the events in question, a
different point of view, and, as pointed out by Slobin (1985a: 1181),
"the ability to view scenes from different perspectives" is a salient
feature of human cognition, clearly reflected in all languages. "
Ok, with that as background I am going to present my understanding of
this with vector tense.
Since the mu [*] has moved rightward it first made x happen and then
y. Here Y is the first argument, the nominative, we visualize it
center stage in the "figure" position. AFTER puts Y front and center.
1. Y happened AFTER X:
_ _ _ _ _ _ _ _ _ _
| |
ground | | ground
x happened | y happened |
BB---@========>----------------@===========>---|----*-------------->
| |
ground | figure | ground
| _ _ _ _ _ _ _ _ _|
O
\|/
|
/ \
AFTER puts Y front and center.
2. X happened BEFORE Y:
_ _ _ _ _ _ _ _ _ _
| |
| |
ground | | ground
| x happened | y happened
BB -------------------------|-----@========>---|-----------@===========>-*-
| |
ground | figure | ground
|_ _ _ _ _ _ _ _ _|
O
\|/
|
/ \
BEFORE puts X front and center.
So that although these two statements are logically equivalent, they
look different in our imaginations due to what I call
"The linearization of perception by language"; Somehow the first
predication gets the center of attention, even though the BEFORE and
AFTER are symmetrical relations.
Well, there is another interesting part to this post, dealing with the
now popular expression, "let's put that behind us" as a vector metaphor
that ties into the Wierzbicka primitives, it's too much for tonite.
I hope this message finds Jack, Stephen, Julian, Carlos, Mia, et al
well.
----------------------------------------------------------------------
Jerry | Without careful communication
Gerald Lea Koenig | jlkatnetcomdotcom There is boundless demonization.
----------------------------------------------------------------------
--------------------------- ONElist Sponsor ----------------------------
GRAB THE GATOR! FREE SOFTWARE DOES ALL THE TYPING FOR YOU!
Tired of filling out forms and remembering passwords? Gator fills in
forms and passwords with just one click! Comes with $50 in free coupons!
<a href=" http://clickme.onelist.com/ad/gator4 ">Click Here</a>
------------------------------------------------------------------------
Jack Durst
Date: Tue, 30 Nov 1999 00:41:01 -0330
From: Stephen DeGrace <c72...@morgan.ucs.mun.ca>
Subject: Re: [NGL] Wierzbica Semantic Primes in NGL
From: Stephen DeGrace <c72...@morgan.ucs.mun.ca>
On Sat, 27 Nov 1999 23:49:21 -0800 (PST), Gerald Koenig
<j...@netcom.com> wrote:
>From: Gerald Koenig <j...@netcom.com>
>
[...]
>Si ace letu, vu pa dinfe.
>Simultaneous to that time interval, you past thing-do. OR:
>
>Me ace letu, vu am dinfe.
>Mu a member of that time-vector, you were thing-doing.
>During that interval, you did something.
>
>
>Ex 2: When I did these things, I felt something bad.
> Si mip ecos dinsfe, mip *luil din tom.
^^^^^^
Needs to be {dinesfe}, the -e- is needed to obey a syllable pattern
rule forbidding clusters of three consonants inside a word.
> Does <din> mean "something"? If not I would use ax, X.
> *ax::- X, some thing or some number, according to context.
According to the semi-module that was hammered out last year,
"something" is handled with {yerdin}, "something" and "anything" being
handled by the same word with {òl} or {q} to be used if necessary to
help differentiate meaning.
>naes=feeling, n.
>Surprisingly, we lack a verb for "feel", one of Wierzbicka's
>55 "old" semantic primitives along with "want", "know" and "think".
>I propose <luil>.
>
>*luil::- to feel, v.
_Currently_, I just verb {naes} it give {naesit}, "feel". My
_personal_ opinion is that this has been sufficient.
>Ex 3: When will it happen?
> Kuol atik gol fu xolcit.
> What is the value of the time-atom for it will happen.
> Ans: 10^999. [a guess]. I'm watching "An Enemy of the State" on
> television in the background and this form seems appropriate :).
> Or an equivalent can be given: 9:10 pm.
So while the meaning of "time atom" in tis literal interpretation is
very specific, for practical purposes this specificity can be relaxed,
I take it...
Something I hate to confess my ignorance to, but I don't feel I quite
understand where {kuol} comes from.
[...]
> *kuo::-which is the value of a marker event proposition for...;
> or just plain "when?".
Right now, this has been handled using the question-word-making
particle {-ne}, so "when" in the question word sense has been {cikne}
and "when" in the relative pronoun sense has been {cike} ({cik} +
{ke}).
[...]
>Again surprisingly we do not have a word equivalent to "when?", another
>Wierzbicka semantic primitive universal concept. I propose
>
>*kuo::-when
>
>Since it's a primitive, we all know what it means and we can't define it
>further, although I thought I did above.
Well, we can define it with great exactitude I guess, but ultimately
we need to also be able to define our terms roughly like this,
otherwise people will be intimidated by the level of detail and maybe
experience trouble in actually understanding and using the terms as
they are envisioned to be used. A range of levels of detail is needed
in describing words...
[...]
>BEFORE puts X front and center.
>So that although these two statements are logically equivalent, they
>look different in our imaginations due to what I call
>"The linearization of perception by language"; Somehow the first
>predication gets the center of attention, even though the BEFORE and
>AFTER are symmetrical relations.
>
>Well, there is another interesting part to this post, dealing with the
>now popular expression, "let's put that behind us" as a vector metaphor
>that ties into the Wierzbicka primitives, it's too much for tonite.
>I hope this message finds Jack, Stephen, Julian, Carlos, Mia, et al
>well.
It finds me fairly well, although recovering from a slight touch of
flu :-). The exposition on before and after was very interesting, an
important illustration of how the human entity of language often lives
by a "logic" all its own. That "logic" may not always make perfect
sense from the point of view of actual logic, but yet as language
designers we risk our creations having impeded functionality if we
fail to pay attention to it. Focus of attention is something that is
very important to people in communication. PVS gives a bit of a nod to
this fact in that compound verbs inherently make a distinction between
the time we're talking about, i.e., focused upon, versus the time in
which the verb takes place... VTT, with its fine logic and potential
for great precision has the potential to not only express these
concepts very well to to maybe teach us something about the way we
think...
Naesverig,
Stephen
--------------------------- ONElist Sponsor ----------------------------
*** Get Your FREE Post it Notes from iPrint.com! ***
Offer for new iPrint.com customers only. Shipping as low as $3.20.
Click on <a href=" http://clickme.onelist.com/ad/iPrint34 ">FREE OFFER</a>
------------------------------------------------------------------------
Jack Durst
Date: Tue, 30 Nov 1999 01:36:53 -0800 (PST)
From: Gerald Koenig <j...@netcom.com>
From: Gerald Koenig <j...@netcom.com>
kuol::- What is the value of VARIABLE_TYPE for [abstract] OBJECT..
maol::- The value of VARIABLE_TYPE for [abstract] OBJECT is: ___
variable type::- VEC, VIS, VIT, ATIK, NOM, CARDINAL, ORDINAL,
PROPOSITIONAL (P), UNKNOWN (X).
Don't be embarassed about questioning <kuol>. Carlos had a problem with
this concept and frankly I sometimes find it very elusive myself,
although I have that strong feeling that I need it and it is correct.
As for its sources, I _think I got the idea from a combination of the
Spanish word <cual> which is an interrogative word asking "which one"
of a collection or array, such as ?Cual is the date of today?, and from
the Unix environment variable statements. For example if I type the
command <env> at the Unix command prompt I get this:
TERM=vt100
HOME=/u/u13/jlk
SHELL=/usr/bin/tcsh
USER=jlk
PATH=/usr/ucb:/usr/bin:/usr/5bin:/usr/local/bin:/u/u13/jlk/bin:.
LOGNAME=jlk
HOSTTYPE=sun4
VENDOR=sun
OSTYPE=sunos4
MACHTYPE=sparc
SHLVL=2
PWD=/u/u13/jlk
HOST=netcom10.netcom.com
LESSCHARSET=latin1
NNTPSERVER=nntp.ix.netcom.com
If someone else were to type this same command, <env> on their machine
they would get a different set of assignments for the capitalized
variables. For example, if you could do it it might give
USER=sad
for Steven Arthur DeGrace.
The USER variable has a range of names of users on the system.
The HOME variable has a range of home directory paths on the system,
and so on.
So the value of the USER variable for you might be <sad>.
Another way of saying this could be:
<Maol> USER sad. (using the above nilenga definition)
The value of the variable type USER is currently sad.
Which is a way of saying that in the current environment
(your system) sad is the user.
Or, the value of USER is your initials.
Now, <kuol> is the interrogative form that asks, what is the actual
assignment of USER.
It's like you pick up on a conversation, the interlocutors are speaking
of "he". You want to know who "he" refers to. Then you could ask,
<kuol> "he" ? (What is the current value assignment for "he"?)
Ans: John Doe, or whoever. We know that the range of "he" is male
persons.
Just like you could ask,
<kuol> LOGNAME? What is the current value for LOGNAME?
Ans: jlk, taken from the above environment or situation. The range of
LOGNAME is the user names on the system.
<kuol> asks for the current assignment of a variable type. So there is
always an implicit or explicit range set from which the answer
can be selected. But it doesn't always have to be a number.
kuol lesi? What is the value of the short space vector?
How close is it? (in context)
kuol leto? What is the value of the long space vector?
How far away is it? (context)
kuol letu? What is the value of the time vector?
How long does it take? (requires context).
kuol letu pai? What is the value of the past time vector?
How long did it take?
Here I consider <pa> an adverb, <pa-i> makes the adj.
modifying <letu>, here noun, but sometimes a verb.
kuol nom? Which name?
kuol nom ivu? Which one is your name?
kuol P? Which statement?
kuol atik? Which time-atom? What is the value of ATIK? What time is it?
And so on.
Well this got a lot longer than I imagined, so I will stop here before
I become really obtuse. I'll respond to the rest of your posts later.
I'd be very happy to see the writekit linked to the central
repository. I want to add a paragraph from your last post on the
design of PVS to the kit also. I'd like to see the kit well enough
balanced between systems so that any translation/composition can be
done, even though it's primarily about vector tense. If you could point
me to a summary of PVS I'll include it. Ultimately it is up to the
people in our democratic language to decide how they want to use the
systems provided.
Jerry