Difference between 'I' and 'me'

12 views
Skip to first unread message

Luka Crnkovic

unread,
Jan 29, 1997, 3:00:00 AM1/29/97
to
Hi,
 
I am currently writing a program that processes nat-lang. and I have
a problem. At one stage of my program I would like to reverse the the
pronouns so that you becomes 'I', 'your' becomes 'my' and so on.
The actual problem is in the conversion of the word 'you' 
- it can be translated as both 'me' and 'I'.
 
ex:
Original:Do you know anything?
Reversed: Do I know anything?
 
ex2:
 
Original: Do I know you?
Reversed: Do you know me?
 
---
Does anyone have a solution to this problem?
-- 
------------------------------------ 
mailto:deva...@geocities.com
http://www.geocities.com/CapeCanaveral/1630/
------------------------------------
 

Norbert E. Fuchs

unread,
Jan 30, 1997, 3:00:00 AM1/30/97
to

In article <32EF91...@geocities.com>, deva...@geocities.com wrote:
>
> I am currently writing a program that processes nat-lang. and I have
> a problem. At one stage of my program I would like to reverse the the
> pronouns so that you becomes 'I', 'your' becomes 'my' and so on.
> The actual problem is in the conversion of the word 'you'
> - it can be translated as both 'me' and 'I'.
>
> ex:
> Original:Do you know anything?
> Reversed: Do I know anything?
>
> ex2:
>
> Original: Do I know you?
> Reversed: Do you know me?
>
> ---
> Does anyone have a solution to this problem?
> --

Check the rule-based Eliza system of Weizenbaum which does what you want.

------------------------------------------------------------------------
Norbert E. Fuchs Telephone +41-1-257 43 13
Department of Computer Science Fax +41-1-363 00 35
University of Zurich Email fu...@ifi.unizh.ch
CH-8057 Zurich, Switzerland WWW www.ifi.unizh.ch/staff/fuchs.html
------------------------------------------------------------------------

Stewart Gilles

unread,
Jan 30, 1997, 3:00:00 AM1/30/97
to

Luka Crnkovic <deva...@geocities.com> writes:


>------------8C6498519760
>Content-Transfer-Encoding: 7bit
>Content-Type: text/plain; charset=us-ascii

>Hi,

>I am currently writing a program that processes nat-lang. and I have
>a problem. At one stage of my program I would like to reverse the the
>pronouns so that you becomes 'I', 'your' becomes 'my' and so on.
>The actual problem is in the conversion of the word 'you'
>- it can be translated as both 'me' and 'I'.

>ex:
>Original:Do you know anything?
>Reversed: Do I know anything?

>ex2:

>Original: Do I know you?
>Reversed: Do you know me?

>---
>Does anyone have a solution to this problem?
>--

When "you" is subject ==> I
When "you" is object ==> me

Gilles
--
Gilles Stewart VROUM __o stew...@ere.umontreal.ca
Services informatiques VROUM _`\<,_ (514)343-6111 poste 5248
Pavillon administratif (*)/ (*)
Universite de Montreal ~~~~~~~~~~~~~~~~~~~~~~~~

Luka Crnkovic

unread,
Jan 31, 1997, 3:00:00 AM1/31/97
to
I understand, but how do I differ the subject and the object when parsing
such a text?


>         When "you" is subject ==> I
>         When "you" is object  ==> me

> Gilles
> --
> Gilles Stewart             VROUM    __o      stew...@ere.umontreal.ca
> Services informatiques    VROUM   _`\<,_     (514)343-6111 poste 5248
> Pavillon administratif           (*)/ (*)
> Universite de Montreal   ~~~~~~~~~~~~~~~~~~~~~~~~

Luka Crnkovic

unread,
Jan 31, 1997, 3:00:00 AM1/31/97
to
I am afraid that Eliza cannot handle it. The winner of the Loebner prize 1996 HeX cannot handle it. Actually I havn't found any program that can.


> Check the rule-based Eliza system of Weizenbaum which does what you want.

> ------------------------------------------------------------------------
> Norbert E. Fuchs                 Telephone               +41-1-257 43 13
> Department of Computer Science   Fax                     +41-1-363 00 35
> University of Zurich             Email                fu...@ifi.unizh.ch
> CH-8057 Zurich, Switzerland      WWW   www.ifi.unizh.ch/staff/fuchs.html
> ------------------------------------------------------------------------

John M. Lawler

unread,
Jan 31, 1997, 3:00:00 AM1/31/97
to

Luka Crnkovic <deva...@geocities.com> writes:
> Gilles Stewart <stew...@ere.umontreal.ca> writes:

>> When "you" is subject ==> I
>> When "you" is object ==> me

> I understand, but how do I differ the subject and the object when
> parsing such a text?

Why not just use the assigned theta roles?
You could have your parser look them up in the latest LI.

-John Lawler http://www.lsa.umich.edu/ling/jlawler/ U Michigan Linguistics
---------------------------------------------------------------------------
"Language is the most massive and inclusive art we know, a - Edward Sapir
mountainous and anonymous work of unconscious generations." Language (1921)

Stewart Gilles

unread,
Jan 31, 1997, 3:00:00 AM1/31/97
to

Luka Crnkovic <deva...@geocities.com> writes:


>------------527024D77DB31


>Content-Transfer-Encoding: 7bit
>Content-Type: text/plain; charset=us-ascii

>I understand, but how do I differ the subject and the object when


>parsing
>such a text?
>>
>> When "you" is subject ==> I
>> When "you" is object ==> me
>>

>> Gilles

You said it : you have to *parse* the text, not only *tokenize*
it. In order to parse a text you need a *grammar* describing the
universe of texts you must account for. Then you need a parser that,
given your grammar and a text, will give you the *structure* of the
text in which you can find where the subjects and objects are.

Gilles

>> --
>> Gilles Stewart VROUM __o stew...@ere.umontreal.ca
>> Services informatiques VROUM _`\<,_ (514)343-6111 poste 5248
>> Pavillon administratif (*)/ (*)
>> Universite de Montreal ~~~~~~~~~~~~~~~~~~~~~~~~

>--


>------------------------------------
>mailto:deva...@geocities.com
>http://www.geocities.com/CapeCanaveral/1630/
>------------------------------------


>------------527024D77DB31
>Content-Transfer-Encoding: 7bit
>Content-Type: text/html; charset=us-ascii

><HTML><BODY>

><DT>I understand, but how do I differ the subject and the object when parsing</DT>

><DT>such a text?</DT>

><DT><BR>
>&gt;&nbsp;<BR>
>&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; When &quot;you&quot;
>is subject ==&gt; I<BR>
>&gt;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; When &quot;you&quot;
>is object&nbsp; ==&gt; me<BR>
>&gt;&nbsp;<BR>
>&gt; Gilles<BR>
>&gt; --<BR>
>&gt; Gilles Stewart&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
>VROUM&nbsp;&nbsp;&nbsp; __o&nbsp;&nbsp;&nbsp;&nbsp;&nbsp; stew...@ere.umontreal.ca<BR>
>&gt; Services informatiques&nbsp;&nbsp;&nbsp; VROUM&nbsp;&nbsp; _`\&lt;,_&nbsp;&nbsp;&nbsp;&nbsp;
>(514)343-6111 poste 5248<BR>
>&gt; Pavillon administratif&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;&nbsp;
>(*)/ (*)<BR>
>&gt; Universite de Montreal&nbsp;&nbsp; ~~~~~~~~~~~~~~~~~~~~~~~~<BR>
><BR>
>--&nbsp;<BR>
>------------------------------------&nbsp;<BR>
>mailto:deva...@geocities.com<BR>
>http://www.geocities.com/CapeCanaveral/1630/<BR>
>------------------------------------<BR>
>&nbsp;</DT>

></BODY>
></HTML>
>------------527024D77DB31--

SCN User

unread,
Feb 1, 1997, 3:00:00 AM2/1/97
to

On WED.29JAN.1997, Luka Crnkovic (deva...@geocities.com> queried:

> I am currently writing a program that processes nat-lang. and I
> have a problem. At one stage of my program I would like to reverse

> the pronouns so that you becomes 'I', 'your' becomes 'my' and so on.


> The actual problem is in the conversion of the word 'you'
> - it can be translated as both 'me' and 'I'.

In October 1994 this same problem arose in the development of the
public-domain (Amiga-only) Mind.rexx program of Project Mentifex.

The solution in Mind.rexx was to set up three separate arrays for
the representation of concepts and lexical items, as shown below:

/^^^^^^^^^^^\ Syntax Strings Together a Thought /^^^^^^^^^^^\
/visual memory\ semantic ________ / auditory \
| /--------|-------\ memory / syntax \ |episodic memory|
| | recog-|nition | \________/<--|-------------\ |
| ___|___ | | flush-vector| | _______ | |
| /image \ | __|___ ___V___ | /stored \ | |
| / percept \ | /deep \<-----/lexical\<---|--/ phonemes\| |
| \ engrams /<--|-->/concepts\--->/concepts \---|->\ of words/ |
| \_______/ | \________/ \_________/ | \_______/ |

The array of "deep concepts" holds primitive ideas: self; other.
The intermediate "lexical" array controls onset-tags of phonemes.
The "auditory" array stores words, recallable by associative tag.

Activation of the deep "self" concept activates the lexical fiber
(neuronal fibergang) of "I" which in turn activates the word "I".

When Andru the Mentifex cyborg hears the word "I" externally, his
intermediate language-area interprets "I" as referring to "other"
(a member of the non-self class of "other") in the deep mindcore,
and so the appropriate links are tagged during the comprehension.

Arthur T. Murray

unread,
Feb 3, 1997, 3:00:00 AM2/3/97
to

In a previous article, ment...@scn.org (SCN User) says:

>
>On WED.29JAN.1997, Luka Crnkovic (deva...@geocities.com> queried:
>
>> I am currently writing a program that processes nat-lang. and I
>> have a problem. At one stage of my program I would like to reverse
>> the pronouns so that you becomes 'I', 'your' becomes 'my' and so on.
>> The actual problem is in the conversion of the word 'you'
>> - it can be translated as both 'me' and 'I'.
>
>In October 1994 this same problem arose in the development of the
>public-domain (Amiga-only) Mind.rexx program of Project Mentifex.


Am 9. Oktober 1994, Sonntag in der Corliss-Strasse. At Vaierre I
have been trying to figure out the triple-depth mind. When I fi-
nally thought that I understood it, I was able to rise and come
hither.
Although I have not settled finally on particular names, let's
say that the inmost concepts "deep.f" interact with the intermedi-
ate concepts "fiber.n", which in turn activate the "ph.t" audito-
ry engrams.
Let's imagine that a nascent mind forms two primordial con-
cepts, "self" and "other."
Now let's discuss an advanced mind that hears the sentence,
"You eat fish."
As I lay thinking at Vaierre today, I examined the difference
between uttering the word "you" and hearing the word "you." Af-
ter much thought, interspersed with falling asleep several times
and even dreaming, I came to the conclusion that the difference
will consist chiefly in the activation of reference-flags within
the flag-panels of the fiber.n concepts. There will be perhaps
an "endo" (Greek for "interior") flag within the fiber.n flag-
panel to direct the activation of deep concepts when the interme-
diate fiber.n concepts are activated.
Some of my reasoning follows.
The first deep concept, that of "self," probably arises when
the mental organism first feels anything with the senses. An
"ego" senses something, and therefore "ego" becomes a concept.
That is to say, an initial "ego" concept-fiber becomes active.
Immediately a second concept becomes active, a vague, undif-
ferentiated notion of the non-self "other" that the mind senses.
This massively unknown "other" will gradually be differentia-
ted into multifarious ontology, in a process whereby conceptual
aggregates join and divide and move and scatter, not unlike pro-
tozoa [Maybe "proto.f" is the name which I am seeking.] being ob-
served through a microscope.
I further reasoned that the word "you" is first learned as re-
ferring to the concept of self, and only much later does the ego
turn the word "you" around to refer in speech to some "other."
Whereas for some recent months I was worried that the ambiva-
lent uses of "you" could not be kept apart, now today I can see
that the endo-flag will be like a logic-valve to make sure that
"self" you and "other" you are not mixed up.
If the machine hears the word "you," then the endo-flag of the
fiber.n for "you" will activate the "self" proto-concept.
But if the ego uses the word "you" in speech, then a facet of
some "other" concept will be activating the fiber.n "you" in or-
der to activate the auditory engram "you" during the "THINKING"
modality. Even when the thought undergoes re-entry, there should
be an associative mechanism of suppressing the endo-flag activa-
tion of the "self" concept.
Likewise the word "I" is first heard in a way which activates
the "other" concept. [Isn't it lucky that language is a social
phenomenon?] So hearing the word "I" activates the proto-concept
of "other," but thinking about "other" activates the word-concept
of "you."
These inmost proto-concepts are like the heaving crankshafts
and lurching gears of massively heavy machinery, and similar also
to the flitting patterns of sunlight amid the leaves of a tree.

--

Joseph S. Wisniewski

unread,
Feb 3, 1997, 3:00:00 AM2/3/97
to Arthur T. Murray, ment...@scn.org, uj...@freenet.victoria.bc.ca

Arthur,

Luka posted his question only to comp.ai.nat-lang. Why did you
find it necessary to crosspost your reply to comp.ai, comp.speech,
sci.cognitive, sci.lang, and soc.history.science? Your post is off
topic in most of those groups.

Also, why did you use your account at the Seattle Community Network
(ment...@scn.org) for the reply on comp.ai.nat-lang, and your Los
Angeles Freenet account (ba...@lafn.org) to add the crossposts to all
the other groups?

This is my theory, please tell me if I'm wrong. You have your cool,
official looking "ment...@scn.org" account on the Seattle freenet. The
other freenets only give you accounts with code numbers, such as ba672
or uj797, they won't let you call yourself "mentifex," so the Seattle
account is the one you protect: it's the one you use for replies in the
same group as the posts you reply to. Los Angeles and Vancouver are your
throwaway, expendible accounts, so they are the ones you use for the
spam crossposts. Have I got it?

Arthur T. Murray wrote:
>
> On WED.29JAN.1997, Luka Crnkovic (deva...@geocities.com> queried:
>
> > I am currently writing a program that processes nat-lang. and I
> > have a problem. At one stage of my program I would like to reverse
> > the pronouns so that you becomes 'I', 'your' becomes 'my' and so on.
> > The actual problem is in the conversion of the word 'you'
> > - it can be translated as both 'me' and 'I'.
>
> In October 1994 this same problem arose in the development of the
> public-domain (Amiga-only) Mind.rexx program of Project Mentifex.

[yet another Mentifex diagram deleted]

--
Joseph S. Wisniewski | Views expressed are my own, and don't reflect
Ford Motor Company | those of the Ford Motor Co. or affiliates.
Project Sapphire | LeMans, Daytona, Bonneville, and Sebring are
jwis...@ford.com | just races, won by people driving Ford cars!

Sean Walton - D26

unread,
Feb 12, 1997, 3:00:00 AM2/12/97
to

This is the most useless waste of usenet bandwidth. While I actually
understand what the poster is trying to say, I think that he could
possibly put down the thesaurus, drop the metaphor, metaphysical and
allegory and just answer the simple question:

"When is 'you' a subject and when is it an object?"

Syntactically it is relatively simple. Use the following rules:

1) Identify the nouns and other sentence parts.
2) By placement in the sentence (order), identify whether each
noun is a subject or an object (direct/indirect).
3) All subject-"you's" will become "I's"; all object "you's"
will be "me's".

As you work on more complex sentences, you will see that #2 is not so
fast and clear.

Now back to the responder:
Listen, I know you want to prove yourself on the net and show that
you are older than you really are. People here are trying to solve
problems not listen to you yabber. If you do not know the answer,
do not talk--listen!

-Sean Walton, KB7rfa

In article <1997Feb3.0...@lafn.org>, ba...@lafn.org (Arthur T. Murray) writes:
>
> In a previous article, ment...@scn.org (SCN User) says:
>
> >

> >On WED.29JAN.1997, Luka Crnkovic (deva...@geocities.com> queried:
> >
> >> I am currently writing a program that processes nat-lang. and I
> >> have a problem. At one stage of my program I would like to reverse
> >> the pronouns so that you becomes 'I', 'your' becomes 'my' and so on.
> >> The actual problem is in the conversion of the word 'you'
> >> - it can be translated as both 'me' and 'I'.
> >
> >In October 1994 this same problem arose in the development of the
> >public-domain (Amiga-only) Mind.rexx program of Project Mentifex.
>
>

Reply all
Reply to author
Forward
0 new messages