Connections between units of meaning

[Peter Olcott]

I am trying to derive a complete list of every type of syntactic or
semantic connection between units of meaning (defined below). This is my
current list:

1) Modification (Adverb, Adjective, Quantifier)
2) Anaphora (Pronouns)
3) Tense (location in time)
4) Aspect (fabric of time)
5) Mood (modality, evidentiality)
6) Thematic Roles (Agent, Patient, Instrument, etc)
7) Agreement (Person, Number, Gender, Case)
8) Subcatorization (Verbs select their syntactic arguments)
9) Selection Restriction (Predicates select their semantic arguments)

Some of the items on the above list may be imperfectly specified, I am
just learning these terms for the first time. My primary goal (at this
point) is to make the basic list of categories of types of connections
as complete as possible.

Are there any categories of types of connections between units of
meaning that I have left out of the above list?

AtomicUnitOfMeaning: All of the meaning associated with as a single
sense meaning of a single morpheme.
http://en.wikipedia.org/wiki/Morpheme

This atomic unit of meaning is larger than the meaning of a single Seme.
http://en.wikipedia.org/wiki/Seme_(semantics)

A UnitOfMeaning is comprised (AtomicUnitOfMeaning or UnitsOfMeaning)
possibly combined with (AtomicUnitOfMeaning or UnitsOfMeaning).

UnitOfMeaning:
| AtomicUnitOfMeaning
| AtomicUnitOfMeaning * AtomicUnitOfMeaning
| UnitOfMeaning * AtomicUnitOfMeaning
| AtomicUnitOfMeaning * UnitOfMeaning
| UnitOfMeaning * UnitOfMeaning

[Peter T. Daniels]

On Aug 19, 1:30 pm, Peter Olcott <OCR4Screen> wrote:
> I am trying to derive a complete list of every type of syntactic or
> semantic connection between units of meaning (defined below). This is my
> current list:
>
>     1) Modification          (Adverb, Adjective, Quantifier)
>     2) Anaphora              (Pronouns)
>     3) Tense                 (location in time)
>     4) Aspect                (fabric of time)
>     5) Mood                  (modality, evidentiality)
>     6) Thematic Roles        (Agent, Patient, Instrument, etc)
>     7) Agreement             (Person, Number, Gender, Case)
>     8) Subcatorization       (Verbs select their syntactic arguments)
>     9) Selection Restriction (Predicates select their semantic arguments)
>
> Some of the items on the above list may be imperfectly specified, I am
> just learning these terms for the first time. My primary goal (at this
> point) is to make the basic list of categories of types of connections
> as complete as possible.

Why would "Agreement" have anything at all to do with semantics?

> Are there any categories of types of connections between units of
> meaning that I have left out of the above list?
>
> AtomicUnitOfMeaning: All of the meaning associated with as a single

> sense meaning of a single morpheme.http://en.wikipedia.org/wiki/Morpheme

[Glenn Knickerbocker]

On Sun, 19 Aug 2012 10:33:56 -0700 (PDT), Peter T. Daniels wrote:
>Why would "Agreement" have anything at all to do with semantics?

Why would you expect "every type of syntactic or semantic connection" to
have something to do with semantics?

ŹR - we don't have branes made of clockworks, chocolate maidens
http://users.bestweb.net/~notr/arkville.html --the Ur-beatle

[Peter Olcott]

On 8/19/2012 12:33 PM, Peter T. Daniels wrote:
> On Aug 19, 1:30 pm, Peter Olcott <OCR4Screen> wrote:
>> I am trying to derive a complete list of every type of syntactic or
>> semantic connection between units of meaning (defined below). This is my
>> current list:
>>
>> 1) Modification (Adverb, Adjective, Quantifier)
>> 2) Anaphora (Pronouns)
>> 3) Tense (location in time)
>> 4) Aspect (fabric of time)
>> 5) Mood (modality, evidentiality)
>> 6) Thematic Roles (Agent, Patient, Instrument, etc)
>> 7) Agreement (Person, Number, Gender, Case)
>> 8) Subcatorization (Verbs select their syntactic arguments)
>> 9) Selection Restriction (Predicates select their semantic arguments)
>>
>> Some of the items on the above list may be imperfectly specified, I am
>> just learning these terms for the first time. My primary goal (at this
>> point) is to make the basic list of categories of types of connections
>> as complete as possible.
> Why would "Agreement" have anything at all to do with semantics?

This whole thread is about connections between units of meaning. Syntax
is an aspect of the connections between units of meaning, and
"Agreement" is an aspect of this syntactic connection.

[Peter Olcott]

On 8/19/2012 3:26 PM, Glenn Knickerbocker wrote:
> On Sun, 19 Aug 2012 10:33:56 -0700 (PDT), Peter T. Daniels wrote:
>> Why would "Agreement" have anything at all to do with semantics?
> Why would you expect "every type of syntactic or semantic connection" to
> have something to do with semantics?

The purpose of syntax is to connect units of meaning together.

[Peter T. Daniels]

On Aug 19, 4:30 pm, Peter Olcott <OCR4Screen> wrote:
> On 8/19/2012 12:33 PM, Peter T. Daniels wrote:
>
>
>
> > On Aug 19, 1:30 pm, Peter Olcott <OCR4Screen> wrote:
> >> I am trying to derive a complete list of every type of syntactic or
> >> semantic connection between units of meaning (defined below). This is my
> >> current list:
>
> >>      1) Modification          (Adverb, Adjective, Quantifier)
> >>      2) Anaphora              (Pronouns)
> >>      3) Tense                 (location in time)
> >>      4) Aspect                (fabric of time)
> >>      5) Mood                  (modality, evidentiality)
> >>      6) Thematic Roles        (Agent, Patient, Instrument, etc)
> >>      7) Agreement             (Person, Number, Gender, Case)
> >>      8) Subcatorization       (Verbs select their syntactic arguments)
> >>      9) Selection Restriction (Predicates select their semantic arguments)
>
> >> Some of the items on the above list may be imperfectly specified, I am
> >> just learning these terms for the first time. My primary goal (at this
> >> point) is to make the basic list of categories of types of connections
> >> as complete as possible.
> > Why would "Agreement" have anything at all to do with semantics?
>
> This whole thread is about connections between units of meaning. Syntax
> is an aspect of the connections between units of meaning, and
> "Agreement" is an aspect of this syntactic connection.

Why would you list "Agreement" (a property of some languages) in a
list of putative universals? Likewise for gender, number, case)? Why
would you list "Case" as a thing separate from "Thematic Roles"?
"Theta theory" was an attempt to accommodate Fillmore's Case Grammar
within orthodox (of the moment) Chomskyan practice.

> >> Are there any categories of types of connections between units of
> >> meaning that I have left out of the above list?
>
> >> AtomicUnitOfMeaning: All of the meaning associated with as a single
> >> sense meaning of a single morpheme.http://en.wikipedia.org/wiki/Morpheme
>
> >> This atomic unit of meaning is larger than the meaning of a single Seme.
> >>    http://en.wikipedia.org/wiki/Seme_(semantics)
>
> >> A UnitOfMeaning is comprised (AtomicUnitOfMeaning or UnitsOfMeaning)
> >> possibly combined with (AtomicUnitOfMeaning or UnitsOfMeaning).
>
> >> UnitOfMeaning:
> >>     | AtomicUnitOfMeaning
> >>     | AtomicUnitOfMeaning * AtomicUnitOfMeaning
> >>     | UnitOfMeaning * AtomicUnitOfMeaning
> >>     | AtomicUnitOfMeaning * UnitOfMeaning

> >>     | UnitOfMeaning *  UnitOfMeaning-

[Harlan Messinger]

He didn't. He said nothing about universals.

[Nathan Sanders]

In article
<29b5b62e-cee0-4dc7...@t12g2000vbe.googlegroups.com>,

"Peter T. Daniels" <gram...@verizon.net> wrote:

> Why
> would you list "Case" as a thing separate from "Thematic Roles"?

Because of passives:

He saw her.
She was seen by him.

And verb pairs that express the same semantic proposition with the
same semantic roles, but assign them to different syntactic positions:

He fears her.
She scares him.

And verbs that alternate on their own which syntactic position a given
semantic role is assigned:

He worries about her.
She worries him.

And the middle construction in general:

He scares easily.
She scares him.

Plus, certain types of raising predicates can show an alternation:

He is tough to scare.
It is tough to scare him.

In all four pairs of sentences, both "he" and "him" have the
experiencer semantic role, but "he" have subject case, while "him" is
in object case.

Nathan

--
Department of Linguistics
Swarthmore College
http://sanders.phonologist.org/

[Robert Bannister]

On 20/08/12 8:02 AM, Nathan Sanders wrote:
> In article
> <29b5b62e-cee0-4dc7...@t12g2000vbe.googlegroups.com>,
> "Peter T. Daniels" <gram...@verizon.net> wrote:
>
>> Why
>> would you list "Case" as a thing separate from "Thematic Roles"?
>
> Because of passives:
>
> He saw her.
> She was seen by him.

And don't forget

He spoke to her.
She was spoken to by him.

The second sentence cannot be formed quite like that in languages that
have a real dative case.


--
Robert Bannister

[Peter Olcott]

This is my s\ource:
http://en.wikipedia.org/wiki/Agreement_(linguistics)#Functions

Common types of characteristics that may trigger grammatical agreement - in nouns - are:

• Grammatical person: e.g. Isn't he the king? versus Aren't you the king?
• Grammatical number: e.g. Isn't he? versus Aren't they?; or: One big car, versus Two big cars.
• Grammatical gender: e.g. She loves her cat, versus He loves his cat.
• Grammatical case e.g. Whose is he a friend of? versus Whom is he looking at?

would you list "Case" as a thing separate from "Thematic Roles"?

This case is the one listed under Agreement, (see above link).

"Theta theory" was an attempt to accommodate Fillmore's Case Grammar
within orthodox (of the moment) Chomskyan practice.

Are there any categories of types of connections between units of
meaning that I have left out of the above list?

AtomicUnitOfMeaning: All of the meaning associated with as a single
sense meaning of a single morpheme.http://en.wikipedia.org/wiki/Morpheme

This atomic unit of meaning is larger than the meaning of a single Seme.
   http://en.wikipedia.org/wiki/Seme_(semantics)

A UnitOfMeaning is comprised (AtomicUnitOfMeaning or UnitsOfMeaning)
possibly combined with (AtomicUnitOfMeaning or UnitsOfMeaning).

UnitOfMeaning:
    | AtomicUnitOfMeaning
    | AtomicUnitOfMeaning * AtomicUnitOfMeaning
    | UnitOfMeaning * AtomicUnitOfMeaning
    | AtomicUnitOfMeaning * UnitOfMeaning
    | UnitOfMeaning *  UnitOfMeaning-

0

[Peter T. Daniels]

On Aug 19, 8:02 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article
> <29b5b62e-cee0-4dc7-8070-1c8b61f12...@t12g2000vbe.googlegroups.com>,

>  "Peter T. Daniels" <gramma...@verizon.net> wrote:
>
> > Why
> > would you list "Case" as a thing separate from "Thematic Roles"?
>
> Because of passives:
>
>      He saw her.
>      She was seen by him.

Many (perhaps most) languages don't allow expression of the agent with
a passive. Contrary to what the early Relational Grammar people
claimed (the ones with "chomeurs"), passivization is not demotion of a
subject, but promotion of an object (or perhaps better, not demotion
of an agent but promotion of a patient).

> And verb pairs that express the same semantic proposition with the
> same semantic roles, but assign them to different syntactic positions:
>
>      He fears her.
>      She scares him.

How do you express that difference in languages with no morphological
case? (Such as English outside the pronouns.)

> And verbs that alternate on their own which syntactic position a given
> semantic role is assigned:
>
>      He worries about her.
>      She worries him.
>
> And the middle construction in general:
>
>      He scares easily.
>      She scares him.
>
> Plus, certain types of raising predicates can show an alternation:
>
>      He is tough to scare.
>      It is tough to scare him.
>
> In all four pairs of sentences, both "he" and "him" have the
> experiencer semantic role, but "he" have subject case, while "him" is
> in object case.

You are so incurably English-oriented! You reflect your upbringing
well.

[Peter T. Daniels]

>   * Grammatical person </wiki/Grammatical_person>: e.g. /_Is_n't _he_
>     the king?/ versus /_Are_n't _you_ the king?/
>   * Grammatical number </wiki/Grammatical_number>: e.g. /_Is_n't _he_?/
>     versus /_Are_n't _they_?/; or: /_One_ big _car_/, versus /_Two_ big
>     _cars_/.
>   * Grammatical gender </wiki/Grammatical_gender>: e.g. /_She_ loves
>     _her_ cat/, versus /_He_ loves _his_ cat/.
>   * Grammatical case </wiki/Grammatical_case> e.g. /_Whose_ is he a
>     friend _of_/? versus /_Whom_ is he looking _at_?/

>
> > would you list "Case" as a thing separate from "Thematic Roles"?
>
> This case is the one listed under Agreement, (see above link).

So you too are concerned only with English?

[Nathan Sanders]

In article
<3d2088de-ee81-4109...@z4g2000vby.googlegroups.com>,

"Peter T. Daniels" <gram...@verizon.net> wrote:

> On Aug 19, 8:02 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> > In article
> > <29b5b62e-cee0-4dc7-8070-1c8b61f12...@t12g2000vbe.googlegroups.com>,
> >  "Peter T. Daniels" <gramma...@verizon.net> wrote:
> >
> > > Why
> > > would you list "Case" as a thing separate from "Thematic Roles"?
> >
> > Because of passives:
> >
> >      He saw her.
> >      She was seen by him.
>
> Many (perhaps most) languages don't allow expression of the agent with
> a passive.

So? The fact that any language at all does it shows that case is
something separate from semantic roles.

> > And verb pairs that express the same semantic proposition with the
> > same semantic roles, but assign them to different syntactic positions:
> >
> >      He fears her.
> >      She scares him.
>
> How do you express that difference in languages with no morphological
> case? (Such as English outside the pronouns.)

With different verbs, as in the English pair above. Was this comment
meant for the next pair of sentences instead?

> > And verbs that alternate on their own which syntactic position a given
> > semantic role is assigned:
> >
> >      He worries about her.
> >      She worries him.

If your previous comment was supposed to go here, then my answer would
have been:

Sometimes languages have ambiguity, such as in the sentence "he felt
hot", where the experiencer of feeling the heat might be the subject
"he" or might be some unexpressed entity who is touching him, in which
case, he might not even being experiencing hot, which is common with
people who are running a fever but experience coldness.

> > And the middle construction in general:
> >
> >      He scares easily.
> >      She scares him.
> >
> > Plus, certain types of raising predicates can show an alternation:
> >
> >      He is tough to scare.
> >      It is tough to scare him.
> >
> > In all four pairs of sentences, both "he" and "him" have the
> > experiencer semantic role, but "he" have subject case, while "him" is
> > in object case.
>
> You are so incurably English-oriented!

Since all that's required to prove that case and semantic roles are
separate is data from one language in which they do differ, I'll pick
(i) the one in which it's easiest for me to find and type the data
(this isn't a research project, but a quick Usenet post, so I'd rather
not have to track down a native speaker, get their permission to
conduct research on them, collect the data from them, and then write
out the data, plus word-by-word glosses, plus full sentence glosses),
(ii) the one that will save time for the readers (so that they don't
have to work through glosses of an unfamiliar language, double-check
with their own native speakers to verify I haven't lied to them about
the data), and (iii) the one that usually saves space (except of
course when some uncooperative nincompoop decides that English data is
going to be his Issue of the Day, because he has to have some excuse
to start an argument).

[benl...@ihug.co.nz]

"Thematic Roles" correspond to Fillmore's proposed inventory of "deep
cases" -- possible participant roles. "Case" here means the simpler
system of "surface case" -- mainly subject/object -- which may be
overtly marked by morphology, word order etc. As Nathan's examples
show, Thematic Roles are variously mapped onto Case, so the two are
not the same thing.

[Peter Olcott]

For the moment, I am only concerned with English.

[Peter Olcott]

Are there any types of connections that are missing from my list?

[Peter Olcott]

Are there any types of connections between units of meaning that are
missing from my list?

[PeteOlcott]

On Aug 20, 12:21 am, "benli...@ihug.co.nz" <benli...@ihug.co.nz>
wrote:

> not the same thing.- Hide quoted text -
>
> - Show quoted text -

Thanks for your support.

[Peter T. Daniels]

On Aug 19, 11:51 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article

> <3d2088de-ee81-4109-a60f-fa06c30aa...@z4g2000vby.googlegroups.com>,

>  "Peter T. Daniels" <gramma...@verizon.net> wrote:
>
> > On Aug 19, 8:02 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> > > In article
> > > <29b5b62e-cee0-4dc7-8070-1c8b61f12...@t12g2000vbe.googlegroups.com>,
> > >  "Peter T. Daniels" <gramma...@verizon.net> wrote:
>
> > > > Why
> > > > would you list "Case" as a thing separate from "Thematic Roles"?
>
> > > Because of passives:
>
> > >      He saw her.
> > >      She was seen by him.
>
> > Many (perhaps most) languages don't allow expression of the agent with
> > a passive.
>
> So?  The fact that any language at all does it shows that case is
> something separate from semantic roles.
>
> > > And verb pairs that express the same semantic proposition with the
> > > same semantic roles, but assign them to different syntactic positions:
>
> > >      He fears her.
> > >      She scares him.
>
> > How do you express that difference in languages with no morphological
> > case? (Such as English outside the pronouns.)
>
> With different verbs, as in the English pair above.  Was this comment
> meant for the next pair of sentences instead?

Ok, looks like I have to explain. "He/him," "she/her" are "pronouns,"
and they are "inflected" for "case" -- quite exceptionally in English,
where the vast majority of nominals do _not_ exhibit inflection for
subjective and objective case.

By using them in your examples of, ostensibly, pairs of complementary
verbs, you muddle your example.

Squink for laziness.

[Peter T. Daniels]

> not the same thing.-

Theta-theory was an attempt to deal with the data assembled by
Fillmore. (In "The Case for Case," 1968.)

[Bill McCray]

On 8/19/2012 10:46 PM, Peter Olcott wrote:
>
> Common types of characteristics that may trigger grammatical agreement -
> in nouns - are:
>

> * Grammatical person </wiki/Grammatical_person>: e.g. /_Is_n't _he_
> the king?/ versus /_Are_n't _you_ the king?/
> * Grammatical number </wiki/Grammatical_number>: e.g. /_Is_n't _he_?/
> versus /_Are_n't _they_?/; or: /_One_ big _car_/, versus /_Two_ big
> _cars_/.
> * Grammatical gender </wiki/Grammatical_gender>: e.g. /_She_ loves
> _her_ cat/, versus /_He_ loves _his_ cat/.
> * Grammatical case </wiki/Grammatical_case> e.g. /_Whose_ is he a
> friend _of_/? versus /_Whom_ is he looking _at_?/

I've never heard of anything like "Whose is he a friend of?" Would
anyone actually say that?

These sound more normal:

"Whose friend is he?"
"Whom is he a friend of?"
"He is a friend of whom?"
"He is a friend of whose?"

You could make it "He is a friend of whose?" vs. "He is looking at whom?"

Bill in Kentucky

[Nathan Sanders]

In article
<bd13cb78-8bbb-462c...@g2g2000vba.googlegroups.com>,

"Peter T. Daniels" <gram...@verizon.net> wrote:

Well, yes. That's a fundamental assumption of all of these examples.

How else would you expect me to demonstrate that case and semantic
roles are separate, if not to use words that are inflected for case?
Had I used "John" or "the man" instead of "he/him", how could you tell
that the syntactic positions received different cases, and thus, that
case and semantic roles are separate?

> -- quite exceptionally in English,
> where the vast majority of nominals do _not_ exhibit inflection for
> subjective and objective case.

I'm not sure I understand why you're mentioning this. Do you think
that "exceptional" case isn't an example of case?

> By using them in your examples of, ostensibly, pairs of complementary
> verbs, you muddle your example.

The point of this pair was to show that two different verbs can

express the same semantic proposition with the same semantic roles,

but with different cases. It's parallel to the passive example, but
using two different lexical items in the active, rather than the same
lexical item in two different voices.

As I suspected, your comment would have gone better with the next
pair, and I answered your comment there (and you didn't respond).

For (i), it's really much more about practicality and speed (I don't
ordinarily have quick access to native speakers of other languages
when I'm writing Usenet posts[1]), and it's definitely not my laziness
for (ii) (that would be "accommodation for other people's laziness")
or (iii) (that would be "accommodation for concerns about storage
space, download time, and screen space, for those who care about any
of those things").

[1] Do you really think linguists just have a bunch of native speakers
of different languages sitting around in their closets, waiting to be
pulled out on a whim to have their syntax sufficiently explored to
determine whether or not their language ever has examples of case
patterning separately from semantic roles? As far as I know, the
closest native speaker of a non-English language at the time I wrote
that post was my neighbor, who speaks Lingala, but I'm pretty sure
Lingala doesn't have case, so that wouldn't have helped any!

[Nathan Sanders]

In article <gvKdnQVfdZkQk6_N...@giganews.com>,

> Are there any types of connections between units of meaning that are
> missing from my list?

I think instead of trying to be exhaustive from the get-go, you should
focus on making some small subset of things working well.

[PeteOlcott]

On Aug 20, 10:07 am, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <gvKdnQVfdZkQk6_NnZ2dnUVZ_usAA...@giganews.com>,

>  Peter Olcott <OCR4Screen> wrote:
>
>
>
>
>
> > On 8/19/2012 10:51 PM, Nathan Sanders wrote:
> > > In article

> > > <3d2088de-ee81-4109-a60f-fa06c30aa...@z4g2000vby.googlegroups.com>,

> Swarthmore Collegehttp://sanders.phonologist.org/- Hide quoted text -

>
> - Show quoted text -

I already have simple things working well with first order predicate
logic.
I can see how these can be extended with higher order predicate logic
(or Type Theory).

Therefore I am beginning with an attempt to define the criterion
measure of not(exhaustively expressive notation).

[Peter T. Daniels]

On Aug 20, 11:03 am, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article

> <bd13cb78-8bbb-462c-894d-8bf1a895f...@g2g2000vba.googlegroups.com>,

But that wasn't the question. Peter O appears to be trying to assemble
a universal set of categories, and Agreement and Morphological Case
are clearly not universal.

> > -- quite exceptionally in English,
> > where the vast majority of nominals do _not_ exhibit inflection for
> > subjective and objective case.
>
> I'm not sure I understand why you're mentioning this.  Do you think
> that "exceptional" case isn't an example of case?
>
> > By using them in your examples of, ostensibly, pairs of complementary
> > verbs, you muddle your example.
>
> The point of this pair was to show that two different verbs can
> express the same semantic proposition with the same semantic roles,
> but with different cases.  It's parallel to the passive example, but
> using two different lexical items in the active, rather than the same
> lexical item in two different voices.
>
> As I suspected, your comment would have gone better with the next
> pair, and I answered your comment there (and you didn't respond).

I answered it the first time it appeared.

In the olden days, linguists were trained on things like Gleason's
Workbook that accompanied his Textbook, and Elson & Pickett's Workbook
of Morphology. Between them, they contained scores of examples of the
many kinds of phonemic systems and morphological patterning that have
been described from around the world. They constituted a file of ready-
to-hand examples of just about every phenomenon that might come up in
a discussion.

These days, phonologists behave as though phonemic systems are simply
somehow just "given," without even a paragraph about how one goes
about determining the phonemes of a language.

[Nathan Sanders]

In article
<d16abcad-b0c0-46b5...@k20g2000vbk.googlegroups.com>,

"Peter T. Daniels" <gram...@verizon.net> wrote:

Yes it was. In fact, it was your question. You asked, "Why would you

list 'Case' as a thing separate from 'Thematic Roles'?"

> Peter O appears to be trying to assemble
> a universal set of categories,

No he wasn't. He was trying to assemble a "complete list of every
type". That's a very different thing, a difference that was already
pointed out to you by Harlan (who you ignored), and by Peter himself
(who you ridiculed), and would be obvious to you if you had more
interest in paying attention to what people were actually saying,
rather than in trying find an argument to start.

> and Agreement and Morphological Case
> are clearly not universal.

If it's so clear, then clearly he must not be talking about
universals. Why would you assume he was? Especially given that
"universal" appears nowhere in his message, but "complete list of
every type" does.

how on earth could get from "complete list of every type" to
"universal"? Would you think a "complete list of every type" of
fricative is a "universal" set of fricatives that every language has?
Certainly not, so why would you think so here?

> > > -- quite exceptionally in English,
> > > where the vast majority of nominals do _not_ exhibit inflection for
> > > subjective and objective case.
> >
> > I'm not sure I understand why you're mentioning this. �Do you think
> > that "exceptional" case isn't an example of case?
> >
> > > By using them in your examples of, ostensibly, pairs of complementary
> > > verbs, you muddle your example.
> >
> > The point of this pair was to show that two different verbs can
> > express the same semantic proposition with the same semantic roles,
> > but with different cases. �It's parallel to the passive example, but
> > using two different lexical items in the active, rather than the same
> > lexical item in two different voices.
> >
> > As I suspected, your comment would have gone better with the next
> > pair, and I answered your comment there (and you didn't respond).
>
> I answered it the first time it appeared.

You gave no response to the issue of ambiguity (your use of
"complementary verbs" in your response pretty much guarantees you
aren't talking about ambiguity). Just look:

> > > > > > And verbs that alternate on their own which syntactic position a
> > > > > > given
> > > > > > semantic role is assigned:
> >
> > > > > > � � �He worries about her.
> > > > > > � � �She worries him.
> >
> > > > If your previous comment was supposed to go here, then my answer would
> > > > have been:
> >
> > > > Sometimes languages have ambiguity, such as in the sentence "he felt
> > > > hot", where the experiencer of feeling the heat might be the subject
> > > > "he" or might be some unexpressed entity who is touching him, in which
> > > > case, he might not even being experiencing hot, which is common with
> > > > people who are running a fever but experience coldness.

The preceding is the part that you didn't respond to.

Sure, and as a phonologist, I can come up with such examples off the
top of my head for most kinds of phonological phenomena.

But this is a discussion about the syntax-semantics interface, and I
am neither a syntactician nor a semanticist, let alone both. My
memorized cache of non-English examples relevant to this discussion is
much more limited.

> These days, phonologists behave as though phonemic systems are simply
> somehow just "given," without even a paragraph about how one goes
> about determining the phonemes of a language.

Obvious red herring alert! Variations in the alignment between case
and semantic roles (the current topic) has pretty much nothing at all
to do with how anyone's particular pet phonological theory organizes a
language's sounds into abstract units.

[Glenn Knickerbocker]

On Mon, 20 Aug 2012 12:57:45 -0700 (PDT), Peter T. Daniels wrote:
>But that wasn't the question. Peter O appears to be trying to assemble
>a universal set of categories, and Agreement and Morphological Case
>are clearly not universal.

A universal set of categories is not necessarily a set of universal
categories.

ŹR "I love Blip just because it's the absolute opposite of fun"
http://users.bestweb.net/~notr/travelog/19990710.html --Kibo

[Nathan Sanders]

In article <scb5385tm1ekuqnqv...@4ax.com>,

Glenn Knickerbocker <No...@bestweb.net> wrote:

> On Mon, 20 Aug 2012 12:57:45 -0700 (PDT), Peter T. Daniels wrote:
> >But that wasn't the question. Peter O appears to be trying to assemble
> >a universal set of categories, and Agreement and Morphological Case
> >are clearly not universal.
>
> A universal set of categories is not necessarily a set of universal
> categories.

Thank you!

[Jerry Friedman]

On Aug 20, 3:36 pm, Glenn Knickerbocker <N...@bestweb.net> wrote:
> On Mon, 20 Aug 2012 12:57:45 -0700 (PDT), Peter T. Daniels wrote:
> >But that wasn't the question. Peter O appears to be trying to assemble
> >a universal set of categories, and Agreement and Morphological Case
> >are clearly not universal.
>
> A universal set of categories is not necessarily a set of universal
> categories.

In fact, since set theory is clearly on topic in both groups, a union
is not the same as an intersection.

--
Jerry Friedman

[Peter Olcott]

The purpose of this list is to end up with an I-Language notational
convention that is capable of fully representing all of the semantic
meaning of every concept that can be expressed in words.

>> and Agreement and Morphological Case
>> are clearly not universal.
> If it's so clear, then clearly he must not be talking about
> universals. Why would you assume he was? Especially given that
> "universal" appears nowhere in his message, but "complete list of
> every type" does.
>
> how on earth could get from "complete list of every type" to
> "universal"? Would you think a "complete list of every type" of
> fricative is a "universal" set of fricatives that every language has?
> Certainly not, so why would you think so here?
>

I may have mentioned the I-language aspect of my goal in other postings.

[Nathan Sanders]

In article <TKidnV5Uv4UOJq_N...@giganews.com>,

Which is probably so ambitious as to be impossible (which I why I
cautioned you elsewhere to limit the scope of your investigation; at
minimum, I would suggest you limit your *discussion* of your intended
scope of your investigation, because the more often you say things
like "fully representing all of the semantic meaning of every concept
that can be expressed in words", the less credible you sound).

> >> and Agreement and Morphological Case
> >> are clearly not universal.
> > If it's so clear, then clearly he must not be talking about
> > universals. Why would you assume he was? Especially given that
> > "universal" appears nowhere in his message, but "complete list of
> > every type" does.
> >
> > how on earth could get from "complete list of every type" to
> > "universal"? Would you think a "complete list of every type" of
> > fricative is a "universal" set of fricatives that every language has?
> > Certainly not, so why would you think so here?
> >
> I may have mentioned the I-language aspect of my goal in other postings.

And I-language is speaker/language-specific, so it is mostly composed
of things that are not universal.

[Nathan Sanders]

In article <sanders-A9B2DE...@free.teranews.com>,

Nathan Sanders <san...@alum.mit.edu> wrote:

> In article <TKidnV5Uv4UOJq_N...@giganews.com>,
> Peter Olcott <OCR4Screen> wrote:
>
> > The purpose of this list is to end up with an I-Language notational
> > convention that is capable of fully representing all of the semantic
> > meaning of every concept that can be expressed in words.
>
> Which is probably so ambitious as to be impossible (which I why I
> cautioned you elsewhere to limit the scope of your investigation; at
> minimum, I would suggest you limit your *discussion* of your intended
> scope of your investigation, because the more often you say things
> like "fully representing all of the semantic meaning of every concept
> that can be expressed in words", the less credible you sound).

By the way, I don't intend to mean here, just direct and honest. In
general, the smaller the scope of your project, the more likely it is
to succeed.

[António Marques]

On Aug 20, 11:47 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <sanders-A9B2DE.18441020082...@free.teranews.com>,
>  Nathan Sanders <sand...@alum.mit.edu> wrote:
>
> > In article <TKidnV5Uv4UOJq_NnZ2dnUVZ_r6dn...@giganews.com>,

> >  Peter Olcott <OCR4Screen> wrote:
>
> > > The purpose of this list is to end up with an I-Language notational
> > > convention that is capable of fully representing all of the semantic
> > > meaning of every concept that can be expressed in words.
>
> > Which is probably so ambitious as to be impossible (which I why I
> > cautioned you elsewhere to limit the scope of your investigation; at
> > minimum, I would suggest you limit your *discussion* of your intended
> > scope of your investigation, because the more often you say things
> > like "fully representing all of the semantic meaning of every concept
> > that can be expressed in words", the less credible you sound).
>
> By the way, I don't intend to mean here, just direct and honest.  In
> general, the smaller the scope of your project, the more likely it is
> to succeed.

But what if the all-encompassing-ness of it is the [chief] purpose? Mr
Olcott has shown willingness to reduce the breadth of his goal, but
not the depth (or the other way round). There _are_ already people
working on limited scopes, isn't it?

[Nathan Sanders]

In article
<afb30f9e-9ee3-4a26...@cf4g2000vbb.googlegroups.com>,

As I said, "at minimum, I would suggest you limit your *discussion* of
your intended scope of your investigation".

There's only so much discussion of A Grand Theory of Everything that
reasonable can tolerate, when the first steps haven't even been
finished.

[benl...@ihug.co.nz]

Yes, I know, you already said that. I was trying to explain why "Case"
and "Thematic Roles", as used by present day Chomskyans, are not the
same thing.

[Peter Olcott]

So far I have not yet encountered a single counter-example showing that
a system such as this does not already exist in several of the current
ontology languages, such as CycL.

Designing a system capable of representing any possible conception is
not nearly as difficult as populating this system with the finite set of
every currently written conception. This has to be automated, it is
infeasible otherwise. To automate this requires teaching the machine to
read.

I am starting with groking compositionality. That is the purpose of this
thread.
http://plato.stanford.edu/entries/compositionality/
http://en.wikipedia.org/wiki/Grok

Problems of enormous complexity can be broken down into a hierarchy of
simple steps.

> minimum, I would suggest you limit your *discussion* of your intended
> scope of your investigation, because the more often you say things
> like "fully representing all of the semantic meaning of every concept
> that can be expressed in words", the less credible you sound).

Credibility is a fake measure of truth.
I am thinking that my insights into compositionality will establish
"credibility".
http://plato.stanford.edu/entries/compositionality/#4.2.1

(3) Most students will succeed if they work hard.
<Most> x (WorkHard(x)--->Succeed(x))

(3′) Most students who work hard will succeed.
<Most> x (WorkHard(x) & Succeed(x))

>
>>>> and Agreement and Morphological Case
>>>> are clearly not universal.
>>> If it's so clear, then clearly he must not be talking about
>>> universals. Why would you assume he was? Especially given that
>>> "universal" appears nowhere in his message, but "complete list of
>>> every type" does.
>>>
>>> how on earth could get from "complete list of every type" to
>>> "universal"? Would you think a "complete list of every type" of
>>> fricative is a "universal" set of fricatives that every language has?
>>> Certainly not, so why would you think so here?
>>>
>> I may have mentioned the I-language aspect of my goal in other postings.
> And I-language is speaker/language-specific, so it is mostly composed
> of things that are not universal.
>
> Nathan
>

I did not get that from:

Functional vs. lexical: a cognitive dichotomy, Ronnie Cann.

A theory of categorisation at different levels of the grammar is
proposed that allows language specific, E-linguistic,
expressions to be related to universal (I-linguistic) categories that
can be manipulated by principles of Universal Grammar.

I thought that he was referring to I-Language as universal, he must have
meant a universal instance of I-Language.

[Peter Olcott]

I try not to take offense even when offense is clearly intended.
I could tell that you did not intend any offense.

I have learned to think within a hierarchy of goals, that may be my
strongest strength.
The sub-topic of compositionality would seem to lay the foundation for
my higher goal.

[Peter Olcott]

It seems that focusing on compositionality may be the best limited scope
goal that would lay the foundation for the next larger goal.
http://plato.stanford.edu/entries/compositionality/

[Peter Olcott]

On 8/20/2012 6:05 PM, Nathan Sanders wrote:
> In article

> <afb30f9e-9ee3-4a26...@cf4g2000vbb.googlegroups.com>,

Someone has to focus on the Grande Scheme, even if at an extremely high
level of abstraction.

It is my opinion that All of the failures of computer science in the
field of natural language processing are because the scope was always
too limited to see enough details to get the design anywhere near the
ball-park of sufficiently specified.

[António Marques]

On Aug 21, 12:51 am, Peter Olcott <OCR4Screen> wrote:
> On 8/20/2012 6:05 PM, Nathan Sanders wrote:
>
> > In article

> > <afb30f9e-9ee3-4a26-8445-bfe94f011...@cf4g2000vbb.googlegroups.com>,

> Someone has to focus on the Grande Scheme, even if at an extremely high
> level of abstraction.
>
> It is my opinion that All of the failures of computer science in the
> field of natural language processing are because the scope was always
> too limited to see enough details to get the design anywhere near the
> ball-park of sufficiently specified.

I wouldn't say all, but that's certainly the reason some approaches
fail. Put otherwise, I think it is more legitimate to hold that 'How
to Produce the Grand Scheme?' is in itself a field of study, than to
hold that partial limited approaches will just combine gracefully at
some point (someone please CSC that).

But it's very hard to discuss it in terms that don't eventually annoy
everyone. Perhaps the first step should be 'how can I discuss it in
limited, digestible terms?'. (I think that's not very different from
what Nathan's saying above.)

[Nathan Sanders]

In article <KbmdnWSbX-gKV6_N...@giganews.com>,

Peter Olcott <OCR4Screen> wrote:

> On 8/20/2012 5:44 PM, Nathan Sanders wrote:
>
> > minimum, I would suggest you limit your *discussion* of your intended
> > scope of your investigation, because the more often you say things
> > like "fully representing all of the semantic meaning of every concept
> > that can be expressed in words", the less credible you sound).
>
> Credibility is a fake measure of truth.

But credibility is what helps bring you into meaningful contact with
knowledgeable people, who can help you improve your project through
exchange of ideas (whether it's by direct contribution or through the
process of debate).

If everyone of any intellectual value to you thinks you are a crank,
you run the risk of them ignoring you, and depriving you of their
insights.

> >> I may have mentioned the I-language aspect of my goal in other postings.
> >
> > And I-language is speaker/language-specific, so it is mostly composed
> > of things that are not universal.
>

> I did not get that from:
>
> Functional vs. lexical: a cognitive dichotomy, Ronnie Cann.
>
> A theory of categorisation at different levels of the grammar is
> proposed that allows language specific, E-linguistic,
> expressions to be related to universal (I-linguistic) categories that
> can be manipulated by principles of Universal Grammar.
>
> I thought that he was referring to I-Language as universal, he must have
> meant a universal instance of I-Language.

As he states in section 3.4, "I-language may be construed as a
metalanguage that generates (or otherwise characterises) E-language
and is equated in Chomsky 1986 with a parametrised state of Universal
Grammar". That is, it is not equivalent to UG itself, but rather, it
is a particular, language-specific state of UG, with parameters set in
one configuration.

But overall, Cann talks about E- and I-language in a rather confusing
and misleading way. I wouldn't rely on his work for understanding
these notions, and I certainly wouldn't quote from him for definitions.

[Nathan Sanders]

In article <sanders-4E1800...@free.teranews.com>,

As an example of how I-language is not universal, I-language is
generally taken to include the lexicon, but obviously, there is no
universal lexicon.

[Peter Olcott]

Compositionality! (ie How do units of meaning fit together?)

[Peter Olcott]

On 8/20/2012 7:19 PM, Nathan Sanders wrote:
> In article <KbmdnWSbX-gKV6_N...@giganews.com>,
> Peter Olcott <OCR4Screen> wrote:
>
>> On 8/20/2012 5:44 PM, Nathan Sanders wrote:
>>
>>> minimum, I would suggest you limit your *discussion* of your intended
>>> scope of your investigation, because the more often you say things
>>> like "fully representing all of the semantic meaning of every concept
>>> that can be expressed in words", the less credible you sound).
>> Credibility is a fake measure of truth.
> But credibility is what helps bring you into meaningful contact with
> knowledgeable people, who can help you improve your project through
> exchange of ideas (whether it's by direct contribution or through the
> process of debate).
>
> If everyone of any intellectual value to you thinks you are a crank,
> you run the risk of them ignoring you, and depriving you of their
> insights.

Not if I can prove myself the way that Saul Kripke did when he was only 17.

>>>> I may have mentioned the I-language aspect of my goal in other postings.
>>> And I-language is speaker/language-specific, so it is mostly composed
>>> of things that are not universal.
>> I did not get that from:
>>
>> Functional vs. lexical: a cognitive dichotomy, Ronnie Cann.
>>
>> A theory of categorisation at different levels of the grammar is
>> proposed that allows language specific, E-linguistic,
>> expressions to be related to universal (I-linguistic) categories that
>> can be manipulated by principles of Universal Grammar.
>>
>> I thought that he was referring to I-Language as universal, he must have
>> meant a universal instance of I-Language.
> As he states in section 3.4, "I-language may be construed as a
> metalanguage that generates (or otherwise characterises) E-language
> and is equated in Chomsky 1986 with a parametrised state of Universal
> Grammar". That is, it is not equivalent to UG itself, but rather, it
> is a particular, language-specific state of UG, with parameters set in
> one configuration.
>

I am just thinking of it as the structure of knowledge within the mind.
It is a hierarchial structure, an acyclic di-graph.

> But overall, Cann talks about E- and I-language in a rather confusing
> and misleading way. I wouldn't rely on his work for understanding
> these notions, and I certainly wouldn't quote from him for definitions.
>
> Nathan
>

Although It is helpful to have some sort of lingua franca to talk about
my ideas in forums such as this.

I don't really need correct terminology for the work that I am doing, I
only need concepts with labels.
I am not taking the words that I am reading as constituents of my
design, I only need the concepts.

[Peter Olcott]

My design requires it, yet I call it an ontology rather than a lexicon.
It only has to have enough knowledge to know how to read.
I used to call this BootStrap English.

[Nathan Sanders]

In article <eMWdne2JDZyFR6_N...@giganews.com>,

Peter Olcott <OCR4Screen> wrote:

> On 8/20/2012 7:19 PM, Nathan Sanders wrote:
> > In article <KbmdnWSbX-gKV6_N...@giganews.com>,
> > Peter Olcott <OCR4Screen> wrote:
> >
> >> On 8/20/2012 5:44 PM, Nathan Sanders wrote:
> >>
> >>> minimum, I would suggest you limit your *discussion* of your intended
> >>> scope of your investigation, because the more often you say things
> >>> like "fully representing all of the semantic meaning of every concept
> >>> that can be expressed in words", the less credible you sound).
> >>
> >> Credibility is a fake measure of truth.
> >
> > But credibility is what helps bring you into meaningful contact with
> > knowledgeable people, who can help you improve your project through
> > exchange of ideas (whether it's by direct contribution or through the
> > process of debate).
> >
> > If everyone of any intellectual value to you thinks you are a crank,
> > you run the risk of them ignoring you, and depriving you of their
> > insights.
>
> Not if I can prove myself the way that Saul Kripke did when he was only 17.

The number of geniuses who will turn out to be a giant in their field
the way Kripke did is astronomically small. Comparing oneself to such
a figure is a sure-fire way to make everyone think you're a crank.

> >>>> I may have mentioned the I-language aspect of my goal in other postings.
> >>> And I-language is speaker/language-specific, so it is mostly composed
> >>> of things that are not universal.
> >> I did not get that from:
> >>
> >> Functional vs. lexical: a cognitive dichotomy, Ronnie Cann.
> >>
> >> A theory of categorisation at different levels of the grammar is
> >> proposed that allows language specific, E-linguistic,
> >> expressions to be related to universal (I-linguistic) categories that
> >> can be manipulated by principles of Universal Grammar.
> >>
> >> I thought that he was referring to I-Language as universal, he must have
> >> meant a universal instance of I-Language.
> >
> > As he states in section 3.4, "I-language may be construed as a
> > metalanguage that generates (or otherwise characterises) E-language
> > and is equated in Chomsky 1986 with a parametrised state of Universal
> > Grammar". That is, it is not equivalent to UG itself, but rather, it
> > is a particular, language-specific state of UG, with parameters set in
> > one configuration.
> >
> I am just thinking of it as the structure of knowledge within the mind.
> It is a hierarchial structure, an acyclic di-graph.

How you choose to represent it in your personal theory has no bearing
on whether it is universal or not.

> > But overall, Cann talks about E- and I-language in a rather confusing
> > and misleading way. I wouldn't rely on his work for understanding
> > these notions, and I certainly wouldn't quote from him for definitions.
>

> Although It is helpful to have some sort of lingua franca to talk about
> my ideas in forums such as this.
>
> I don't really need correct terminology for the work that I am doing, I
> only need concepts with labels.
>
> I am not taking the words that I am reading as constituents of my
> design, I only need the concepts.

If you use existing terminology incorrectly or insist on using your
own idiosyncratic set of novel jargon for things that already have
perfectly good labels, at best, people aren't going to understand you,
and at worst, they're going to think you're a crank.

Either way, you're going to have a lot of trouble having useful
interactions with them!

[Peter Olcott]

Knowledge already has its own natural structure. The structure could be
fit into a connected set of nodes, much like the Theory of Types.

>>> But overall, Cann talks about E- and I-language in a rather confusing
>>> and misleading way. I wouldn't rely on his work for understanding
>>> these notions, and I certainly wouldn't quote from him for definitions.
>> Although It is helpful to have some sort of lingua franca to talk about
>> my ideas in forums such as this.
>>
>> I don't really need correct terminology for the work that I am doing, I
>> only need concepts with labels.
>>
>> I am not taking the words that I am reading as constituents of my
>> design, I only need the concepts.
> If you use existing terminology incorrectly or insist on using your
> own idiosyncratic set of novel jargon for things that already have
> perfectly good labels, at best, people aren't going to understand you,
> and at worst, they're going to think you're a crank.
>
> Either way, you're going to have a lot of trouble having useful
> interactions with them!
>
> Nathan
>

I am trying to use the words as I understand them from their context.
Some contexts are too limited to infer the meanings.

[Peter Olcott]

I will try to conform to the common language, feel free to correct me.
Some of these words I have only seen on one context.

Here are two feasible areas for investigation:

Determine the current degree of progress on forming an ontology capable
of storing every element in the set of conceptions. Find the boundaries
of this research.

Determine the different types of connections that can be made using
units of meaning from language.

[Peter T. Daniels]

On Aug 20, 6:34 pm, Peter Olcott <OCR4Screen> wrote:
> On 8/20/2012 4:04 PM, Nathan Sanders wrote:
> >   "Peter T. Daniels" <gramma...@verizon.net> wrote:

> >> Peter O appears to be trying to assemble
> >> a universal set of categories,
> > No he wasn't.  He was trying to assemble a "complete list of every
> > type".  That's a very different thing, a difference that was already
> > pointed out to you by Harlan (who you ignored), and by Peter himself
> > (who you ridiculed), and would be obvious to you if you had more
> > interest in paying attention to what people were actually saying,
> > rather than in trying find an argument to start.
>
> The purpose of this list is to end up with an I-Language notational
> convention that is capable of fully representing all of the semantic
> meaning of every concept that can be expressed in words.

I-Language is, by definition, pretty much universal, isn't it? Is it
anything more than a newer label for UG?

> >> and Agreement and Morphological Case
> >> are clearly not universal.
> > If it's so clear, then clearly he must not be talking about
> > universals.  Why would you assume he was?  Especially given that
> > "universal" appears nowhere in his message, but "complete list of
> > every type" does.
>
> > how on earth could get from "complete list of every type" to
> > "universal"?  Would you think a "complete list of every type" of
> > fricative is a "universal" set of fricatives that every language has?
> > Certainly not, so why would you think so here?
>

> I may have mentioned the I-language aspect of my goal in other postings.-

I think you did. Nathan probably missed it, since he seems to look in
most threads only at my postings, for the purpose of finding something
to misunderstand and pick at.

At extremely excessive length.

[Peter Olcott]

On 8/19/2012 3:26 PM, Glenn Knickerbocker wrote:

> On Sun, 19 Aug 2012 10:33:56 -0700 (PDT), Peter T. Daniels wrote:
>> Why would "Agreement" have anything at all to do with semantics?

> Why would you expect "every type of syntactic or semantic connection" to
> have something to do with semantics?

Language is all about connecting smaller units of meaning into higher
units of meaning and the connections are often syntactical.

>
> ¬R - we don't have branes made of clockworks, chocolate maidens
> http://users.bestweb.net/~notr/arkville.html --the Ur-beatle

[Peter Olcott]

On 8/19/2012 12:33 PM, Peter T. Daniels wrote:
> On Aug 19, 1:30 pm, Peter Olcott <OCR4Screen> wrote:
>> I am trying to derive a complete list of every type of syntactic or
>> semantic connection between units of meaning (defined below). This is my
>> current list:
>>
>> 1) Modification (Adverb, Adjective, Quantifier)
>> 2) Anaphora (Pronouns)
>> 3) Tense (location in time)
>> 4) Aspect (fabric of time)
>> 5) Mood (modality, evidentiality)
>> 6) Thematic Roles (Agent, Patient, Instrument, etc)
>> 7) Agreement (Person, Number, Gender, Case)
>> 8) Subcatorization (Verbs select their syntactic arguments)
>> 9) Selection Restriction (Predicates select their semantic arguments)
>>
>> Some of the items on the above list may be imperfectly specified, I am
>> just learning these terms for the first time. My primary goal (at this
>> point) is to make the basic list of categories of types of connections
>> as complete as possible.

> Why would "Agreement" have anything at all to do with semantics?

I am trying to simplify the problem by dividing the categories of syntax
and semantics a little differently.
I have had some of these ideas in my mind for 25 years.

Now what we have is many different kinds of connections between units of
meaning. (Listed above)
We simply consider all of these Connections to be Syntax, and the rest
are Semantics.

>
>> Are there any categories of types of connections between units of
>> meaning that I have left out of the above list?
>>
>> AtomicUnitOfMeaning: All of the meaning associated with as a single
>> sense meaning of a single morpheme.http://en.wikipedia.org/wiki/Morpheme
>>
>> This atomic unit of meaning is larger than the meaning of a single Seme.
>> http://en.wikipedia.org/wiki/Seme_(semantics)
>>
>> A UnitOfMeaning is comprised (AtomicUnitOfMeaning or UnitsOfMeaning)
>> possibly combined with (AtomicUnitOfMeaning or UnitsOfMeaning).
>>
>> UnitOfMeaning:
>> | AtomicUnitOfMeaning
>> | AtomicUnitOfMeaning * AtomicUnitOfMeaning
>> | UnitOfMeaning * AtomicUnitOfMeaning
>> | AtomicUnitOfMeaning * UnitOfMeaning
>> | UnitOfMeaning * UnitOfMeaning

[PeteOlcott]

On Aug 19, 5:08 pm, Harlan Messinger <h.rem...@gavelcade.com> wrote:
> On 8/19/2012 4:37 PM, Peter T. Daniels wrote:

>
>
>
>
>
> > On Aug 19, 4:30 pm, Peter Olcott <OCR4Screen> wrote:
> >> On 8/19/2012 12:33 PM, Peter T. Daniels wrote:
>
> >>> On Aug 19, 1:30 pm, Peter Olcott <OCR4Screen> wrote:
> >>>> I am trying to derive a complete list of every type of syntactic or
> >>>> semantic connection between units of meaning (defined below). This is my
> >>>> current list:
>
> >>>>       1) Modification          (Adverb, Adjective, Quantifier)
> >>>>       2) Anaphora              (Pronouns)
> >>>>       3) Tense                 (location in time)
> >>>>       4) Aspect                (fabric of time)
> >>>>       5) Mood                  (modality, evidentiality)
> >>>>       6) Thematic Roles        (Agent, Patient, Instrument, etc)
> >>>>       7) Agreement             (Person, Number, Gender, Case)
> >>>>       8) Subcatorization       (Verbs select their syntactic arguments)
> >>>>       9) Selection Restriction (Predicates select their semantic arguments)
>
> >>>> Some of the items on the above list may be imperfectly specified, I am
> >>>> just learning these terms for the first time. My primary goal (at this
> >>>> point) is to make the basic list of categories of types of connections
> >>>> as complete as possible.
> >>> Why would "Agreement" have anything at all to do with semantics?
>

> >> This whole thread is about connections between units of meaning. Syntax
> >> is an aspect of the connections between units of meaning, and
> >> "Agreement" is an aspect of this syntactic connection.
>
> > Why would you list "Agreement" (a property of some languages) in a
> > list of putative universals?
>

> He didn't. He said nothing about universals.- Hide quoted text -
>
> - Show quoted text -

Ultimately the semantics that underlies the syntax would be universal.

[PeteOlcott]

On Aug 20, 8:11 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <eMWdne2JDZyFR6_NnZ2dnUVZ_hGdn...@giganews.com>,

>  Peter Olcott <OCR4Screen> wrote:
>
>
>
>
>
> > On 8/20/2012 7:19 PM, Nathan Sanders wrote:

> > > In article <KbmdnWSbX-gKV6_NnZ2dnUVZ_j2dn...@giganews.com>,

> > >   Peter Olcott <OCR4Screen> wrote:
>
> > >> On 8/20/2012 5:44 PM, Nathan Sanders wrote:
>
> > >>> minimum, I would suggest you limit your *discussion* of your intended
> > >>> scope of your investigation, because the more often you say things
> > >>> like "fully representing all of the semantic meaning of every concept
> > >>> that can be expressed in words", the less credible you sound).
>
> > >> Credibility is a fake measure of truth.
>
> > > But credibility is what helps bring you into meaningful contact with
> > > knowledgeable people, who can help you improve your project through
> > > exchange of ideas (whether it's by direct contribution or through the
> > > process of debate).
>
> > > If everyone of any intellectual value to you thinks you are a crank,
> > > you run the risk of them ignoring you, and depriving you of their
> > > insights.
>
> > Not if I can prove myself the way that Saul Kripke did when he was only 17.
>
> The number of geniuses who will turn out to be a giant in their field
> the way Kripke did is astronomically small.  Comparing oneself to such
> a figure is a sure-fire way to make everyone think you're a crank.

Kripke has established that credentials are not required for
credibility.
Whether or not I can derive anything of significance in this field has
not yet been established.

[Nathan Sanders]

In article
<75752086-cfca-4e9f...@n19g2000yqc.googlegroups.com>,

I never said anything about credentials.

What I am talking about is whether knowledgeable people will choose to
interact with you in meaningful ways, a decision they will make in
large part based on how you present yourself and your ideas.

If you present yourself as a crank, you may find yourself receiving
far less help than you might want, let alone need, and your road will
consequently be a much harder one to travel.

[Nathan Sanders]

In article
<7556e7d8-7bed-4e2f...@q35g2000yqm.googlegroups.com>,

> Ultimately the semantics that underlies the syntax would be universal.

Unlikely (especially if you're including pragmatics and lexical
semantics, which you need to), but the syntax itself definitely is
not. So a "complete list of every semantic or syntactic connection"
can't possibly be a list of universals.

[PeteOlcott]

On Aug 22, 2:17 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article

> <75752086-cfca-4e9f-99ed-63b650acc...@n19g2000yqc.googlegroups.com>,

No you were talking about my credibility.

The reason that I brought up Kripke in the first place was that he has
established the precedent that credentials are not required for
credibility. In other words it is possible to skip the credibility
step entirely and move directly to validity.

> What I am talking about is whether knowledgeable people will choose to
> interact with you in meaningful ways, a decision they will make in
> large part based on how you present yourself and your ideas.
>
> If you present yourself as a crank, you may find yourself receiving
> far less help than you might want, let alone need, and your road will
> consequently be a much harder one to travel.
>
> Nathan
>
> --
> Department of Linguistics

> Swarthmore Collegehttp://sanders.phonologist.org/- Hide quoted text -

[Nathan Sanders]

In article
<62617a03-0b02-4764...@d6g2000pbr.googlegroups.com>,

You still seem not to understand what I'm saying, and I'm not sure how
many more different ways I can say it. I'll try again:

This isn't about credentials at all. Mentioning them yet again will
only further demonstrate that you still don't understand.

What I'm talking about is how a person presents himself and his ideas
has a significant effect on his ability to have meaningful
interactions with people who might be able to provide valuable
contributions to his work (helping to make the work go easier and
quicker, having more robust results, etc.).

[PeteOlcott]

On Aug 22, 2:58 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article

> <62617a03-0b02-4764-8faf-1a4bcbb1c...@d6g2000pbr.googlegroups.com>,

Only in the case of most people whom have not dismissed credibility as

a fake measure of truth.

The inventor of the steam boat was dismissed as a crank, "Fulton's
folly."

>
> > > What I am talking about is whether knowledgeable people will choose to
> > > interact with you in meaningful ways, a decision they will make in
> > > large part based on how you present yourself and your ideas.
>
> > > If you present yourself as a crank, you may find yourself receiving
> > > far less help than you might want, let alone need, and your road will
> > > consequently be a much harder one to travel.
>
> Nathan
>
> --
> Department of Linguistics

[Nathan Sanders]

In article
<d9e4d266-c07e-4a09...@wq14g2000pbb.googlegroups.com>,

I give up.

[PeteOlcott]

On Aug 22, 2:38 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article

> <7556e7d8-7bed-4e2f-9181-c89642efb...@q35g2000yqm.googlegroups.com>,

> Swarthmore Collegehttp://sanders.phonologist.org/- Hide quoted text -

>
> - Show quoted text -

When I use the word {semantics} I am referring to any unit of meaning
that was ever expressed.
Is there a better term (within linguistics) that fully encapsulates
this concept?

[pauljk]

"Nathan Sanders" <san...@alum.mit.edu> wrote in message
news:sanders-C8AEC1...@free.teranews.com...

> In article
> <d9e4d266-c07e-4a09...@wq14g2000pbb.googlegroups.com>,
> PeteOlcott <peteo...@gmail.com> wrote:
>> On Aug 22, 2:58 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
>> > In article
>> > <62617a03-0b02-4764-8faf-1a4bcbb1c...@d6g2000pbr.googlegroups.com>,
>> > PeteOlcott <peteolc...@gmail.com> wrote:
>> > > On Aug 22, 2:17 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
>> > > > In article
>> > > > <75752086-cfca-4e9f-99ed-63b650acc...@n19g2000yqc.googlegroups.com>,
>> > > > PeteOlcott <peteolc...@gmail.com> wrote:

[...]

>> > > > > Kripke has established that credentials are not required for
>> > > > > credibility.
>> >
>> > > > I never said anything about credentials.
>> >
>> > > No you were talking about my credibility.
>> >
>> > > The reason that I brought up Kripke in the first place was that he has
>> > > established the precedent that credentials are not required for
>> > > credibility. In other words it is possible to skip the credibility
>> > > step entirely and move directly to validity.
>> >
>> > You still seem not to understand what I'm saying, and I'm not sure how
>> > many more different ways I can say it. I'll try again:
>> >
>> > This isn't about credentials at all. Mentioning them yet again will
>> > only further demonstrate that you still don't understand.
>> >
>> > What I'm talking about is how a person presents himself and his ideas
>> > has a significant effect on his ability to have meaningful
>> > interactions with people who might be able to provide valuable
>> > contributions to his work (helping to make the work go easier and
>> > quicker, having more robust results, etc.).
>>
>> Only in the case of most people whom have not dismissed credibility as
>> a fake measure of truth.
>> The inventor of the steam boat was dismissed as a crank, "Fulton's
>> folly."
>
> I give up.

You've persevered long enough, that's for sure.

On the subject of Olcott's folly:
I'd describe his use of "whom" as a wilful abuse of English language.
It doesn't look like a consequence of ignorance, it gives me
the impression of being carefully calculated to offend, shock and awe.

pjk

[Brian M. Scott]

On Thu, 23 Aug 2012 16:12:54 +1200, pauljk
<paul....@xtra.co.nz> wrote in
<news:k14ak6$f7a$1...@dont-email.me> in
sci.lang,alt.usage.english,alt.english.usage:

> "Nathan Sanders" <san...@alum.mit.edu> wrote in message
> news:sanders-C8AEC1...@free.teranews.com...

>> In article
>> <d9e4d266-c07e-4a09...@wq14g2000pbb.googlegroups.com>,
>> PeteOlcott <peteo...@gmail.com> wrote:

>>> On Aug 22, 2:58 pm, Nathan Sanders
>>> <sand...@alum.mit.edu> wrote:

[...]

>>>> You still seem not to understand what I'm saying, and
>>>> I'm not sure how many more different ways I can say
>>>> it. I'll try again:

>>>> This isn't about credentials at all. Mentioning them
>>>> yet again will only further demonstrate that you still
>>>> don't understand.

>>>> What I'm talking about is how a person presents himself
>>>> and his ideas has a significant effect on his ability
>>>> to have meaningful interactions with people who might
>>>> be able to provide valuable contributions to his work
>>>> (helping to make the work go easier and quicker,
>>>> having more robust results, etc.).

>>> Only in the case of most people whom have not dismissed
>>> credibility as a fake measure of truth. The inventor of
>>> the steam boat was dismissed as a crank, "Fulton's
>>> folly."

>> I give up.

> You've persevered long enough, that's for sure.

> On the subject of Olcott's folly:

> I'd describe his use of "whom" as a wilful abuse of
> English language. It doesn't look like a consequence of
> ignorance, it gives me the impression of being carefully
> calculated to offend, shock and awe.

Oh, it definitely looks like ignorance to me. And this long
exchange clearly marks him as a crank irrespective of any
possible merits of his ideas.

Brian

[Franz Gnaedinger]

On Aug 22, 10:04 pm, PeteOlcott <peteolc...@gmail.com> wrote:
>
> Only in the case of most people whom have not dismissed credibility as
> a fake measure of truth.
> The inventor of the steam boat was dismissed as a crank, "Fulton's
> folly."

Instead of outlining your machine you are building
walls of words around a promise. You said you have
solved some cases. Why don't you speak of those
instead of always repeating your totalitarian promise?
Prof. Dr. Nathan Sanders and I have our hay not in
the same stack (as we say in Switzerland), so if we
give you the same advice you better head it: find
a realistic project, or reduce your machine to actual
problems it can solve, show us how it behaves in
a specific case you claim to be able to handle.

[pauljk]

"Brian M. Scott" <b.s...@csuohio.edu> wrote in message
news:1dm9ouja8csuv.p3t4mk37jv4h$.dlg@40tude.net...

Okay, I agree.
I was just trying to fit him up with a murder rather than
a lesser charge of accidental killing.

pjk

[Peter Olcott]

It is just too tedious of a detail to remember, and it is not
semantically significant.
All of the aspects of grammar that are not semantically significant are
too tedious to remember.

[Peter Olcott]

On 8/23/2012 2:25 AM, Franz Gnaedinger wrote:
> On Aug 22, 10:04 pm, PeteOlcott <peteolc...@gmail.com> wrote:
>> Only in the case of most people whom have not dismissed credibility as
>> a fake measure of truth.
>> The inventor of the steam boat was dismissed as a crank, "Fulton's
>> folly."
> Instead of outlining your machine you are building
> walls of words around a promise. You said you have
> solved some cases. Why don't you speak of those
> instead of always repeating your totalitarian promise?

I did several times and no one noticed.

> Prof. Dr. Nathan Sanders and I have our hay not in
> the same stack (as we say in Switzerland), so if we
> give you the same advice you better head it: find
> a realistic project, or reduce your machine to actual
> problems it can solve, show us how it behaves in
> a specific case you claim to be able to handle.

Designing a fully expressive meta language is likely already an
accomplished fact. No one noticed that either.

[António Marques]

Brian M. Scott wrote (23-08-2012 07:56):
> On Thu, 23 Aug 2012 16:12:54 +1200, pauljk <paul....@xtra.co.nz>
> wrote in <news:k14ak6$f7a$1...@dont-email.me> in
> sci.lang,alt.usage.english,alt.english.usage:
>

>> On the subject of Olcott's folly:
>>
>> I'd describe his use of "whom" as a wilful abuse of English language.
>> It doesn't look like a consequence of ignorance, it gives me the
>> impression of being carefully calculated to offend, shock and awe.
>
> Oh, it definitely looks like ignorance to me. And this long exchange
> clearly marks him as a crank irrespective of any possible merits of his
> ideas.

Sometimes one's formaulation can be spot on and one still refrain from
speaking it aloud (again, someone please CSC that).

[Guy Barry]

"António Marques" wrote in message news:k14tuo$44s$1...@dont-email.me...

> Sometimes one's formaulation can be spot on and one still refrain from
> speaking it aloud (again, someone please CSC that).

The CSC doesn't apply to that sentence, since there's no extraction
involved.

--
Guy Barry

[António Marques]

And what I suggested was that someone could fix that.

[PeteOlcott]

On Aug 23, 2:25 am, Franz Gnaedinger <f...@bluemail.ch> wrote:
> On Aug 22, 10:04 pm, PeteOlcott <peteolc...@gmail.com> wrote:
>
>
>
> > Only in the case of most people whom have not dismissed credibility as
> > a fake measure of truth.
> > The inventor of the steam boat was dismissed as a crank, "Fulton's
> > folly."
>
> Instead of outlining your machine you are building
> walls of words around a promise. You said you have
> solved some cases. Why don't you speak of those
> instead of always repeating your totalitarian promise?

http://www.lel.ed.ac.uk/~ronnie/publications.shtml
R Cann (2001) 'Functional versus lexical: a cognitive dichotomy'. In R
D Borsley (ed.) The Nature and Function of Syntactic Categories
Academic Press:37-78.

This paper outlines the almost impossible difficulty of dividing words
up into their functional (syntactic) and lexical (semantic) uses.

At least one of the things that makes this difficult is that some
words simultaneously fill both roles:
1) functional (syntactic)
2) lexical (semantic)

This is clearly the case of using a verb as the predicate to a meaning
postulate.

The way to solve this problem is to simply view syntax as pertaining
to aspects of connections between units of meaning, and semantics as
providing these units of meaning.

> Prof. Dr. Nathan Sanders and I have our hay not in
> the same stack (as we say in Switzerland), so if we
> give you the same advice you better head it: find
> a realistic project, or reduce your machine to actual
> problems it can solve, show us how it behaves in
> a specific case you claim to be able to handle.

I have already reduced the problem to a single manageable aspect,
focus on compositionality. I have said this several times and no one
noticed.

[PeteOlcott]

On Aug 22, 3:19 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article

> <d9e4d266-c07e-4a09-bed0-70b9e44d3...@wq14g2000pbb.googlegroups.com>,

> Swarthmore Collegehttp://sanders.phonologist.org/- Hide quoted text -
>
> - Show quoted text -

Challenging every aspect of the status quo is the only way that
substantial progress can be made in this field. This requires (to some
degree) and anti-conformist mindset.

Some of the fundamental concepts of linguistics are defined in ways
(such as the division of syntax from semantics) that add enormous
unnecessary complexity. Rather that conforming to these conventional
divisions, a new division must be formed.

[PeteOlcott]

On Aug 23, 1:56 am, "Brian M. Scott" <b.sc...@csuohio.edu> wrote:
> On Thu, 23 Aug 2012 16:12:54 +1200, pauljk

> <paul.kr...@xtra.co.nz> wrote in
> <news:k14ak6$f7a$1...@dont-email.me> in
> sci.lang,alt.usage.english,alt.english.usage:
>
> > "Nathan Sanders" <sand...@alum.mit.edu> wrote in message
> >news:sanders-C8AEC1...@free.teranews.com...
> >> In article
> >> <d9e4d266-c07e-4a09-bed0-70b9e44d3...@wq14g2000pbb.googlegroups.com>,

> Brian- Hide quoted text -

>
> - Show quoted text -

So credibility carries more weight than validity?
This is a very common error of reasoning.

[António Marques]

Short answer: Obviously. Credibility is an impression, validity requires
costly evaluation. Who will spend time evaluating something which didn't
make a favourable impression?

> This is a very common error of reasoning.

Long answer: Do you not see there is a hierarchy? Don't expect a n y o n e
at all to invest time in ascertaining validity unless there is some hint
that it will be worthwhile. We're all pressed for resources, we can't waste
them on every salesman. The only field where you can expect support based
only on your promises is politics: parties thrive mostly on what they
profess to desire rather than their actual behaviour. In science, you have
to have something to show, and until you have something you can only point
to the appearance of credibility.

[PeteOlcott]

> to the appearance of credibility.- Hide quoted text -

>
> - Show quoted text -

Establishing credibility and making substantial progress within the
field of linguistics may be mutually incompatible goals.

Whereas the latter requires actively challenging every aspect of the
status quo, (including the definition of every term) this may severely
negatively impact the former.

Do you have any idea how to best reconcile these competing goals?

[PeteOlcott]

On Aug 23, 6:20 am, António Marques <antonio...@sapo.pt> wrote:

> to the appearance of credibility.- Hide quoted text -

>
> - Show quoted text -

Brian said:
>> Oh, it definitely looks like ignorance to me. And this long
>> exchange clearly marks him as a crank irrespective of any
>> possible merits of his ideas.

That even my valid ideas should be discarded because I have not passed
the credibility test.
This is clearly an error of reasoning.

[António Marques]

PeteOlcott wrote (23-08-2012 13:02):
> Establishing credibility and making substantial progress within the field
> of linguistics may be mutually incompatible goals.
>
> Whereas the latter requires actively challenging every aspect of the
> status quo, (including the definition of every term) this may severely
> negatively impact the former.
>
> Do you have any idea how to best reconcile these competing goals?

As I think Nathan said (now, and possibly in the future, I don't presume to
explicate Nathan, but neither do I wish to be stealing someone's stance
without giving credit), you don't have to compromise on your goals or ideas,
but you could limit your _discussion_ of them. I think if you keep the
discussions narrowly focused and technical enough, you can build your
edifice without hassles and _then_ impress everyone with the results.

Cf.
'way to convert stack-based bytecode into register-based bytecode'
vs
'way to build an overnight dominant framework for mobile app development by
hijacking a programming environment which is widely used but inefficient and
has always failed to deliver successful end-user applications'

[PeteOlcott]

Way to reconcile the differing perspectives and lack of consensus
pertaining to compositionality within linguistics.

[Peter T. Daniels]

> This is clearly an error of reasoning.-

There's always the matter of figuring out _which_ of a multitude of
ideas are valid and which aren't.

(Back in 1979, at a conference celebrating the 50th anniversary of the
discovery of Ugarit, mention was made of the biblical commentaries of
Fr. Mitchell Dahood, a highly respected scholar, who made very
imaginative use of the Ugaritic literary texts in interpreting
difficult biblical passages -- his commentary on Psalms runs to three
volumes and is largely built on such comparisons and inferences.
Someone said, "If only 25% of his suggestions are valid ...," and the
ever-acerbic Prof. (now Rabbi) Stephen Kaufman interjected, "Ah, but
_which_ 25%?")

[PeteOlcott]

So the idea to sharply focus on a much more restricted domain should
(at least mostly) solve the problem.

[Peter T. Daniels]

> (at least mostly) solve the problem.-

Sounds like an excellent starting point.

[Glenn Knickerbocker]

On Thu, 23 Aug 2012 04:11:44 -0500, Peter Olcott wrote:
>It is just too tedious of a detail to remember, and it is not
>semantically significant.
>All of the aspects of grammar that are not semantically significant are
>too tedious to remember.

Your repeated insistence that the forms of connection in your own native
language are too difficult to learn and use makes it hard even for other
dilettantes like me to take your claims seriously about examining and
categorizing all the possible forms of connection in all possible human
languages. If an electrician wired an outlet for you with the neutral
and ground wires crossed and complained that it was just too tedious to
remember how they're different, would you pay attention to anything he
had to say about devising a new, more efficient way to transfer power
over electrical wires?

¬R / "We know the difference between good and bad corn early in life and
/ have the confidence that comes from such discernment." Chris Squire
http://users.bestweb.net/~notr/ (of Ontario's *London Free Press*)

[Peter Olcott]

On 8/23/2012 3:15 PM, Glenn Knickerbocker wrote:
> On Thu, 23 Aug 2012 04:11:44 -0500, Peter Olcott wrote:
>> It is just too tedious of a detail to remember, and it is not
>> semantically significant.
>> All of the aspects of grammar that are not semantically significant are
>> too tedious to remember.
> Your repeated insistence that the forms of connection in your own native
> language are too difficult to learn and use makes it hard even for other
> dilettantes like me to take your claims seriously about examining and
> categorizing all the possible forms of connection in all possible human

I make it a general rule to never ever memorize anything. If an idea can
not be expressed as connected reasoning, then it is generally not worth
knowing. This aspect of grammar** forms one of those instances.

** Grammar that is not semantically significant.

[pauljk]

"Peter Olcott" <OCR4Screen> wrote in message
news:B46dnWp-GYlNbqjN...@giganews.com...

That is not true, "who" and "whom" do not have the same meaning.

> All of the aspects of grammar that are not semantically significant are too tedious
> to remember.

So why do you bother to remember spelling of thousands
of other words like "tedious" when "teedeus" would be easier
to remember AND it wouldn't clash with another word?

pjk

[Peter Olcott]

The difference in the meanings only looks syntactical to me. I could be
wrong.

>
>> All of the aspects of grammar that are not semantically significant
>> are too tedious to remember.
>
> So why do you bother to remember spelling of thousands
> of other words like "tedious" when "teedeus" would be easier
> to remember AND it wouldn't clash with another word?
>
> pjk
>
>

I use these words enough that I naturally remember without any effort,
the spell checker mostly finds typos.

[Nathan Sanders]

In article <k16p6v$tkh$1...@dont-email.me>,

"pauljk" <paul....@xtra.co.nz> wrote:

> That is not true, "who" and "whom" do not have the same meaning.

There are even minimal pairs:

(1) I don't know who I want to visit first.
("who" is the semantic subject of "visit";
they will be coming to visit me)

(2) I don't know whom I want to visit first.
("who" is the semantic object of "visit";
I will be going to visit them)

Of course, this distinction is nearly completely lost for most
speakers in ordinary casual conversation (except sporadically,
especially as a hypercorrection).

[Peter Olcott]

On 8/23/2012 10:18 PM, Nathan Sanders wrote:
> In article <k16p6v$tkh$1...@dont-email.me>,
> "pauljk" <paul....@xtra.co.nz> wrote:
>
>> That is not true, "who" and "whom" do not have the same meaning.
> There are even minimal pairs:
>
> (1) I don't know who I want to visit first.
> ("who" is the semantic subject of "visit";
> they will be coming to visit me)
>
> (2) I don't know whom I want to visit first.
> ("who" is the semantic object of "visit";
> I will be going to visit them)
>
> Of course, this distinction is nearly completely lost for most
> speakers in ordinary casual conversation (except sporadically,
> especially as a hypercorrection).
>
> Nathan
>

I accept that I stand corrected.

[Glenn Knickerbocker]

On Thu, 23 Aug 2012 21:54:18 -0500, Peter Olcott wrote:
>On 8/23/2012 9:34 PM, pauljk wrote:
>> That is not true, "who" and "whom" do not have the same meaning.
>The difference in the meanings only looks syntactical to me.

And yet you list case in your inventory of types of connection. Your
argument makes it seem as though you don't understand your own list.

ŹR http://users.bestweb.net/~notr/arkville.html
God damn it, now I have a .sig, Andrew Pearson.

[Guy Barry]

"Peter Olcott" wrote in message
news:cfGdnWkeZo-GbqvN...@giganews.com...

> On 8/23/2012 10:18 PM, Nathan Sanders wrote:

> > There are even minimal pairs:
>
> > (1) I don't know who I want to visit first.
> > ("who" is the semantic subject of "visit";
> > they will be coming to visit me)
>
> > (2) I don't know whom I want to visit first.
> > ("who" is the semantic object of "visit";
> > I will be going to visit them)

Are you sure about this? Consider the corresponding non-relativized
sentences:

(1a) I want him to visit first.
(2a) I want to visit him first.

In both of these, "him" is in the object case (even though it is a semantic
subject in (1a)). Hence I would expect to use an object relative pronoun in
both (1) and (2). A speaker who used "whom" as the object relative would
presumably use "whom" in both, and a speaker (like me) who used "who" would
use "who" in both. The case-marking on the relative reflects syntactic
function, not semantic role.

--
Guy Barry

[Nathan Sanders]

In article <ThEZr.2291$mc1...@fx12.am4>,

"Guy Barry" <guy....@blueyonder.co.uk> wrote:

> "Peter Olcott" wrote in message
> news:cfGdnWkeZo-GbqvN...@giganews.com...
>
> > On 8/23/2012 10:18 PM, Nathan Sanders wrote:
>
> > > There are even minimal pairs:
> >
> > > (1) I don't know who I want to visit first.
> > > ("who" is the semantic subject of "visit";
> > > they will be coming to visit me)
> >
> > > (2) I don't know whom I want to visit first.
> > > ("who" is the semantic object of "visit";
> > > I will be going to visit them)
>
> Are you sure about this?

I only have "who" in both cases, so no, I'm not sure. :-)

> Consider the corresponding non-relativized sentences:
>
> (1a) I want him to visit first.
> (2a) I want to visit him first.

Good point, but is the choice of "who/whom" always determined by its
base (logical/deep-structure) position (marked by __)?

(1b) Who(m?) do you want __ to visit?
(2b) Whom do you want to visit __?

(1c) That's who(m?) I want __ to visit.
(2c) That's who(m?) I want to visit __.

(1d) Who(m?) I want __ to visit is none of your business.
(2d) Who(m?) I want to visit __ is none of your business.

I don't really know the answer; I'm going well beyond the bounds of my
native speaker intuition now. All of these are "who" for me.

[Franz Gnaedinger]

On Aug 23, 11:13 am, Peter Olcott <OCR4Screen> wrote:
>
> Designing a fully expressive meta language is likely already an
> accomplished fact. No one noticed that either.

One more totalism, fully expressive meta-language.
You never speak of actual language, you never give us
an example of language and how it is being analyzed
by your machine, all we get are ever more bricks of
your wall, plenty of termini technici and ontology, your
favorite term. You believe that lists of definitions given
in lexica cover the full range of meaning and function,
not understanding that such definitions are circular
and approximative, ultimately requiring a human mind
and gathered experience to understand. You can't
separate all of language from life. I wonder how your
ontological definition of love as a semantic and
functional element would look like. You can't define
love, but you understand what the word means if you
love and are loved and have been loved. The part
of experience can't be fully rendered by machines.

[Paul {Hamilton Rooney}]

On 24-Aug-12 3:04 PM, Franz Gnaedinger wrote:
The part
> of experience can't be fully rendered by machines.

Or by words.

It used to be - maybe still is - fashionable to think that thoughts
relied on words, but it really doesn't take a genius to show that that
is false.

More so for machines.

[pauljk]

"Peter Olcott" <OCR4Screen> wrote in message

news:cfGdnWkeZo-GbqvN...@giganews.com...

Good on you.

In this and similar cases, I would expect people for whom
there is no semantic distinction between "who" and "whom"
would keep using the simple "who" exclusively.

That is why I found your use of "whom" puzzling because
I expected it to have distinct meaning and I couldn't figure
what it was.

pjk

[Franz Gnaedinger]

On Aug 24, 9:12 am, Paul {Hamilton Rooney} <paulvloo...@snotmail.com>
wrote:

Poems and songs can say a lot about love that will be understood
by humans - if the words resonate with experience. A good poet
and singer can make them resonate in ourselves. The full range
of meanings of words goes beyond machine logic. Peter Olcott
has given just one single sample of language, in his very first
message to sci.lang:

Treat your neighbor as your friend.

I read this as an equivalent of the Biblical commandment

Thou shalt love thy neighbor as thyself.

Peter forbids this reading, and gave the link to a Wikipedia
page on Indian philosophy. How can he command how
we must understand language? He did not explain how he
got his sentence, whether it is a quote, or whether he run
a paper simulation of his machine on the Wikipedia page.
If the latter is the case, his machine makes the philosophy
collapse instead of reducing a maxim to its most compact
form. Apart from that single sample we get no example
of language from Peter Olcott, always and only big words
of ontological dimensions.

[Peter T. Daniels]

On Aug 24, 2:08 am, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <ThEZr.2291$mc1....@fx12.am4>,

The case of who/whom is determined by its function in the relative
clause.

He visits you --> Who do you want to visit?

You visit him --> Whom do you want to visit?

(But sentence-initial "whom" is quite unusual.)

[Peter Olcott]

On 8/23/2012 11:50 PM, Glenn Knickerbocker wrote:
> On Thu, 23 Aug 2012 21:54:18 -0500, Peter Olcott wrote:
>> On 8/23/2012 9:34 PM, pauljk wrote:
>>> That is not true, "who" and "whom" do not have the same meaning.
>> The difference in the meanings only looks syntactical to me.
> And yet you list case in your inventory of types of connection. Your
> argument makes it seem as though you don't understand your own list.

I was not aware of any semantic distinction between who and whom.

[António Marques]

Peter Olcott wrote (24-08-2012 15:04):
> On 8/23/2012 11:50 PM, Glenn Knickerbocker wrote:
>> On Thu, 23 Aug 2012 21:54:18 -0500, Peter Olcott wrote:
>>> On 8/23/2012 9:34 PM, pauljk wrote:
>>>> That is not true, "who" and "whom" do not have the same meaning.
>>> The difference in the meanings only looks syntactical to me.
>> And yet you list case in your inventory of types of connection. Your
>> argument makes it seem as though you don't understand your own list.
> I was not aware of any semantic distinction between who and whom.

Then 'whom' isn't part of your language. WHY do you ever use it?

[Peter Olcott]

On 8/24/2012 2:04 AM, Franz Gnaedinger wrote:
> On Aug 23, 11:13 am, Peter Olcott <OCR4Screen> wrote:
>> Designing a fully expressive meta language is likely already an
>> accomplished fact. No one noticed that either.
> One more totalism, fully expressive meta-language.
> You never speak of actual language, you never give us

You have not read all of my posts or you would realize that I have
mentioned CycL.

> an example of language and how it is being analyzed
> by your machine, all we get are ever more bricks of

The proper way to design a complex system is to reverse-engineer the
details on the basis of the desired final end-result. Starting with a
toy system and trying to generalize it later has proven to provide an
insufficient basis for comprehending the details of natural language
processing.

> your wall, plenty of termini technici and ontology, your
> favorite term. You believe that lists of definitions given
> in lexica cover the full range of meaning and function,
> not understanding that such definitions are circular
> and approximative, ultimately requiring a human mind

The would be true of dictionary definitions for humans. Because most
words are defined in terms of other words and this recursive
specification can run enormously deep an extremely complex ontology must
be specified to provide comprehension by machines.

> and gathered experience to understand. You can't
> separate all of language from life. I wonder how your
> ontological definition of love as a semantic and
> functional element would look like. You can't define
> love, but you understand what the word means if you
> love and are loved and have been loved. The part
> of experience can't be fully rendered by machines.

I am limiting the current research to the set of concepts that can be
expressed within words. It seems that the only limit to this is that
direct physical sensations can not be fully expressed within words. One
can not completely describe the taste of strawberries such that this
actual taste can be communicated from mind to mind.

[Peter Olcott]

Even though I was apparently incorrect, it still seems to be a nit picky
detail.

[Jerry Friedman]

On Aug 23, 9:18 pm, Nathan Sanders <sand...@alum.mit.edu> wrote:
> In article <k16p6v$tk...@dont-email.me>,

>  "pauljk" <paul.kr...@xtra.co.nz> wrote:
> > That is not true, "who" and "whom" do not have the same meaning.
>
> There are even minimal pairs:
>
>      (1) I don't know who I want to visit first.
>           ("who" is the semantic subject of "visit";
>           they will be coming to visit me)
>
>      (2) I don't know whom I want to visit first.
>           ("who" is the semantic object of "visit";
>           I will be going to visit them)
>
> Of course, this distinction is nearly completely lost for most
> speakers in ordinary casual conversation (except sporadically,
> especially as a hypercorrection).

I don't buy this (not that I'm a "whom" user for most purposes). The
sentences corresponding to the dependent clauses are

(1a) I want them to visit first.

(2a) I want to visit them first.

As both have "them", I believe the traditional rules call for "whom"
in both.

--
Jerry Friedman

[António Marques]

It's not nitpicking, it's puzzlement.