[NGL-project] Proposal: Adverbs of means and manner
Jack Durst
Date: Fri, 17 Mar 2000 23:16:24 -0800 (PST)
From: Gerald Koenig <j...@NETCOM.COM>
Subject: Re: [NGL-project] Proposal: Adverbs of means and manner
From: Gerald Koenig <j...@NETCOM.COM>
>Date: Thu, 16 Mar 2000 02:39:03 -0800
>From: Jack Durst <sp...@sierra.net>
>Subject: [NGL-project] Proposal: Adverbs of means and manner
>
>From: Jack Durst <sp...@sierra.net>
>
>I use the proposal tag with some uncertainty here, as I don't have a
>proposal here so much as an idea that's been rolling around my head for a
>while that I'm uncertain of the implementation for.
Hi, Jack, and all,
I'm still tripping to the hospital every day and nite to help my
brotherinlaw so haven't been able to think. But here are a few short
comments.
>
>CONCEPT:
>There are a number of different classes into which an adverb can fall,
>this is generally unclear from the meaning alone. To name only three,
>adverbs can be adverbs because they're forced to be by a rule of grammar
>(eg. {q senat Kanadai' inlìnuzig} "the unbelieving canadian senate" where
>what would normally be an adjective is derived into an adverb to describe
>a pseudo-genitive),
This corresponds to the VXT participle with adjective role:
inl`inuz-iem (unbelieve-p-participle-adj) or unbelieving-adj.
The -i in -iem specifically identifies the word as an adjective.
>they can be adverbs because they describe the manner
>in which something was done ({ta paem vonig} "he/she walked quickly"), or
>they can be adverbs which describe the means by which something was done.
>
>In the equivilants for "with" we already find this distinction between
>means and manner (contrast {Paem (col/ezo) òl vurem}, one indicates that
>he had the stick with him as he walked (manner), the other that he used
>the stick in order to walk (means)), but in the equivilant adverbial form
>{Paem vurig}, though it may imply one case or the other lexically, is
>ambiguous grammatically, and though the paraphrase works well for
>sentences, it cannot be applied to adjectives, conjunctions and plain
>verbs as easily.
A possible solution would be the use of the POS tagged participles in
VXT. For your means or instrumental case of stick, ("in order to"), we
can use an adverbial tag which insures that the stick is modifying or
restricting the action of the verb, <walk>. "He/she walked
sticked-adv."
1.Ad paem vonoad.
>From the writekit: ftp://ftp.netcom.com/pub/jl/jlk/writekit.nov
------------------------
3 ADVERB o o+ad=oad
She lived <loved> by all her family and friends.
za pa je? <filisoad> a? zupo zas ubos et hixes.
<Loved> is an adverb of manner modifying the verb live, answering the
question, how?
This example shows that the tense of the participial is not tied to
the (past indicative) tense of the main verb. The love goes on without
her being alive. Infinitive tense is independent of the tense of the
main clause verb.
----------------------------------------------
This English example is taken from a grammar book so I assume the POS
analysis is correct. (I have not used past indicative, which would be
pa paem vonoad.)
I have to differ with you in that the second case would not be an
adverb of manner according to the definition I have:
"An adverb of manner indicates manner of action or state...and thus the
answer to the question, ...by what method."
The method of walking here would be by the help of a stick. Just
carrying a stick is not a method.
However whatever these two cases are named, I understand your point that
they may need to be distinguished.
For the "manner" (your usage) case the stick is an object, a noun,
that s/he had with her/him during the walking.
2. Ad paem vonuad.
"He/she walked sticked-noun."
While this doesn't rule out that the stick was used in walking, it
doesn't directly assert it either, as the previous adverbial form
does. Given that the adverbial form is available, I think the use of
the noun suffix form might be enough to make it clear the stick is
merely accompaning the walker.
I think this question deserves a lot more discussion before a proposal
stage is reached. There is so much more to adverbs than I have touched
on here, as you've said. This is just one possible use for the role
tagged participles I have defined. I'd like to respond to your other
proposal too but I am in Hospital Hell right now.
Vus,
Jerry.
>CONCLUSION
>So, I'd like y'all to take a look at the possibilities and report back
>which you prefer... Thanks.
>
>Sincerely,
> Jack Durst
>Sp...@sierra.net
>[this posting written in Net English]
------------------------------------------------------------------------
You have a voice mail message waiting for you at iHello.com:
http://click.egroups.com/1/2377/1/_/_/_/953363786/
------------------------------------------------------------------------
Jack Durst
Date: Sat, 18 Mar 2000 19:13:11 -0330
From: Stephen DeGrace <c72...@mun.ca>
Subject: Re: [NGL-project] Proposal: Adverbs of means and manner
From: Stephen DeGrace <c72...@mun.ca>
On Thu, 16 Mar 2000 02:39:03 -0800 (PST), Jack Durst
<sp...@sierra.net> wrote:
>POSSIBLE PROPOSALS:
>1. Do nothing. The language has gotten along fine without this
>distinction in adverbs so far, and the extra morphemes aren't worth it.
>Deal with it in the rough cases by idioms (like the {laus/laus ke}
>distinction) or derivations relying on lexical biases...
Hmm. Well, I *think* we can probably get along fine without it.
However, that's not to say that I oppose adding something. If a
proposal for some explicit system were made, I would certainly look at
it very seriously. It's just that I don't _seem_ to feel a pressing
need.
As well, as Jerry said, a solution may exist in greater employment of
VXT's derivation system...
Stephen
------------------------------------------------------------------------
Group buying power teams you with others to save more. Get the
lowest prices on all the latest Palms and accessories when you
buy through Accompany, the Web's leading destination for great
deals on group buys.
http://click.egroups.com/1/2512/1/_/_/_/953419278/
------------------------------------------------------------------------
Jack Durst
Date: Sun, 19 Mar 2000 22:10:56 -0800 (PST)
From: Gerald Koenig <j...@NETCOM.COM>
From: Gerald Koenig <j...@NETCOM.COM>
>Date: Sat, 18 Mar 2000 20:33:24 -0800 (PST)
>From: Jack Durst "<sp...@sierra.net>"
>Subject: Re: [NGL-project] Proposal: Adverbs of means and manner
Hi, Jack and all,
Things are a lot better at the hospital today, my brotherinlaw
got his kidney function back and 3 antibiotics have been found which
together work against his septicemia.
>> >CONCEPT:
>> >(eg. {q senat Kanadai' inlěnuzig} "the unbelieving canadian senate" where
>> This corresponds to the VXT participle with adjective role:
>> inl`inuz-iem (unbelieve-p-participle-adj) or unbelieving-adj.
>> The -i in -iem specifically identifies the word as an adjective.
>Actually, grammatically it couldn't without a shift or meaning... If it's
>a plain adjective (of any type) then it can only describe a noun, making
>{Kanadai' & inlěnuziem} both describe the noun. To describe the
>pseudo-genitive (my intent being "the senate, composed of unbelieving
>Canadians..." NOT "the unbelieving senate of Canada" as I translated it at
>2AM) it would still have to be an adverb.
Well I was waffling on the correct tag for the participle because
I did see "inlinuzig" as an adverb. Anyway, it can be expressed in
VXT as as a present participle which has an adverbial role with the
-oem suffix: <inl`inuzoem>. I must say though I'm still not sure why
"unbelieving" in "unbelieving Canadians" isn't an adjective. Maybe it's
the elided "composed of"; Canadians of a kind that unbelievably do
exist. I was quite surprised at the outcome of the California
initiatives.
>
>> A possible solution would be the use of the POS tagged participles in
>> VXT. For your means or instrumental case of stick, ("in order to"), we
>> can use an adverbial tag which insures that the stick is modifying or
>> restricting the action of the verb, <walk>. "He/she walked
>> sticked-adv."
>Yes, this would work, but linguisticly I have reservations about allowing
>cases to be derivational. I've never studied a natlang in which this was
>true, nor have I ever read any theoretical discussion which would permit
>it, so I have trouble on universality grounds for this.
I probably shouldn't have used the word case here. I was thinking only
of the varieties of roles I have identified for the VXT participle and
infinitive forms. I haven't thought through how they can or could be
used for grammatical case, this is the first example to come up. These
roles are all derived from common english usage, so I don't see any
universality problem. These endings are just tags that classify common
english uses of the participle and gerund. By so doing the grammar of
the sentence becomes more parseable and the semantics more definite.
>> 3 ADVERB o o+ad=oad
>Is VXT {-ad} accepted? If not I'd like to offer a late objection to the
>form on the grounds that final {-ad} is already a case marker in wide
>use and allowing it would be unacceptably ambiguous. Using it to mark
>adverbs feels perverse as the other morpheme quite explicitly marks nouns.
>(It's in wider use, and is also older...)
Well I really worked hard when I designed these endings to not impact
your existing system, though I've forgotten now exactly how I did it. I
hope this is not an exception. I don't think it is because <-ad> never
appears alone in these suffixes. I just wrote it that way to show its
derivation from pAst Dextro, english verb-ed. It must be -uad, (noun),
-iad (adjective) or -oad (adverb). Also, if the last letter of the verb
stem is a vowel, an <s> must be inserted. The stem must be a verb. The
stem for TVS <-ad>, dative case, must be a noun. It seems to me that
<-oad> as a participle verb suffix is quite distinct from <-ad> as a
dative noun suffix. I hope the esthetic will grow acceptable to you.
I never have proposed any of VXT formally, because I want to keep it
fluid to work out possible problems such as this one. Eventually I will
propose it for NGL if it comes into use and is fairly debugged.
I created it specifically for NGL.
>term. The distinction I was aiming at was instrumental/attributive
>(though manner is a fine example of an additional class...)
>
>Take the ambiguous English sentence "I wanna play guitar like Hendrix."
>It could mean:
>Yeva kaodom q gitar laus Hendriksa. (Same style, maybe genre)
>(attributive)
>Yeva kaodom q gitar laus ke Hendriksa. (Left handed, with thumb crossing)
>(instrumental/means)
>Yeva kaodom q gitar laus leh ke Hendriksa. (of the same quality)
>(qualitative)
Heh, did you know I am a Hendrix fan? :). All except his ending.
Somewhere I have B. Rempt's quite exhaustive list of cases, I translated
the spatial ones to VXT. It's a perennial conlang topic for a reason, we
should deal with it. One thing we could do for the lang as an
exercise is to translate a list of cases, creating new grammar as
needed. Anything that survives a test drive by Stephen will be quite
sturdy :). Just kidding, Stephen, your tryouts are the best.
>Understood, and thanks for your response, hixo.
Vus hixci je inxer umi,
Jerry
>
>Sincerely,
> Jack Durst
>Sp...@sierra.net
>[this posting written in Net English]
>
------------------------------------------------------------------------
Group buying power teams you with others to save more. Get the
lowest prices on all the latest Palms and accessories when you
buy through Accompany, the Web's leading destination for great
deals on group buys.
http://click.egroups.com/1/2512/1/_/_/_/953532927/
------------------------------------------------------------------------
Jack Durst
> Hi, Jack and all,
> Things are a lot better at the hospital today, my brotherinlaw
> got his kidney function back and 3 antibiotics have been found which
> together work against his septicemia.
Good to hear.
> >> >CONCEPT:
> >> >(eg. {q senat Kanadai' inlěnuzig} "the unbelieving canadian senate" where
> >> This corresponds to the VXT participle with adjective role:
> >> inl`inuz-iem (unbelieve-p-participle-adj) or unbelieving-adj.
> >> The -i in -iem specifically identifies the word as an adjective.
> >Actually, grammatically it couldn't without a shift or meaning... If it's
> >a plain adjective (of any type) then it can only describe a noun, making
> >{Kanadai' & inlěnuziem} both describe the noun. To describe the
> >pseudo-genitive (my intent being "the senate, composed of unbelieving
> >Canadians..." NOT "the unbelieving senate of Canada" as I translated it at
> >2AM) it would still have to be an adverb.
> Well I was waffling on the correct tag for the participle because
> I did see "inlinuzig" as an adverb. Anyway, it can be expressed in
> VXT as as a present participle which has an adverbial role with the
> -oem suffix: <inl`inuzoem>. I must say though I'm still not sure why
> "unbelieving" in "unbelieving Canadians" isn't an adjective. Maybe it's
This is interesting... I guess it's the stronger typing in my dialect
which makes this feel odd. In Tokcir, where POS typing is strong due to
the widespread use of derivation and the highly varriable word order, I'd
worry about ambiguity. In Nilenga, with its fairly well fixed word order
and weaker typing by POS, however, it feels perfectly at home.
It'll get sorted out in the merge, I suppose.
> the elided "composed of"; Canadians of a kind that unbelievably do
> exist. I was quite surprised at the outcome of the California
> initiatives.
As was I, our neighbors to the west are a bit off...
> >> A possible solution would be the use of the POS tagged participles in
> >> VXT. For your means or instrumental case of stick, ("in order to"), we
> >> can use an adverbial tag which insures that the stick is modifying or
> >> restricting the action of the verb, <walk>. "He/she walked
> >> sticked-adv."
> >Yes, this would work, but linguisticly I have reservations about allowing
> >cases to be derivational. I've never studied a natlang in which this was
> >true, nor have I ever read any theoretical discussion which would permit
> >it, so I have trouble on universality grounds for this.
> I probably shouldn't have used the word case here. I was thinking only
> of the varieties of roles I have identified for the VXT participle and
> infinitive forms. I haven't thought through how they can or could be
> used for grammatical case, this is the first example to come up. These
The role structure of verbs is similar linguistically to the case
structure of nouns. There are languages which let the role structure be
derivational, but it's usually a lexical, not a grammatical distinction in
such languages. The only latin examples I can think of are modern
coinings...
> roles are all derived from common english usage, so I don't see any
> universality problem. These endings are just tags that classify common
> english uses of the participle and gerund. By so doing the grammar of
> the sentence becomes more parseable and the semantics more definite.
I have no trouble with the conception of verb roles, most languages have
them in some form or another and they're quite generally usefull. The
question is whether they ought to be grammatical or lexical/derivational.
> >> 3 ADVERB o o+ad=oad
> >Is VXT {-ad} accepted? If not I'd like to offer a late objection to the
> >form on the grounds that final {-ad} is already a case marker in wide
> >use and allowing it would be unacceptably ambiguous. Using it to mark
> >adverbs feels perverse as the other morpheme quite explicitly marks nouns.
> >(It's in wider use, and is also older...)
> Well I really worked hard when I designed these endings to not impact
> your existing system, though I've forgotten now exactly how I did it. I
I understand and I appreciate your effort.
> hope this is not an exception. I don't think it is because <-ad> never
> appears alone in these suffixes. I just wrote it that way to show its
> derivation from pAst Dextro, english verb-ed. It must be -uad, (noun),
> -iad (adjective) or -oad (adverb). Also, if the last letter of the verb
> stem is a vowel, an <s> must be inserted. The stem must be a verb. The
> stem for TVS <-ad>, dative case, must be a noun. It seems to me that
> <-oad> as a participle verb suffix is quite distinct from <-ad> as a
> dative noun suffix. I hope the esthetic will grow acceptable to you.
Seeing your explanation, I am less concerned by the polysemy. Given the
fairly strong typing of nouns and verbs in the language, I can see an
arguement to allow it as reasonable.
However, I do have a good compromise proposal, which would remove all
polysemy. Change {-ad} to {-'ad} (with the ' being elideable except in
case of double-vouel violations.) This removes the need for a special
rule in double-vouel situations (as ' is never elideable in that case
anyway...) while at the same time always making the two forms
distinguishable and allowing the forms to coexist peacefully with minimal
change (any incompatibility would appear to be poor spelling...)
> I never have proposed any of VXT formally, because I want to keep it
> fluid to work out possible problems such as this one. Eventually I will
> propose it for NGL if it comes into use and is fairly debugged.
> I created it specifically for NGL.
Cool. Frankly, I think the best way to debug the system would be a
proposal debate, but it's your perogative as to when you propose.
> >term. The distinction I was aiming at was instrumental/attributive
> >(though manner is a fine example of an additional class...)
> >
> >Take the ambiguous English sentence "I wanna play guitar like Hendrix."
> >It could mean:
> Heh, did you know I am a Hendrix fan? :). All except his ending.
Me too...
> Somewhere I have B. Rempt's quite exhaustive list of cases, I translated
> the spatial ones to VXT. It's a perennial conlang topic for a reason, we
> should deal with it. One thing we could do for the lang as an
> exercise is to translate a list of cases, creating new grammar as
Cases can be extremely usefull, but a language must balance size with
utility. A huge case system is more difficult to use even for native
speakers. Vectors proper are versatile enough to handle most of the
spacial cases without modification.
Jack Durst
Date: Tue, 21 Mar 2000 00:27:42 -0800 (PST)
From: Gerald Koenig <j...@NETCOM.COM>
Subject: Re: [NGL-project] Proposal: Adverbs of means and manner
From: Gerald Koenig <j...@NETCOM.COM>
>Subject: Re: [NGL-project] Proposal: Adverbs of means and manner
>From: Jack Durst <sp...@sierra.net>
>On Sun, 19 Mar 2000, Gerald Koenig wrote:
>> hope this is not an exception. I don't think it is because <-ad> never
>> appears alone in these suffixes. I just wrote it that way to show its
>> derivation from pAst Dextro, english verb-ed. It must be -uad, (noun),
>> -iad (adjective) or -oad (adverb). Also, if the last letter of the verb
>> stem is a vowel, an <s> must be inserted. The stem must be a verb. The
>> stem for TVS <-ad>, dative case, must be a noun. It seems to me that
>> <-oad> as a participle verb suffix is quite distinct from <-ad> as a
>> dative noun suffix. I hope the esthetic will grow acceptable to you.
>
>Seeing your explanation, I am less concerned by the polysemy. Given the
>fairly strong typing of nouns and verbs in the language, I can see an
>arguement to allow it as reasonable.
>
>However, I do have a good compromise proposal, which would remove all
>polysemy. Change {-ad} to {-'ad} (with the ' being elideable except in
>case of double-vouel violations.) This removes the need for a special
>rule in double-vouel situations (as ' is never elideable in that case
>anyway...) while at the same time always making the two forms
>distinguishable and allowing the forms to coexist peacefully with minimal
>change (any incompatibility would appear to be poor spelling...)
>
I'm not sure I understand what you have in mind, so I give examples
to verify:
tok::- v, speak, write.
vaib::- unit, word
saho::- discover, look for, find
Current VXT grammar:
ku tokiad vaib
the spoken word. "spoken" is an adjectival participle.
[cf. <tokad>, TVS dative object;
and <tokiad>, VXT adjectival participle.]
Proposed grammar?
ku toki'ad vaib? Is this an elidable ' ?
Current:
ku sahosiad jiuz.
the sought answer.
Proposed?
ku sahoi'ad jiuz?
ku saho'i'ad jiuz? Is this the elidable ' ?
Are you suggesting substituting " ' " for "s" in all forms?
------------------------------------
From the writekit for reference:
PARTICIPLES, GERUNDS, VERBALS.
GENERAL FORM OF THE PARTICIPLE:
<STEM->(s){u|i|o}{em|ad|ek}
(| means or.)
(s) is interposed between 1st and 2nd triplicate vowels: vsvv.
Three vowels in a row are not permissible.
SUFFIX ELEMENTS FOR COMBINING AS V1+VC: vowel1, vowel-consonant.
V1 may be:
u::- declares a noun function of the participle. AKA a gerund.
i::- declares an adjective function of the participle.
o::- declares an adverbial function of the participle.
VC may be:
em::- declares a Present Participle, -ing, -ando,-iendo..[add your language]
ad::- declares a Past Participle, -ed, en, ..-ado..-ido,.[add your language]
ek::- declares a Perfect Participle, having <verb>-ed...
------------------------------------------------
>> I never have proposed any of VXT formally, because I want to keep it
>> fluid to work out possible problems such as this one. Eventually I will
>> propose it for NGL if it comes into use and is fairly debugged.
>> I created it specifically for NGL.
>Cool. Frankly, I think the best way to debug the system would be a
>proposal debate, but it's your perogative as to when you propose.
Perhaps it is my experience with endless debates over flaws frozen into
Loglangs that make me so reluctant to accept top-down design, even my
own, without thorough testing at each step. The implications of every
design decision are so far reaching they can hardly be known without a
period of use. This is not to discount in any way having a grand
design; that is what makes a grand language. Ultimately it will be the
quality of our lang that will determine its usefulness and acceptance.
And I take your point that there are other standards besides English.
>> Somewhere I have B. Rempt's quite exhaustive list of cases, I translated
>> the spatial ones to VXT. It's a perennial conlang topic for a reason, we
>> should deal with it. One thing we could do for the lang as an
>> exercise is to translate a list of cases,
>Cases can be extremely usefull, but a language must balance size with
>utility. A huge case system is more difficult to use even for native
>speakers. Vectors proper are versatile enough to handle most of the
>spacial cases without modification.
>
I fully agree with you that a huge, complex case system is a problem
and probably bound to disintegrate, as Latin did. I haven't given any
thought to non-spatial cases though and it may be ok to take a look at
them. I must modestly :) agree with you that vectors will do the job
for spatial cases. At least I have not yet found any that cannot be
expressed well with vector and I try all the examples I see.
Sahosoem jiuxes, je
Jerry
------------------------------------------------------------------------
FreeLotto wants to give you a brand new Mazda Miata. Plus
tonight and every night you can win $1,000,000. Sign up
today to win! Drawing will be held on March 22, 2000. play for FREE!
http://click.egroups.com/1/2301/1/_/_/_/953627544/
------------------------------------------------------------------------
Jack Durst
is to change {-ad} in your system to {-'ad} with the ' being elidable in
all cases except when preceeded by two vouels, and to simultaniously do
away with (or make optional) the s in double-vouwl situations.