Dans un courrier daté du 21/06/00 21:56:28  , Jim a écrit :

(I don't reject all semantic analysis.   I merely think that adverbs and
adjectives are easiest to distinguish by syntactic criteria.)
that's for sure.
from your definitions of "process", "action", etc.
i think that i ignore the proper english vocab needed
to discuss this however leniently you rate my english :-)

so let me start from scratch with very basic french
words that will certainly conjure up in your mind
their english definitions i ignore myself.

i suggest that your post was about syntactic (?)
adv and adj while my post tried to discuss
what linguists call here "comportement" ("behaviour"?) as
opposed to "entité" ("entity"?).
an "énoncé" ("phrase"?) is made by pairing an "entité"
with a "comportement".

the comportement might be "active" ("process"?) :
"the sun shines"

or "attributif":
"the sun is shiny"

the "énoncé" is also expressed in various "modes"
(some claim there are 6 of them)
like "présentatif" (ex: "there is a misunderstanding").

as you can see, whether there actually and physically is
a process ongoing is not really important from this theory.
and the "behaviour" may be expressed at any level of "énoncé":
sentence, subclause, phrase, word order, word, affix, flexion.
so whether you consider noun phrase, verb or noun or adj or adv
does only change your starting point of the analysis, not the outcome.
i mean: analysing a noun that way will readily start from "entity",
a verb from "behaviour", and an adjective-- well-- depends on whether
you're a romance or an english speaker :-)

i agree that it's all semantics and you didn't write about adj
and adv in this respect, but i'm interested in it for conlanging
purpose and also because i've read a thread on adjective being
either verb or noun, which i think roots in this question.

any concept expressed with language (it's called here a "lexie"-
of course it's not necessarily a word)
can be taken as either an "entity" or a "behaviour".
when a noun (which is always itself a lexie) is a paired
as a "behaviour"of/ with an "entity", then this noun is
taken as an "actor" of a range of process(es);
but of which "process" exactly is determined by the lexie
it builds with the other "entity".

the fact that a weapon has many different uses doesn't mean
that "deadly" doesn't refer to one of them.

in the lexie "deadly weapon", "weapon" can be taken as an
entity qualified by a behaviour ("deadly").
if you dig enough and forget syntax for a minute
you can consider that weapon is the
instrument of, and refers to, various "behaviours" like
protect, kill, etc. that you would call "actions"
(but i can't use the word "action" here precisely because
"deadly weapon" is not an "énoncé actif" but an
"énoncé attributif").

but paired with "deadly", weapon likely refers to "to kill".
to give a silly example, "deadly" refers to the behaviour
"kill" in "weapon (entity) kills (behaviour)", "kill" is
taken as an entity, namely the causative actor of "to die"
(behaviour) implied in word "death".
or reversely you start with "death" as an entity result/goal
of behaviour "kill", etc.

there are plenty of categories of actors.
i think UNL's list of them is wrong (i.e., it's right if you're
a Japanese scholar who's fluent in english)-- but that's another

as another example, in "the corn is crunchy", you have to
consider "corn" as a "natural" patient of "to eat" rather
than a patient of "to be genetically modified", then switch
from, or maybe skip, "to eat food" to "to crunch on food"
(through a mode nobody knows quite well yet), then consider
that patient of "to crunch on food" is an entity paired readily with a
behaviour of yielding a certain resistance to your teeth, etc.

this looks stupid but it's actually why and how french engineers
are currently compending all possible lexies (they say they'll
find a few million) to make the first automatical french-
english translator. and when it's over we can eventually discuss
with each other without misunderstanding, Jim.
in 20 years or so. :-)

Christophe Grand made an interesting lang this way where all
words can be "activated" with a suffix. for instance "mouth+suffix"
means roughly "mouth working as the actor of a certain process".
of course you infer what process it is from "context" (i hate this
word) like in "shut your mouth, Mathias!"

i think i already wrote enough to respond to your lines hereunder,
with the following qualification, though:

"context" is no inaccessible divinity.
it can itself be analysed as a string of lexies chained
together as entity-behaviour pairs.
the only mysteries of "context" are
what is the number of implied énoncés needed to link expressed
lexies with each other and
how is it that the human brain can retrieve and link two lexies
distant of tens of words to pair them.

In "deadly weapon," the adjective refers to the purpose for which the
weapon was designed.
that's why it's adverbal.
So where is the verb that "deadly" modifies?   The fact that a weapon can be
used to kill is said of a weapon, not an action, making "deadly" an
adjective, not an adverb.

A deadly weapon is one designed to seriously injure
or kill,
the process of injuring or killing.
(Interesting that "the process of injuring or killing" is a noun phrase.
Reference to an action can be accomplished with nouns as well as verbs.
The noun "weapon" has a meaning intimately related to its intended use.
This use, injuring or killing, is an action, but "weapon" is nonetheless a

as opposed to a non-lethal weapon such as a net-gun.   The
adjective "deadly" can't refer to the actual use of the weapon, since the
actual uses include many actions not described by the meaning of the
deadly refers to one use among many uses, namely to kill.
instruments are more or less specialized.
(So the adjective "deadly" can only apply to "weapon" when it is being used
to kill?)

One can threaten someone with a deadly weapon, but "deadly"
does not mean "threatening."   One can collect deadly weapons, but "deadly"
does not mean "collectable."    One can sell a deadly weapon, but "deadly"
does not mean "saleable."    A deadly weapon could be used a paperweight
but...well, you get the idea.    The adjective refers to the purpose of the
weapon, and so describes the weapon, not an action.)
you're right : deadly refers to a particular use of the weapon.
did i say there was only one such process?
(Well, I didn't mean to convey that "deadly refers to a particular use of
the weapon."   I meant to convey that "deadly" characterizes the purpose
that the weapon was designed for.   I don't think that a deadly weapon stops
being "deadly" just because it's being used for some purpose other than
killing in a particular instance.)

(I admit, I'm confused by your use of the phrase "a particular use of the
weapon."   Do you mean a particular purpose for which the weapon could be
used in more than one context, or do you mean a particular instance of the
weapon's being used?)

a general is a general, not a nurse.
when you speak of his children, he's a father.
a wise general and a good father are actors of different
fields of activity i think.
(A general is not stripped of his rank merely because he is working with his
children, and reinstated as soon as he resumes his military duties.   He
remains a general all the while.   A wise general who had children might
want to avoid the pitfalls of the constant moving about that military duty
often entails.   Being a general and being a father are both considered
relatively stable statuses.   Though he cannot perform military and parental
duties at the same time (which duties might be described by adverbs), our
man retains the titles of both general and father over a long, stable period
of time (which titles are descibed by adjectives like "wise.")

"you play tennis well" is not quite the same as "you play tennis, which
is good". in the first instance "well" refers to how the process works
and in the second "good" qualify the entity of "playing tennis".

(I disagree on both counts.   "well" in this context does not tell us how a
process works;   it evaluates the person's tennis games
you mean the games he plays?
btw, according to your round-about method does a player
ever play? is a noun of agent an agent of any process?
Yes, I mean the games he plays.    The noun names the agent, and the verb
names the action.   I'm afraid I see nothing "round-about" in that analysis.

"playing tennis"
is not an entity, but an action,
it's both.
playing tennis is an action.
the fact of playing tennis is an entity.
the action of playing tennis is taken as an entity.
(I suppose a language in which all constituents could be analyzed as noun
phrases could include actions and events in their classes of entities.
Such a language would indeed have no distinction between adverbs and
adjectives.   However, I have never seen the term "entity" used in any
analysis as you have used it here.   Perhaps other listers have.)


maybe we should first agree on what is a substantive,
a predicate, a noun, a verb, an entity, a process, an action, a state, etc.
but we won't because apparently you won't consider the semantic.


I have spent most of this correspondance considering semantics.  But you are
right about our need to agree on the meanings of terms.

I am no expert on case grammar;   but from what I understand ...

I'm not sure what a "substantive" is;   I've heard this word used with
reference to nouns and adjectives with concrete referents.

A predicate, as I understand it, is a symbol or set of symbols that stands
for something ascribed to or equated with something else.   Predicates
usually correspond to comments (defined pragmatically) or verb phrases
(defined syntactically) in sentences.  I'm not sure whether verb phrases
have ever been defined syntactically for all time, but I know a predicate
when I see one in English.

An entity is a thing;   some body or idea that exists or subsists.
An action is what some entity does.
A process is a change that some entity undergoes.
A state is a condition in which some entity remains at least for a time.

A noun can stand for almost anything, including an entity, action, process
or state.   e.g.

dog (entity), war (action), fermentation (process), solidity (state)

That's why I define nouns according to syntactic criteria;   e.g.  can they
be arguments of the verb, do they take adjectives, etc.

A verb can stand for an action, process, or state.

pierce (action), melt (process), vegetate (state)

That's why I define verbs according to syntactic criteria;   e.g.   can they
have arguments, do they take adverbs, etc.

and "good" does not qualify only themention of this action;   it describes
what the entire preceding clausestands for.)
a clause taken as an entity, indeed.

("You play tennis, which is good."   does not mean  "You play tennis, and
playing tennis is good."
did i detail this?

It means "It is good that you play tennis," or
"Your playing tennis is good.")
so what?
(So it's not accurate to say "and in the second 'good' qualify the entities
of 'playing tennis,'" as you wrote in your last post.   That's what.)

i'm sorry to mess up words.
i'm no english native speaker, you know.
shall we keep on this thread in french?
(Your English is vastly superior to my non-existent French.   I have no
doubt that you will learn more as you communicate with more English
speakers.  As a monolingual English speaker, I salute you for becoming
fluent in a language other than your mother tongue.   How many natural
languages do you know, by the way?)

is a horrible accusator a person making horrible accusation
or an ugly accusator?
because you know, accusators like generals are not
always handsome and some are awsome tennis players.
although awsome may refer to the way their children retire.
or reversely.

Ha ha ha!