One of the important differences between
formal semantics and language on this point might be summed up as a distinction
between “proposition” and “predication.” If I
understand correctly, one of the goals of formal semantic representations is to
recast presuppositions and assumptions so that they are explicit – to prevent
the analyst from unconsciously or consciously slipping in a claim that is
implicit but not presented *as* a
claim. With natural language, however, we all take advantage of devices that
let us position information as asserted or assumed in statements. This can, for
example, be seen as one of the functions served by casting something as an
independent declarative clause (easy to contradict) vs. a subordinate clause (harder
or impossible) vs. a relative clause or nonfinite expressions (impossible to
contradict):
A Bjorn
murdered Brunnhild! B. No
he didn’t!
A’ You were there when Bjorn murdered Brunnhild B’ *No
he didn’t! No,
I wasn’t.
A” I
objected to Bjorn’s murdering Brunnhild B” *No
he didn’t No,
you didn’t.
The element that makes something finite –
what some functionalists call a finite operator and what I *think* gets cast as an element like INFL in
generative grammars – is crucial to the presentation of something as a
claim (and its position as well – changing S F to F S yields a question).
The formal-semantic interpretation of the
term “proposition” does not require that it is presented as
a claim, so states of affairs presented implicitly in Adj+N
sequences can count. The interpretation of the term “proposition”
outside of formal semantics, though, has “claim” built into it, I
think.
Bill Spruiell
Dept. of English
From:
Sent: Tuesday, October 11, 2005
12:04 PM
To: [log in to unmask]
Subject: Re: A logical thought
We’re getting into marvelously
irrelevant arcane here, but you’re still defining
“proposition” too syntactically. Wouldn’t “a
green house” in Joanna’s sentence “He lives in a green
house” be, logically something like, (and I don’t know how to type
an existential or a universal quantifier) Ex((x,house)&(x,green))? My
predicate calculus is a little rusty, so please correct at will. But
anyway, that proposition will be realized in English in whatever way the syntax
requires given a number of discourse functions and pragmatic assumptions.
Herb
Good
clarification. Risking driving the point far too deeply into the ground
for the present discussion, however, I would mention again that "green
house" is not the realization of a proposition. There is a proposition
that may be supposed by the author that the house is green, but it is never
intended in the former form as a proposition. (Even in Korean the
adjective is not tensed in attribute position.) One challenge with
semantics is that there are many aspects of meaning that are not expressed, and
even some that cannot be expressed either in symbolic logic (as formalized
by the "giants") or in natural languages. This excellent point
you made about poetry should always be kept in mind by anyone trying to
formalize semantics (describe its grammar).
Bruce
>>> [log in to unmask] 10/11/2005 9:29:42 AM >>>
Bruce has properly skewered me for being
elliptical to the point of near incomprehensibility. “Green
house” is a proposition, or, rather, it is the syntactic realization of
the same proposition that can be realized as “the house is
green”. The point is didn’t make while giving the example is
that the propositions expressed in predicate calculus have nothing whatsoever
to do with syntactic form. They are semantic in nature, but they are also
well short of the scope of natural language semantics, which is a function of
human cognition, and such giants of symbolic logic as Reichenbach, Carnap, and
Whitehead new quite well that formal predicate calculus and its extension
described only a subset of natural language semantics.
The point of the Yoruba and Korean
reference was that what can be a sentence is very different from language to
language. In those languages adjectives are tensed. In Yoruba,
prepositions are verbs and so would come from logical propositions. In
Finnish the negative marker is tensed. So what can be a complete thought
also differs from language to language. According to the strong form of
the Sapir-Whorf Hypothesis, facts such as these should correlate closely with
the culture and cognition of the speakers of these different languages, and in
such a way that the linguistic structures would determine the cognitive.
Granted, this strong form has been thoroughly repudiated in linguistics and
anthropology, although it’s still grist for the popular press.
Herb
From:
Sent: Tuesday, October 11, 2005
10:15 AM
To: [log in to unmask]
Subject: Re: A logical thought
I would like simply to comment on Herb's last paragraph. He
seems to have gotten a little loose with logic. I would be careful
in associating the logical proposition with the complete thought.
"Green house" in Johanna's sentences is not a proposition,
logically. It is a logical argument, meant to serve as a parameter in her
function. The fact that it doesn't have a finite verb is a result of the
fact that the function is unstated. A corresponding propostion would have
to take the form of "that the house be green" or "for the house
to be green" or whatever syntax would be required by the verb (as
expressing a function) that takes a proposition as argument. The fact
that Korean and Yoruba have a syntax in which verbs and adjectives
are nearly identical has little to do with "complete thought."
They are not propositions without arguments. They are still logically
predicates. Herb will have to fill in some more steps for me to see how
this all relates to the Sapir-Whorf Hypothesis. Does it have to do
with grammar being "useful and true"?
Bruce
>>> [log in to unmask] 10/11/2005 8:48:16 AM >>>
Sentences have
structure. Logical propositions have formal internal
relationships. Thoughts are pretty much non-definable. The sense
the
no relation to a proposition or a sentence, and one that probably could
have been expressed so vividly only by a genius of her caliber. Craig's
sense of "move in a series of moves" works a lot better, although
then
we're faced with defining "move", perhaps less daunting but still
knotty.
Actually, with respect to the content of a New Public Grammar, or of a
K12 Language Arts grammar curriculum, I can think of few better starting
points than what Craig lays out below. It fleshes out the catch phrase
I tossed out a year or so ago, "Grammar that's useful and
true." If we
make sure that what we are teaching is useful in talking about texts
students read or write and make sure that it bears some reasonable
resemblance to what students know and observe, we'll get rid of a lot of
junk, including terminology, and we'll force ourselves to think about
this with some clearer goals.
Like Joanna, I'm leery even of associating the logical proposition with
the complete thought. "Green house" in her sentences is a
logical
proposition. The fact that it doesn't have a finite verb is a result of
the particular move she had to make with that sentence. Actually, I
know of languages, like Korean and Yoruba, where there's no significant
difference between a verb and an adjective, which would mean that
"complete thought" would mean rather different things in Korean and
Yoruba than in English, even though the logical propositions would be
the same. This is at best a reduction to the absurd of the Sapir-Whorf
Hypothesis.
Herb
Subject: Re: Each sentence contains a thought
The only reason" complete thought" was ever coined
is to help
students avoid sentence fragments (which now show up in the work of many
fine writers.) We use this misleading notion because we somehow believe
that students can't understand what we really want, at least one
independent clause within a sentence unit. This requires, as Ed has
been pointing out, an understanding of the role of finite verbs and some
sense of subordination. (Finite verbs are necessary, but not always
sufficient.)
The problem with "complete thought" is that it
purports to be about
the rhetoric of the sentence (and not a constraint on form.)
I combat the misinformation with the following, which has become my
FIRST way into grammar for my writing classes.
1) A sentence is not a complete thought. It is a move in a series
of
related moves.
2) Sentence boundaries are very flexible.
3) Sentences vary widely in the amount of information they contain.
4) Sentences vary widely in the way that information is organized.
5) Sentences occur within the context of other sentences.
6) Sentences occur within the context of a writer's evolving purposes.
Students have been encouraged not to repeat themselves. Taken in tandem
with the "complete thought" notion, they believe their job is to
present
fifty or sixty "complete thoughts" in a 1,000 word essay. (And they
often write that way, beginning with sentences that avoid getting to the
point in order to avoid running out of things to say.) Effective
writers, on the other hand, tend to think in terms of saying one (or a
few things) and using the fifty or sixty sentences to carry that out.
If you look closely at professional and student work--and I do exactly
that with my students, the most recent being a paragraph from Stephen
Crane's "The Open Boat"--you see an enormous carryover of meaning in
the
best writing, including a willingness to reassert key ideas and keep the
same topic in focus over long stretches of text. They tend to have more
longer sentences than a typical student writer, but more short sentences
as well. They tend to put the most important information in key places,
especially at the end of intonation units and a the end of the clause.
Student work can often be radically improved when viewed through this
more rhetorical lens. Believe it or not, they tend to buy into it, pay
attention, and put it to work. The aim is in helping them achieve their
own purposes, offering them tools, helping them put their own language
to work.
It would be nice to put complete thought forever to rest, but we
can't do so if we want to insist on eliminating fragments but are afraid
to develop a sufficient base of understanding for that to truly happen.
If it feels like a "complete thought" to a student and
feels like
that to a reader, are we willing to accept that? Every time?
Craig
Johanna Rubba wrote:
> As a few others have noted, "thought" is way to broad a term to
name
> much of anything about a sentence. "Complete thought" is not any
better.
>
> It would be better to say that a _clause_ (or an independent clause)
> expresses a proposition, as Bruce notes. A proposition in logic is a
> statement that predicates something of an entity: "The sky is
blue"
> predicates, or attributes, blueness to the sky. The logical formula is
> f(x), meaning "f is predicated of x".
>
> It might look like this corresponds closely to the subject/predicate
> division in grammar, but it doesn't, once you move on from
> linking-verb sentence patterns. But when you get to transitive and
> other kinds of verbs, it doesn't line up so well:
>
> eat(child, cupcake) would be the formula for "The child is eating a
> cupcake."
>
> I think I have this right.
>
> Ed also raises the question of what looks like propositions inside of
> propositions:
>
> "He lives in a green house."
>
>
> Noun phrases with adjectives can be viewed as compressions of
> propositions, as can nominalized clauses such as the subject of
>
> _The corporation's outsourcing of customer service calls_ has led to
> complaints.
>
> There was once a theory of syntax that proposed that, indeed, even
> noun phrases with adjectival modifiers were derived from "deep"
> clauses; the theory was called generative semantics. As you can
> imagine, the derivation of quite ordinary sentences grew quite
> cumbersome.
>
> In any case, the logical-proposition idea is a good one, because it
> shows the crucial role of the main verb. It is the verb that
> determines the sentence pattern (linking, transitive, and so on).
>
> People concerned with correctness want sentences to "express a
> complete thought". A much better criteria for valid sentencehood
(that
> is, the quality of being able to "stand alone") are (a) presence
of a
> finite (present- or past-marked) verb and (b) the item is not a
> modifier or complement in a larger sentence (e.g., a relative
> [adjective] or adverb clause).
>
> For relatively short sentences, there are two pretty good tests.
> (1) Can the sentence appear in the blank in the following?
> "I am convinced that ____."
> (2) Can you add a tag to the sentence?
> "The hurricane wrought devastation across large areas of the Gulf
> coast, _didn't it_?"
>
>
> Dr. Johanna Rubba, Associate Professor, Linguistics
> Linguistics Minor Advisor
> English Department
>
> E-mail: [log in to unmask]
> Tel.: 805.756.2184
> Dept. Ofc. Tel.: 805.756.2596
> Dept. Fax: 805.756.6374
> URL: http://www.cla.calpoly.edu/~jrubba
>
>
> Dr. Johanna Rubba, Associate Professor, Linguistics
> Linguistics Minor Advisor
> English Department
>
> E-mail: [log in to unmask]
> Tel.: 805.756.2184
> Dept. Ofc. Tel.: 805.756.2596
> Dept. Fax: 805.756.6374
> URL: http://www.cla.calpoly.edu/~jrubba
>
> To join or leave this LISTSERV list, please visit the list's web
> interface at:
> http://listserv.muohio.edu/archives/ateg.html
> and select "Join or leave the list"
>
> Visit ATEG's web site at http://ateg.org/
>
To join or leave this LISTSERV list, please visit the list's web
interface at:
http://listserv.muohio.edu/archives/ateg.html
and select "Join or leave the list"
Visit ATEG's web site at http://ateg.org/
To join or leave this LISTSERV list, please visit the list's web interface at:
http://listserv.muohio.edu/archives/ateg.html
and select "Join or leave the list"
Visit ATEG's web site at http://ateg.org/
------------------------------------------------------------------------------
This message may contain confidential information, and is
intended only for the use of the individual(s) to whom it
is addressed.
------------------------------------------------------------------------------
To join or leave this LISTSERV list, please visit the list's web
interface at: http://listserv.muohio.edu/archives/ateg.html and select
"Join or leave the list"
Visit ATEG's web site at http://ateg.org/ To
join or leave this LISTSERV list, please visit the list's web interface at:
http://listserv.muohio.edu/archives/ateg.html and select "Join or leave
the list"
Visit ATEG's web site at http://ateg.org/