[analytic] Taking Truth as Primitive
Rodrigo Vanegas
At 02:34 PM 8/13/2001 -0300, you wrote:
>A theory can RELY on concepts that it derives from a more
>basic theory, or even, it may even simply postulate an object as, precisely
>a "theoretical object" without need of explicit exposition (if I may be
>redundant).
Oh, I see. You are supposing that I take the concept of truth-conditions
to be primitive. I don't, actually. In fact, I have no use for the
concept at all, being as I am, willing to speak only of
truth-biconditionals instead. Those, I believe, we all agree can be
explained. Of course, meanings are not truth-biconditionals, but then I
don't think that there are meanings.
Davidson, to be sure, does take the concept of truth (if not
truth-conditions) to be primitive. His best argument in defense of this
can be found in "The Folly of Trying to Define Truth" (_Truth_, eds.
Blackburn and Simmons). Not surpringly, I sympathize with his attacks on
those who would demand definitions or explanations of the concept of truth,
beyond the trivial disquotational story and related accounts of theory,
quotation, and interpretation. I only disagree with Davidson in that I
wouldn't describe truth as primitive. Doing so very misleadingly suggests
that one is engaging in first philosophy, separating the primitive concepts
from the derivative ones into a systematic conceptual scheme that builds
upon foundations. To be sure, Davidson is probably aware of this
connotation, so it is not uncharitable to ask what he means when he says
truth is primitive. After all, why shouldn't we take the concept of
satisfaction to be primitive instead, defining truth in terms of
satisfaction? I think the gist of it is that he does not think truth can
be naturalized, that is, explained in terms that are not semantical. This
is consonant with the rest of his work and also makes it clear that there
are a set of related notions which easily hang together, but not easily
apart. To take truth as primitive, then, is just not to expect a
naturalization of truth and to work with the concept as it is interdefined
with the other basic semantical concepts, which include satisfaction,
reference, belief, intention, and mind.
Now, as I understand it, Baker himself is sympathetic to the idea that
semantics cannot be naturalized. Why then should he object to taking truth
as primitive?
Rodrigo <Van...@yahoo.com>
(c) 2001 by Analytic
http://analytic.ontologically.com/
http://groups.yahoo.com/group/analytic/
Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/
J L Speranza
Best,
JL
GC.
Rodrigo: "Now, as I understand it, Baker himself is sympathetic to the idea
that semantics cannot be naturalized. Why then should he object to taking
truth as primitive?"
I'm afraid I don't really know what Baker is sympathetic no more [sic.
JLS]- I tho't the man was English to the backbone, but then TP Uschanov
tells me he knows the man is American. So maybe he has a personal thing
against Davidson!
Rather than go thru Baker's _Language, Sense and Nonsense_, I rather
re-read bits from his contribution to Grice's festschrift (ed by R.
Grandy). Baker's essay is titled "Alternative Mind Styles", where he
opposes one mind style (presumable Grice, but also Frege, Quine, Davidson,
and everybody non-latterlish Wittgensteinian) with _another_ mind style:
viz. the Latter Wittgenstein's one. He writes where Grice fits in in all
these:
"Grice is one of the few philosophers from whom I have learnt something"
[That's nice for a start in a fest, I guess. JLS]... No doubt I have learnt
far less than he taught [Implicature: he taught too much. JLS] ... & hence
I have profited less from his work than HE might reasonably hope" [Trouble
is Grice had no reasonable hopes. Trust Baker to turn a possible compliment
into something quite nasty: the "he" there I find self defeating] ... I
always admired Grice's heresies transformed by his success into orthodoxies
[Baker is here being ironic, since he is so anti-mainstream, that he
probably implicates Grice's heresies had remained so!] ... but my intention
in this veritable Trojan horse of a fest contribution is to raise questions
about the very framework in which he constructs ingenious answers ... Some
intention to honour somebody! ... So this essay may be viewed as an
antidote to the intoxication brought on by a deep draught of "The Logic of
Conversation" [Only Grice never wrote that. He wrote "Logic _and_
conversation". I wrote a whole essay about _mis_taking Grice's "logic and
conversation" as it being "logic of conversation". Chomsky commits it too.
JLS]"
Baker then turns to a predecessor of Grice: Gottlob Frege, qua precursor of
truth-conditional semantics. In this connection, I find it interesting to
relate Baker's discussion to a passage from K Bach's essay on "The Myth of
Conventional Implicature" cited by L Tapper. Bach writes:
"Grice is usually credited with the discovery of conventional implicature,
but it was actually Frege’s idea. Grice merely labelled it. Cfr. Frege:
"Subsidiary clauses beginning with "altho'" express complete thoughts. This
conjunction actually has no sense and does not change the sense of the
clause but only illuminates it in a peculiar fashion. Similarly in the case
of "but" and "yet". We could indeed replace the concessive clause without
harm to the truth of the whole by another of the same truth value; but the
light in which the clause is placed by the conjunction might then easily
appear unsuitable, as if a song with a sad subject were to be sung in a
lively fashion ('On sense & reference', in RM Harnish, _Basic Topics in the
Philo of Lang. Prentice Hall, p. 155) and "With the sentence
1. Alfred has still not come
one really says
2. Alfred has not come.
and, at the same time, hints that his arrival is expected, but it is only
hinted. It cannot be said that, since Alfred’s arrival is not expected, the
sense of the sentence is therefore false. Thus
3. p BUT q
differs from
4. p AND q
in that with (3), the utterer intimates that what follows is in contrast
with what would be expected from what preceded it. Such suggestions make no
difference to the thought. (The Thought: a logical inquiry, in RM Harnish,
op cit, p.522)". Bach comments: "in other words, "still" and "but" (beyond
its conjunctive import) have no bearing on the truth or falsity of what is
said."
=====
Baker then turns to demolish Frege. Baker writes: "Frege is _reputed_ to
have given the first truth-table definitions of the connectives of the
propositional calculus (p.278) ... He employs the notion of "sense" to give
the first analysis of meaning as truth-conditions and introduces the
distinction between sense and FORCE which is indispensable for generalising
truth-conditional semantics to the whole of any natural language such as
English... Yet, some of this lustre appears chipped. First, he lacks the
notion of a "truth-condition" familiar from the _Tractatus_. Hence he was
at one remove from the idea that the concept of sense grows out of the the
conception of meaning as truth conditions. Also, he never actually
formulated _truth_-definitions for each of the connectives. He gives
tabular explanations for them, but they cannot be construed as names of
_truth_ functions. Actually, his demand for completeness atually excludes
the legitimacy of truth table expnaantions. Finally, his thesis that the
_reference_ of a sentence is a "truth-value" -- as Baker quotes from the
diagram from Frege's Philosophical and Mathematical Correspondence. Oxford:
sentence
.
.
.
sense of a
sentence
(thought)
.
.
.
reference of
a sentence
(truth-value)
-- "rests on a sequence of equivocations and question begigings". Baker
refers to his _Logical Excavations_ (co-authored with P. Hacker), where
these "claims are defended in detail". "Yet", Baker admits, Frege did
"discover entities having novel logical forms and he actually identified a
galaxy of new logical forms". Frege's main fault, for Baker, seems to be to
have "declared that his discovery of the two logical objects
1. The True
2. The False
was as "important as the discovery of a pair of new elements in chemistry
... In the Basic Laws "truth value" is treated as a basic value of a
function... Yet Frege dogmatically denies that truth is a property of
thoughts, asserting instead that "the True" is really the reference of what
we misleadingly call TRUE judgements ... (Baker, p.287). ... He then
DECIDES -- or discovers! -- that a pair of a thought and a truth value
correspond to a putative judgeable content!"
Grice writes in the fest: I am moved that the authors of this work are nnot
just colleagues: Every one of them is a friend" (p.45), so don't let us
despair!
Baker quotes Dummett's _Truth & Other Enigmas_. Like him, Baker seems to
defines himself as an "anti-realist. Baker's brand of semantics he calls
"criterion-semantics". from an essy in Ratio, No. 16.
Now, turning to _Language, Sense and Nonsense_ Baker does write, as per
quoted in Rodrigo's post,
"The absence of the term "truth-condition" from the exposition of a theory
does not prove that the theory does not rest on teh concept of a
truth-condition" (p.191).
I looked for "truth" in the index and it is quoted on pp.12, 156, 180-90,
and 198. Let's review: On p.12, Baker states that his strategy will be to
try and UNDERMINE the presupposition of truth-conditional semantics. I.e.
not so much the truth of an hypothesis but the intelligbility of a thesis.
So far so good. On p.156: he notes that "explaining meaning in terms of
truth conditions amounts to a CLARIFICATION of meaning only on the
supposition that the concept of "truth" itself is clear and unproblematic.
This point is now a focus of dispute. Dummett contends that the most
crucial general problem in semantics is to settle "what notion of truth is
admissbile" (In p.xxii of _Truth & Other Enigmas_ he presupposes hat
"truth" is a THEORETICAL CONCEPT). ... The realism/antirealism debate
turns on how to relate TRUTH TO VERIFICATION" [I wish RB Jones would help
us here!] ... "Realism" maintains that the question whether we have any
means to determine the truth of a sentence has no bearing whatever on the
question whether the sentence is true. Truth is said to be
"verification-trascendent ... but Dummett sees this as the product of
slovely analogical thinking" [And Baker seems to agree there! I like
"slovenly"!] ... Viewed thru Dummett's spectaclces the principle of
verification espoused by logical positivists rested on an anti-realist
conception of truth".
On p.198, Baker discusses the notion of "truth" as applied to context
sensitive utterances ("I am speaking"). This may relate L. Tapper's quibble
that it's utterance TOKENS which are primary bearers of true. Baker there
quotes Davidson (!):
5. Truth is a relation between a sentence,
a person, and a time.
('Truth and meaning' p.319)
The substantial bit are those pp.180-90 from where I quote:
"Truth-conditional semantics treats "truth" as a _pivotal_ concept in
understanding meaning. ... Thus Davidson writes, "to know what is for a
sentence to be true amounts to understanding it" ('Truth and Meaning',
p.310). Similarly, Tarski sought for a "definition of "truth" in a a
materially adequate and formally correct definition of the term "true
sentence" (The Concept of Truth" in _Logic, Semantics & Metamathematics_,
p.152) ... Would anything become clearer when we encounter the thesis that
a Tarskian truth-definition does not yeild a complete ecplanation of the
concept of truth? (as we do in Dumett, _Truth_ p. xxi). The philosopher who
wants to regiment may say, argues Baker, that adjective "true" is
vague
ambiguous
and incoherent
one may say. It is applied to
a physical object like a sentence
a psychological phenomenon like a belief
or even
an abstract entitity like a proposition
(Tarski, 'The semantic conception of truth', p.53).
So, it is fair to expect that the philosoher will like to deal with "true"
as it were a "theoretical concept" or term. But, Baker, notes, someone who
says
6. It is true that snow is white.
has notably "not predicated truth of a platonic object. Thus, the idea that
truth is a theoreitcal concept is LUDICROUS [And there was I thinking Baker
was a Dummettian!] ... The philosopher's look on the word "true" as
technical jargon in some explanatory theory is not a source of insight but
DISTORTION. Then there's the polemic whether "true" is predicated of a
sentence or to the structural description of the sentence" But, Baker
writes "I shall charitably ignore this". He notes the nominalist strategy:
"In his attempt to escape abstract entities, the philosopher may treat
"true" as a property of INSCRIPTIONS". And Baker finds this reasonable:
"Given that a special token of a sentence, e.g.
7. The current King of France is bald.
can be "true", "false" or "neither true nor false" on occasion suggests
that the doctrine that truth is a property of utterance TYPES needs to be
clarified... But it is not at all obvious that any coherent explanation
will be forthcoming..."
Baker finds further problems in the fact that "a particular token (e.g. a
signal) can be said to be TRUE: a true assertion may be made by a
particular gesture.". Thus he proposes (tenatively) that we view "true" as
within the family of terms like
8. accurate, right, wrong, exaggerated, exact, clear, certain, refuted,
well-founded, vague, obscure, definite, faithful, literal, metaphorical.
But the truth-conditional semantics will deny (8) and stick to a hateful
"AD HOCNESS": "the treatment of "true" rather than any of (8) as _the_
semantic property predicated of a "sentence" is motivated solely as a
precondition of constructing truth-conditional semantics."
"A truth-conditional theory owes us a clear expalantion of "true"
(p.187)... Should a truth-theorist subscribe to the doctrine that truth is
a simple INEXPLICABLE PROPERTY? This proposal hardly recommends itself as
the basis for introducing a technical term in the construction of a theory.
Introducing a technical terminology for expository purposes is permissible
but such expressions must be clearly and coherently explained. Although
from fear of embarrassment at not comprehending what others seem to find
obvious we run the risk of persuading ourselves that we understand what is
incomphrensible..." [Must say I like that turn of phrase v. much]... The
strategy of treating "true" as a theoreitcal concept is misguided: the
truth-conditional smeanticists's claim that meaning is a matter of "truth
conditions" is interesting only to the extent that the semanticist's
concept of "truth" coincides with mine qua competent native Engish speaker
who knows what I mean by "true". But the semanticist's pressure to declare
his concept to be a technical ((OR A PRIMITIVE. JLS)) one arises from his
recognising important divergence in this matter. He is wishing to bestow on
"true" a novel use. He affirms things like
9. The truth predicate is a device of
disquotation
(Quine, _Philosophy of Logic_, p.12.
cfr Davidson, In Defense of Contention T)
But, that's no way out, since the fact remains that "defining "true" as a
theoretical term deprives truth conditional semantics of its intuitive
support and, what's worse, threatens to make it incomprehensible -- at
least to me."
Unfortunately, there's no reference to "satisfaction" in the index to
Baker's book, but must say, I like Rodrigo's idea (following Tarski, and
Grice) to take, perhaps, "satisfaction" as more primitive than true! A good
thing about that is that we don't have to quarrel with native speakers's
intuitions, seem they don't seem to hve ANY idea about what "SATISFACTION"
in the first place means -- at least my friends don't. I say this is
compatible with Grice since he does prefer in "Logic & Conversation" to
speak of "factual satisfactoRINESS" (not even "satisfaction") as
philosophically more interesting than "truth", no doubt to tease his
Harvard audience...
This began as a discussion as to whether Baker was a semantic naturalist
(Does he think semantics can be naturalised?). I think he is. I.e. he seems
to very much favour (if you excuse me the split infinitive) the idea, as
his "teacher" Grice does, that semantics CAN (and even SHOULD) be
naturalised. In Grice's case, in terms of psychology. In Baker's case, in
terms of "Criteria" -- which are also a psychological notion -- Baker never
uses "criteria" in the singular since he holds that for each lexeme there
must be at least a defeasible set of criteria... Else we're not speaking
proper, natural, English...
====
Rodrigo Vanegas
At 10:35 PM 8/13/2001 -0300, you wrote:
>Rodrigo: "Now, as I understand it, Baker himself is sympathetic to the idea
>that semantics cannot be naturalized. Why then should he object to taking
>truth as primitive?"
>
>I'm afraid I don't really know what Baker is sympathetic no more [sic.
>JLS]-
I think you're right that it's not so simple. You've given me a lot of
material to work with, and I am feeling a little overwhelmed. I think what
I'll do is just pick one or two ideas and comment on those.
>7. The current King of France is bald.
>
>can be "true", "false" or "neither true nor false" on occasion suggests
>that the doctrine that truth is a property of utterance TYPES needs to be
>clarified... But it is not at all obvious that any coherent explanation
>will be forthcoming..."
>
>Baker finds further problems in the fact that "a particular token (e.g. a
>signal) can be said to be TRUE: a true assertion may be made by a
>particular gesture.". Thus he proposes (tenatively) that we view "true" as
>within the family of terms like
That truth is a property of sentences is always a simplification that
idealizes context away. The more accurate rendition has it that truth is a
relation between a sentence-type, a person, and a time. The sentence "the
present King of France is bald" is sometimes true and sometimes false, but
its truth-conditions are always the same: "the present King of France is
bald" is true at time t iff the King of France is bald at time t.
Speaking for myself now, I do not take "truth" or "truth-conditions" to be
technical terms. When I say ""snow is white" is true", I do not intend to
use the term in any special way. I can easily imagine ordinary
conversation in which one person says "snow is white" and another says
"that's true". I mean what they mean. But I also agree with Quine that
the truth predicate is a device of disquotation, and with Davidson that
sentences are meaningful if they have truth-conditions. It seems
provocative, at best, to insist that I think truth can be predicated of
sentences only because I have a philosophical axe to grind. It simply
isn't an accurate charge and I plead innocent.
Rodrigo <Van...@yahoo.com>
Larry Tapper
> After all, why shouldn't we take the concept of
> satisfaction to be primitive instead, defining truth in terms of
> satisfaction?
Rodrigo,
I see this comment of yours caught Speranza's eye. Mine too.
Speranza notes that the beauty of it is that the native speaker knows
even less about satisfaction than he knows about truth-conditions.
About this I'm not sure, though, because an informal notion of
satisfaction seems to me to be one of the main things people
argue about, when they do argue about truths.
For example, I tell you that
(1) The cat is on the mat.
And if you are inclined to doubt me, you may very well ask such
questions as:
Is it really a CAT on the mat?
Is it really a MAT the cat is on?
Is the cat really ON the mat?
Which all seem to be questions about the sets of objects that
satisfy 'x is a cat', 'x is a mat', etc.
Now as far as I can dimly remember, Tarski mostly just cared about
whether one's formal definition of satisfaction covered all the
syntactic possibilities adequately, in the abstract. For example
existential quantifiers are definable in terms of sequences. Once
you've done all this spadework, defining satisfaction schematically
for all the wffs of the object language, you can then move on to a
consideration of truth-definitions at the sentence level. (Please
correct me if I'm wrong, it's been a long time since I've read
Tarski.)
But I think Tarski did say somewhere that it was unlikely we could
define satisfaction adequately for a natural language, partly because
natural language is vague and the relevant sets are, as we would now
say, fuzzy. The set of objects that satisfy 'x is a cat' is not so
well-defined, even in a single idiolect.
Not only that, even our logical connectives are ambiguous, e.g. "or".
Anyway, I like this focus on satisfaction, even if it's based on a
totally erroneous reading of Tarski, because it underscores the sense
in which "truth-conditions" (conceived informally as the kinds of
things normal people might actually argue about) are something like
regions in an attribute space.
That is, I picture the truth-conditions of (1) as being something
like a three-dimensional solid. At the origin you put your
stereotypical cat, your stereotypical mat, and your stereotypical
situation in which something is on something else. Near the outside
surface of the solid are marginal cases like Tibbles with his front
paws on a mat and The Lion King on a Persian rug. The general idea is
that (1) is true if the observed conditions lie within the solid,
else not.
I think this may be a useful metaphor because we can play with the
characteristics of the n-dimensional solid.
For example, my choice of (1) as an example was obviously not so hot,
because "the cat" and "the mat" are expressions that are typically
expected to be used referentially. No problem, substitute the
intended cat for the stereotypical cat at the origin, substitute the
intended mat for the stereotypical mat, unintended referents
disappear, and then (1) is true if the observed conditions lie on the
line segment through the origin representing the normal uses of "is
on". This method I call "speaker's dimension-elimination". Of course
it explains nothing, but what do you expect, it's just a metaphor.
I don't think these notes are much good but I'll post them anyway,
because I have a feeling they might make a decent launching pad for
other posters.
Regards,
Larry
Rodrigo Vanegas
At 01:38 AM 8/16/2001 +0000, you wrote:
>--- In analytic@y..., Rodrigo Vanegas <vanegas@y...> wrote:
> > After all, why shouldn't we take the concept of
> > satisfaction to be primitive instead, defining truth in terms of
> > satisfaction?
>
>Rodrigo,
>
>I see this comment of yours caught Speranza's eye. Mine too.
>
>Speranza notes that the beauty of it is that the native speaker knows
>even less about satisfaction than he knows about truth-conditions.
>About this I'm not sure, though, because an informal notion of
>satisfaction seems to me to be one of the main things people
>argue about, when they do argue about truths.
>
>For example, I tell you that
>
>(1) The cat is on the mat.
>
>And if you are inclined to doubt me, you may very well ask such
>questions as:
>
>Is it really a CAT on the mat?
>Is it really a MAT the cat is on?
>Is the cat really ON the mat?
>
>Which all seem to be questions about the sets of objects that
>satisfy 'x is a cat', 'x is a mat', etc.
Yes, I think so too. Besides, satisfaction and truth are
interdefinable. We know how truth is defined in terms of satisfaction, but
satisfaction is also definable in terms of truth. Roughly,
a satisfies "F_" iff "Fa" is true
for all names and definite descriptions "a".
>Now as far as I can dimly remember, Tarski mostly just cared about
>whether one's formal definition of satisfaction covered all the
>syntactic possibilities adequately, in the abstract.
Well, that's why we say "Tarski-style" and not "according to
Tarski". Davidson borrowed the structure of Tarski's truth-definition, but
not Tarski's project or his philosophy of language.
Now, I am aware of some objections to the idea that truth-conditional
semantics can be applied to natural language, but I have never been
convinced that these objections are not easy to overcome, in the manner of
the so-called Davidsonian program.
That connectives such as "or" are vague or ambiguous, for example, is a
simple case of a problem easy to overcome. What must be understood is that
the truth-definition will be expressed in a meta-language that includes the
object language as a subset. So,
"p or q" is true iff p or q
perfectly captures the ambiguity and vagueness of the connective because
the "or" on the left, is the same as the "or" on the right.
J L Speranza
or finish for that matter.
Best,
JL
=====
Rodrigo:
"You've given me a lot of material to work with, and I am feeling a little
overwhelmed."
My meaning-intention. :) Overwhelm people.
"That truth is a property of sentences is always a simplification that
idealizes context away. The more accurate rendition has it that truth is a
relation between a sentence-type, a person, and a time. The sentence
[adapted]
1. the cat is on the mat.
"is sometimes true and sometimes false, but its truth-conditions are always
the same:
1b. "the cat is on the mat" is true at time t
iff the cat is on the mat at time t.
Fair enough. I guess this leaves possible worlds out of the picture. But I
can do without them, too! (cfr. H20 fills this glass (sic) vs. XYZ fills
this glass).
"I do not take "truth" or "truth-conditions" to be technical terms. When I
say
1c. "The cat is on the mat" is true.
"I do not intend to use "true" in any special way. I can easily imagine
ordinary
conversation in which
1d.
A: the cat is on the mat.
B: that's true
"I mean what they mean."
Well for Strawson -- I think he calls this (or his enemies do) the Ditto
Theory of Truth. I.e. B is just saying,
1e.
B: Ditto!
which leads straight to Ramsey's redundancy theory ("... is true" adds
little to asserting plain "p"). I know there is a way out of this, and
follow that way out!
"I also agree with Quine that the truth predicate is a device of
disquotation, and with Davidson that
sentences are meaningful if they have truth-conditions."
Cfr. my quote from Strawson, "Meaning and Truth" below. I think Davidson
was certainly Strawson's target in 1969, since Davidson had just published
the Synthese essay two years ago.
"It seems provocative, at best, to insist that I think truth can be
predicated of sentences only because I have a philosophical axe to grind.
It simply isn't an accurate charge and I plead innocent."
I know. I liked that "provocative at best". Nice implicature!
Larry:
>Speranza notes that the beauty of it is that the native speaker knows
>even less about satisfaction than he knows about truth-conditions.
>About this I'm not sure,
But then you _are_ a native speaker. So you would NOT be sure, would you
not. QED :)
>though, because an informal notion of
>satisfaction seems to me to be one of the main things people
>argue about, when they do argue about truths.
And also, I submit, when they argue about commands, such as
2. Fly me to the moon"
Is that satisfactible?
>For example, I tell you that
1. The cat is on the mat.
>If you are inclined to doubt me, you may very well ask such
>questions as:
3. Is it really a CAT on the mat?
4. No, it's a nasty woman
(Oxford English Dictionary: "cat. 1. felis domesticus. 2. nasty woman ((I'm
currently studying metaphor -- and I take 1 and 2 to be differences in
USAGE, not senses -- I'm a bit scared of Fregean senses. Also there's no
point in postulating two senseS AND one conversational implicature,
especially when One Implicature AND One Sense seem to do just fine. I see
this is your point indeed below when talking of stereotypes and
speaker-free meaning!).
5. Is it really a MAT the cat is on?
6. No, "on the mat" I mean "she is being punished"
by her boss (since "cat" can also mean
"prostitute").
(From OED: mat: to be on the... 1. to be on a rug. 2. to be punished by
someone in charge).
7. Is the cat really ON the mat?
8. Now, just seems to.
8b. Implicature: Are you _kidding_.
8c. Do you think a cat can float like that?
>Which all seem to be questions about the sets of objects that
>satisfy 'x is a cat', 'x is a mat', etc.
"An utterance is factually satisfactory if the S-item belongs to the
P-class" (Grice, Studies, p.56). And cfr re "formality": "there is this
demand for a THEORETICALLY ADEQUATE specification of conditions which will
authorise the assignment of TRUTH-CONDITIONS to suitably selected
expressions -- thereby endowing those expressions with a conventional
meaning. Such provision is needed. But the fact that [such provision is]
needed is insufficient to show that [it is] available. We might be lift in
the sceptic's position of seeing clearly what is needed, and yet being at
the same time totally unable to attain it!" (idem, p.364). These
conditions, Grice notes should be of course "conditions of truth". We
should suppose that what is being sought is not the meaning, as such, but
"the"/a"
solid guarantee that, in certain conditions,
in calling the S a P the utterer it not be
mispredicating P of the S
Cfr. Strawson on meaning and truth: "It is indeed a generally harmless and
salutary thing to say that to know the meaning of a sentence is to know
under what conditions one who utters it says something true. But if we with
for a philosophical elucidation of the concept of meaning, the dictum
represents not the end but the beginning or our task. It narrows our
problem to inquire what is contained in that little phrase as 11"
"... says something true".
"When we come to try to explain in general what is to say something true,
to express a true proposition, reference to BELIEF or to ASSERTION (and
thereby to belief) is INESCAPABLE"
Note he is saying this as Prof of Metaphysics. See papers by successor in
the chair, CAB Peacocke -- especially inaugural lecture! (_Trascendental
Arguments in the Theory of Content_, OUP)
"Thus we may harmlessly venture
"Someone says something true" iff
things are as he says they are.
"But this "says" already has the force of "asserts". Then we may propound
"Someone propounds a true utterance" iff
things are as anyone who believes what he
propounds would thereby believe them to be!
"or some such... In the latter definition, the reference to belief is
explicit and INESCAPABLE"
Yes, but it's s also kind of meaningless?
>Now as far as I can dimly remember, Tarski mostly just cared about
>whether one's formal definition of satisfaction covered all the
>syntactic possibilities adequately, in the abstract. For example
>existential quantifiers are definable in terms of sequences. Once
>you've done all this spadework, defining satisfaction schematically
>for all the wffs of the object language, you can then move on to a
>consideration of truth-definitions at the sentence level.
But
> The set of objects that satisfy 'x is a cat' is not so
>well-defined, even in a single idiolect.
Well, it is in mine. On any ordinary day, in which I awake with an
unstobbale wish to think in English, by "cat" I mean
9. x is a cat iff x has phenomenalist properties,
A: has chromosome structure felis domesticus
(A being kinda essential)
B: prototype: it purrs, and answers to "Tibbles".
C: it miaows.
Actually, when in doubt (usually when other people use _cat_ or when I'm
writing emails to ANALYTIC, I standardly consult the OED. Do you want me
to!? :) Tomorrow, when/if the cat's away...
>Not only that, even our logical connectives are ambiguous, e.g. "or".
Rodrigo does considers this, but I think Rodrigo is right in saying that
"ambiguity" is out of the question here, even for Grice, and I'm sure you'd
agree. It's basically UNIVOCITY plus IMPLICATURE, ennit. One good thing (or
"the beauty" as you indulging in second-order PC say) about the
"implicature" is that it allows you to get rid of the idea that things like
"or" are "ambiguous". They are "monoguous". Horn has like 5 essays on that.
He also calls words "autohyponyms" (e.g. dog means "dog", but also "male
DOG". Therefore, "dog" is an autohyponym of itself).
The ambiguity thesis (and how to fail it) was defended in Oxford inter alia
by Strawson, and later, by LJ Cohen ("On Grice's analysis of logical
connectives" -- should revise this.).
> I picture the truth-conditions of "the cat is on the mat"
>as being something
>like a three-dimensional solid. At the origin you put your
>stereotypical cat, your stereotypical mat, and your stereotypical
>situation in which something is on something else.
Right -- and illustrated as in SE Toulmin's book on INFERENCE. I love that
illustration. Apparently, Anglophones use (2) as paradigmatic (as a matter
of history) because it featured in reading-books in England in the 1930s
(with emphasis on phonetic/nondyslexic correlatin "kaet"//"maet", etc).
>Near the outside
>surface of the solid are marginal cases like 10 and 11.
10. Tibbles has his front paws on a mat.
11. The Lion King is on a Persian rug.
I agree. But Tibbles is a Yorkie (on closer inspection) -- while the Lion
King is only a 'toon! You cannot help being metaphoric if only while
writing for ANALYTIC, Larry!?
>The general idea is
>that (2) is true if the observed conditions lie within the solid,
>else not.
Mmm. Nothing wrong to say (11) is true for me -- but then I never saw a
lion being king. Or even (15), provided I define, as you say below, "The S
is on SOMETHING if at least part of the S is on the something."
cfr.
12. He ate the apple
13. +> He ate (all) the apple, including the hard inner part.
(whatever is called) (There's a MIT essay on this implicature involving
informativeness/quantity violation!).
> This method I call "speaker's dimension-elimination". Of course
>it explains nothing, but what do you expect, it's just a metaphor.
Loved that. Mind, I'm studying metaphor, and according to say, DE Cooper,
metaphors should explain. His example is
14. "Truth is a woman" (F. Nietzsche).
According to Cooper, even tho' (14) may be a metaphor, the theory which
(14) purports to illustrate may not! Some metaphors are better than others...!
>these notes are much good but I'll post them anyway,
>because I have a feeling they might make a decent launching pad for
>other posters.
Well, I guess I'd be factually satisfactoried if mine at least don't make
an indecent launching pad, or would I? :)
=======
Larry Tapper
Rodrigo,
I'm pretty sure that Tarski thought of "intensional vagueness" as a
real problem, but I can't remember the exact reasoning. Maybe it had
to do with the fact that satisfaction was defined as a relation
between an open sentence and an ordered sequence of n-tuples of
objects. So maybe the sequence corresponding to a vague predicate is
either undefined, or disorderly, or both.
Another classic Tarskian objection, which I didn't mention, was that
natural languages are semantically closed: that is, they contain all
the truth-related terms which according to Tarski, must be exiled to
the metalanguage in order to avoid the semantic paradoxes. This was
one of Tarski's primary motivations for all this Convention T stuff
in the first place, so it needs to be taken seriously, one would
think. What does Davidson say about this?
Anyway, you mention the "Davidsonian project" and I'm curious --- how
is it going? Have you been following it? I'm about 20 years behind ---
I remember when people like Burge and Harman were working on various
pieces of the puzzle. One guy would do oratio obliqua, someone else
would do adverbs. Personally, I'm always a little suspicious of
philosophical enterprises that present themselves as "normal
science", though of course interesting things happen in any event.
Regards, Larry
Rodrigo Vanegas
At 12:46 PM 8/16/2001 +0000, you wrote:
>I'm pretty sure that Tarski thought of "intensional vagueness" as a
>real problem, but I can't remember the exact reasoning. Maybe it had
>to do with the fact that satisfaction was defined as a relation
>between an open sentence and an ordered sequence of n-tuples of
>objects. So maybe the sequence corresponding to a vague predicate is
>either undefined, or disorderly, or both.
Well, again, that could only be a problem if the meta-language is not
itself undefined and disorderly.
>Another classic Tarskian objection, which I didn't mention, was that
>natural languages are semantically closed: that is, they contain all
>the truth-related terms which according to Tarski, must be exiled to
>the metalanguage in order to avoid the semantic paradoxes. This was
>one of Tarski's primary motivations for all this Convention T stuff
>in the first place, so it needs to be taken seriously, one would
>think. What does Davidson say about this?
To be honest, I forget, and I have my own ideas about how to deal with the
semantic paradoxes. I think Davidson says that so long as we restrict the
quantifiers in our truth theory so that the semantic terms always predicate
over *another* language, then there is no trouble.
I do remember that the place to look is page 28 of _Truth and
Interpretation_. :-)
>Anyway, you mention the "Davidsonian project" and I'm curious --- how
>is it going? Have you been following it? I'm about 20 years behind ---
>I remember when people like Burge and Harman were working on various
>pieces of the puzzle.
I have no idea, really. I only vaguely suspect that it has merged with
linguistics. Larson and Segal's _Knowledge of Meaning_ and Ludlow's
_Semantics, Tense, and Time_ certainly look like the Davidsonian
program. The first one even has a blurb by Davidson on the back cover. A
look inside reveals, however, that it has little of Davidson's holism and
anomalous monism. I don't know the inside story or whether there is
anyplace else to look.
>One guy would do oratio obliqua, someone else
>would do adverbs. Personally, I'm always a little suspicious of
>philosophical enterprises that present themselves as "normal
>science", though of course interesting things happen in any event.
Is it "normal science"? It's not as though we haven't seen division of
labor elsewhere in philosophy. I don't think it's much more than that. On
the other hand, the linguists/philosophers I mention above certainly do
seem to think they are riding the transition out of speculative philosophy
and into the hard sciences. I am also suspicious, but have only the most
minimal exposure.
Rodrigo <Van...@yahoo.com>
------------------------ Yahoo! Groups Sponsor ---------------------~-->
The Nissan Sentra
Everything but compact
http://NissanDriven.com
http://us.click.yahoo.com/3vsIKC/txlCAA/ySSFAA/9rHolB/TM
---------------------------------------------------------------------~->
Rodrigo Vanegas
At 05:35 AM 8/16/2001 -0300, you wrote:
>And also, I submit, when they argue about commands, such as
>
>2. Fly me to the moon"
>
>Is that satisfactible?
It sure is. The truth-conditions are trivial, as usual,
"a is flown to the moon" is true iff a is flown to the moon.
Factoring in the force of the utterance, the person, and time,
"fly me to the moon" is obeyed when said by x iff x is flown to the moon.
Of course, depending on the person and the time, we may want to take the
rhetorical force of the speech act to be metaphoric. If the person is
Sinatra and the time is one fine Saturday evening of the 1940s, then we
could not help but notice that Sinatra isn't asking to be transported to
Earth's only natural satellite. Rather, he is asking his baby to hold his
hand, kiss him, and please be true. Remarkably, even syntactic force
markers can be taken metaphorically. He's only telling her that he loves her.
Rodrigo <Van...@yahoo.com>
Larry Tapper
Speranza,
Various light comments on your latest post.
> Larry:
>
> >Speranza notes that the beauty of it is that the native speaker
> >knows even less about satisfaction than he knows about truth-
> >conditions.
> >About this I'm not sure,
>
> But then you _are_ a native speaker. So you would NOT be sure,
> would you not. QED :)
Careful now, those two Latvian guys are reading every Analytic post
very thoroughly in the hope of improving their English.
>
> 3. Is it really a CAT on the mat?
>
> 4. No, it's a nasty woman
>
In our attribute-space model, this is no problem: we just add an
extra dimension for each distinct sense of 'cat'.
> > 5. Is it really a MAT the cat is on?
>
> 6. No, "on the mat" I mean "she is being punished"
> by her boss (since "cat" can also mean
> "prostitute").
Here we just subtract a dimension, treating "on the mat" as a single
predicate. Now we're really on a roll.
> And cfr (Grice) re "formality": "there is this
> demand for a THEORETICALLY ADEQUATE specification of conditions
> which will authorise the assignment of TRUTH-CONDITIONS to suitably
> selected expressions -- thereby endowing those expressions with a
> conventional meaning. Such provision is needed. But the fact that
> [such provision is] needed is insufficient to show that [it is]
> available. We might be left in the sceptic's position of seeing
> clearly what is needed, and yet being at
> the same time totally unable to attain it!" (idem, p.364). These
> conditions, Grice notes should be of course "conditions of truth".
Ah, so here's part of Grice's answer to my earlier question, about
what Griceans count as good "discovery procedures for truth-
conditions." As opposed to non-truth-conditions.
> Cfr. Strawson on meaning and truth: "It is indeed a generally
> harmless and salutary thing to say that to know the meaning of a
> sentence is to know under what conditions one who utters it says
> something true. But if we with for a philosophical elucidation of
> the concept of meaning, the dictum represents not the end but the
> beginning or our task..."
This Meaning and Truth lecture --- er, essay --- keeps coming up. One
thing I noticed in it is that Strawson is always contrasting old-
style "truth-conditional semanticists" with "communication-intention
theorists". Like a rugby scrimmage, with Frege, early Wittgenstein
and Chomsky on one side, and late Wittgenstein, Grice, and Austin on
the other side (Strawson names these people, or people-stages,
explicitly).
I wonder what you think of this. They say Frege was a little guy ---
wouldn't have been much good at rugby --- but I don't know about
Grice.
Also, do these sides make any sense, really? Your posts have often
given me the impression that Grice was fairly cozy with the supposed
enemy here. Not only that, Grice and Austin both seemed to be
believers in incremental progress in philosophy, as opposed to late
W, who was looking for something more like a cure.
> > The set of objects that satisfy 'x is a cat' is not so well-
defined, even in a single idiolect.
> Well, it is in mine. On any ordinary day, in which I awake with an
> unstobbale wish to think in English, by "cat" I mean
>
> 9. x is a cat iff x has phenomenalist properties,
> A: has chromosome structure felis domesticus
> (A being kinda essential)
> B: prototype: it purrs, and answers to "Tibbles".
> C: it miaows.
>
Yikes, you're right! Make that "the set of _mats_ is not so well-
defined" etc. etc.
On the supposed ambiguity of "or":
> Rodrigo does consider this, but I think Rodrigo is right in saying
> that "ambiguity" is out of the question here, even for Grice, and
> I'm sure you'd agree. It's basically UNIVOCITY plus IMPLICATURE,
> ennit.
Well that's the question, idnit. (That's how we say it here in North
Carolina.)
> >I picture the truth-conditions of "the cat is on the mat"
> >as being something
> >like a three-dimensional solid. At the origin you put your
> >stereotypical cat, your stereotypical mat, and your stereotypical
> >situation in which something is on something else.
>
> Right -- and illustrated as in SE Toulmin's book on INFERENCE. I
> love that illustration. Apparently, Anglophones use (2) as
> paradigmatic (as a matter of history) because it featured in
> reading-books in England in the 1930s
> (with emphasis on phonetic/nondyslexic correlation "kaet"//"maet",
> etc).
Now if you ask me, this shows us where we stand on the Turing Test
(subject of an earlier thread). Compare Speranza with a search
engine. No contest. And this is a domain-specific topic search,
supposedly one of the EASIER things for a computer to deal with.
Thank you, JL, I was wondering who else had made use of the attribute-
space metaphor. I knew of some examples in anthropology and
linguistics, but none in philosophy. I think Bas van Fraassen did
some work along these lines but I don't know anything about it.
Regards,
LM Tapper
J L Speranza
explicit reference to Frege vs. Grice, and writes, "Your posts have often
given me the impression that Grice was fairly cozy with the supposed
enemy here."
Yes, but he'd win anyway, as he was a professional sportman. Big bulky
type. Much much much bulkier than Austin. Indeed one of the strongest
philosophers I ever met. And now I'll let Grice give hissself the
relevantly similar impression: He writes:
"My reaction to the [Russell/Strawson, etc] polemic: I have never been
deeply moved by the prospect of a comprehensive and compendious
systematisation of acceptable logical inference, though the tidiness of
modernist logic DOES HAVE SOME APPEAL FOR ME. But what exerts more
influence upon me is my inclination to regard propositions as constructed
entities whose essential character lies in their truth-value, entities
which have an indispensable role to play in a rational and scientific
presentation of the domain of logical inference" (Studies, p.374). This he
said in reply to:
"The Russell/Strawson Polemic. This strand represents what might be deemed
a DIVISION OF SYMPATHY on my part between two DIFFERENT SCHOOLS OF THOUGHT
concerning the nature & content of logic. These consists of the Modernists,
spearheaded by Russell & other mathematically oriented philosopher, & the
traditionalists, particularly the neo-traditionalists led by Strawson in
_An Introduction to Logical Theory_. As MAY BE SEEN, MY INCLINATION HAS
BEEN TO HAVE ONE FOOT IN EACH OF THESE AT LEAST AT ONE TIME WARRING CAMPS"
(p.372).
But this is so to the non-Gricean. Me, as Grice Chair, can authoratively
say that Grice has been pretty clear to me. I said this publicly, and will
say it again. But first a little lexical disgression: In France, you have
to be a "PoMo". i.e. a postmodernist. Not so much in England, less so
Grice's Oxford. Yet, I hold, Grice was a post-modern. For the question in
question, this crucially means that he is _not_ (to be seen as) a
neo-traditionalist -- but he is not a naive modernist (a la Frege, early
Davidson, early Witters, early everybody-but-Grice-and-me either). The
Gricean (essential) charm is going beyond the neo-traditionalists by kind
of accepting all the neo-traditionalists (i.e. Strawson)'s finds in ways
that are kind of consistent with MODERNISM. Impossible to seem, but not to
be. Thus, while the naive non-Gricean modernist would regard Strawson's
discoveries as "metaphysical excrescences", as Grice dubs them in "Logic &
Conversation" (Studies, p.23), Grice spent much of his life analysing them
(excrescences). Thus he writes in "Logic & Conversation":
"On the general question of the place in philosophy of the reformation of
natural language, I shall, in this essay, have nothing to say."
And he does write essay. It's not like I'm editing, or anything...
"I shall confine myself to the dispute in its relation to the alleged
divergences between "and" and "&", etc. I have, moreover, NO INTENDING OF
ENTERING THE FRAY ON BEHALF OF EITHER CONTESTANT [a description of who the
contestants are, he's spent the opening 10 minutes of the essay to] I wish,
rather, to maintain that THE COMMON ASSUMPTION OF THE CONTESTANTS that the
divergences do in FACT EXIST is (broadly speaking) A COMMON MISTAKE..."
Since it is a "common mistake", Grice cannot logically (properly,
reasonably) be dubbed a modernist per se, since he would not like to be
said to be committing the same old trick}, would he...
"AND THAT THE MISTAKE ARISES..."
And this is where Grice (and Griceans make history) -- and I think the
Davidsonians are a kind of Griceans here.
"FROM INADEQUATE ATTENTION TO THE NATURE AND IMPORTANCE of the conditions
governing CONVERSATION. I shall, therefore, inquire into the general
conditions that, in one way or another, apply to conversation as such,
irrespective of its subject matter. I begin with a characterisation of the
notion of "implicature"
((PAUSE TO TAKE A GLASS OF WATER)) He takes.
Later,
Best,
JL
"Fly me to the Moon" is true iff I love you.
Mmm. Rodrigo is getting to get a meaning-liberalist...