A Shaggy Dog

Speranza

The publishing story of Grice's "William James Lectures" is a bit of a  
funny piece. His WoW:vi -- which contains "The shaggy dog story" was published  
in a rather obscure journal, "Foundations of Language". I was able to find 
it  and always studied it with affection. It was repr. in Searle, Philosophy 
of  Language (Oxford UP) which made it a bit of a classic -- and it's now 
safely in  chronological order as WoW:6. I was recently, elsewhere, 
discussing bits of  this, and it may do here to re-transcribe the Grice quotes 
(selected) as it  applies to The S is P  00 "the dog is shaggy", or That dog -- 
i.e. Smith's  dog -- is shaggy -- where "The S" is a nominal phrase, and "P" 
is an adjectival  phrase. THIS IS THE SIMPLE TYPE of syntax Grice wants to 
restrict the lecture  to. "Smith's dog is shaggy", say. 

He uses 

z ∈ y

and 

a ∈ b

embedded in intentional contexts. 

Grice wants to be able to say how such  a conversational move may shed 
light on intentional aspects of the U's behaviour  vis a vis basic postulates of the 

the theory of action. In particular,
"What  d'you mean, 'shaggy'?". "Hairy-coated." So Grice proposes to 
restrict this --  his stage 6 of his project, or grand plan, to "definite 
descriptor" ("Smith's  dog") and an "adjectival phrase" ("is shaggy").

Grice writes: "We need to  be able to apply some such notion of a 
PREDICATION", or indicatation "of beta  (adjectival) on alpha (nominal)". We 
have thus reached the stage where we have  "two species" of co-relation:
i. R-co-relation, where "R" for REFER,  and ii. D-co-relation (for 
DENOTATE). "We want to be able to speak of some  particular [thing] as an 
R-correlate of alpha, and of "each member of some  class" as being a 
D-correlate of 
beta." (WoW:130).

Grice then goes on to  provide an 'intentional' (basic, rather than 
resultant) procedure: that will  co-relate a belief, say, or a desire, with an 
assertion, or an imperative move  ("Bring that shaggy dog over here!"). There 
is 
a P1, then which corresponds to  the R-correlate. This he just formulates 
as an imperative, "To utter s if U  means the S to be P."  A second, P2, 
focuses on the D-correlate: "To utter  a psi-cross-correlated (cfr. P1 and P1' 
predication of beta on alpha)", and here  again he produces conditions which 
do not claim to be necessary and sufficient  jointly, to the effect that U 
intends psi-cross a particular R-correlate of  alpha to be one of a 
particular set of D-correlates of beta" (WoW:  131).

At this point Grice wants to extend BEYOND a merely disquotational  Truth 
scheme (vide M. K. Davies for an extended approach -- his book with RKP on  
Meaning). It's not just 'The dog is shaggy' is true iff the dog is shaggy.  
Rather he wants to say, "Smith's dog (his example, p. 131), called Fido, is  
shaggy iff the thing is hairy coated. So he needs to work on an equivalence, 
in  an intentional context for 'the dog' to mean, first, 'THAT dog that 
Smith owns'  and second that it is shaggy, with a sort of explication for the 
'meaning' of  'shaggy'. Next, Grice applies basic procedures to create a 
'resultant' one: "to  utter "p", a PREDICATION of beta on alpha ... if U 
intends 
to express a  particular R-correlate of alpha to be one of a particular set 
of D-correlates of  beta". Referring to Smith's dog, unimaginatively, as 
"Fido", Grice proceeds with  a more expansive resultant procedure:

"to utter ... a predication of  'shaggy' on 'Fido'" if U intends to express 
the belief that Jones's dog is "ONE  OF THE SET OF" hairy-coated things 
(i.e. is hairy-coated)". "U has the procedure  of uttering a 
psi-cross-corelated predication of 'shaggy' on alpha if ... [he is  expressing 
the belief re 
the psi-cross "a particular R-correlate of alpha to be  one of the set of 
hairy-coated things." At this point, Grice displays an  interest in something 
like intensional isomorphism when in footnote to p. 133 he  notes the caveat 
that reads as a very fine distinction indeed, "To the  definiens, then, we 
should add, within the scope of the initial quantifer, the  following clause: 
'& U's purpose in effecting that (Ax) (......) is that  (ER') (Az) (R' 
shaggy' x iff x [belongs] to y" --, where he uses the  set-theoretical sign for 
'belongs' (as per below). Grice goes on to refer to  'ostending' here (p. 
134) which may relate to the idea of explicit or implicit  definitions. 

An act of ostension makes explicit what is implicit. We are  providing a 
definition of what a correlation is: under what circumstances we  hold the 
'shaggy' = df. 'hairy-coated'? And in doing thus he goes into a  problem. Does 
'shaggy' mean, simpliciter, as it does, 'hairy-coated'? But then  this 
intentional programme seems to yield, rather, and we do not want that, that  
'shaggy' means, "in U's view unmistakably hairy-coated", so we need a tweak  
there (p. 135). So he opts for "non-explicit" correlations.

Grice  concludes the lecture with a nod to what he will later have as the 
PERE, or  principle of economy of rational effort (in "Reply to Richards"): 
The rule --  IMPLICIT (meaning postulate, say) -- is it subterranean?. Grice 
writes: "in some  sense", "implicitly" we DO accept these rules" (p. 136).  
His P.E.R.E makes  sense of that in terms of potential explicitation of what 
we are _deemed_ to  follow or accept implicitly. No subterranean, thanks! 
(This was later the  polemic of, say, Gricean M. K. Davies in the sequel to 
his book with RKP in the  pages of Mind and elsewhere on 'tacit' knowledge of 
a language and what the  thing is supposed NOT to mean!).

OF COURSE GRICE IS RIGHT IN providing  some charming illustration with 
Jones's dog being hairy-coated (colloquially  'shaggy'). For 'shaggy' is, after 
all, a predicate. And what we are dealing with  here is what I think R. Dale 
in his essay on Grice calls 'a first-order  language', i.e. a predicate 
calculus. Dale indeed plays with the idea of  J-English. English, as it 
comprises only one sentence: "June loves to dance".  Similarly one can imagine 
S-English. S-English only contains: "Fido is shaggy".  Grice provides an 
ostensive definition of 'shaggy' -- too, in the remaining bits  of WoW:VI. Dale 
touches on that fascinating point in the theory of Fodor that  the meaning of 
"Mentalese" is 'circumscribed' as it were, by grasping the  non-logical terms 
(i.e. the predicates) involved. Dale plays with 'cows'  (property of being 
a cow -- cfr. Grice, "the property of being hairy-coated",  _sic_ in WoW:VI 
for those who think he is only a committed  extensionalist). Dale also plays 
with 'dog' and Schiffer's 'schmdog'. One  point to consider here may have a 
historical side to it. I recall having to pass  a seminar -- using Greek 
Loeb -- on ancient scepticism. So I read all the  Sextus, and in looking for 
contemporary literature, came across a review by  Dummett on "The language of 
appearance". The idea that there are noumenal- and  phenomenal-predicates 
as it were.

It would seem that 'shaggy' belongs to  the sort of physicalist (or 
physical, or naturalistic) predicates. A asks B:  "What kind of dog are you 
buying?" "A shaggy one" 'Shaggy' does not seem a  _primitive_ predicate. Grice 
I 
think would hold that 'RED' is a primitive  predicate (discussed extensively 
in his "Remarks about the senses", in WoW).  Then there's 'sofa' that Dale 
also mentions! I tend to think that had it not  been for Strawson's 'mistake' 
in "Introduction to Logical Theory" in finding  formal logic otiose, Grice 
would have explored areas that perhaps interested him  more intrinsically, 
like the philosophy of perception.

Why is it that a  PIROT may need to tell another, "That pillar-box isn't 
red", "It SEEMS red"  ("Causal theory of perception" -- unfortunately the 
section II on 'implication'  not repr. in WoW). Talking of 'red', it was good 
to 
find, online, a reply by  Fodor to Schiffer indeed on 'simple 
compositionality', as it were. The concept  of a 'red flag' I think it is -- 
with Fodor 
arguing how this cannot mean but a  'pirot' being equipped with the concept 
'red' AND the concept  'flag'.

Grice seems to have been charmingly obsessed with things like:  "The 
pillar-box seems red"/"The pillar-box looks red". Why is it that '... looks  
...' 
carries this (what Grice calls) 'doubt-or-denial' implicature? Surely  
cancellable. What else can a red pillar box do but LOOK red? (the philosopher 
of  
perception -- Grice, and, why not? I -- wonders). One little bit about the  
politics behind Grice may be in order before too long. 
A beautiful section  in that ch. iv of Dale's PhD dissertation, online, 
"The theory of  meaning"

Recall Grice: "Fido is shaggy" --- R-correlate: By uttering  "Fido" U meant 
Jones' dog. --- D-correlate. By uttering 'shaggy', U meant  'hairy-coated.'

Dale: "[O]n the sort of theory that Fodor argues for, a  predicate like 
"dog" (pretend that's Mentalese) will have as its meaning the  property of 
being a dog." "But, for well known reasons that property will not  suffice as 
the thing assigned to "dog" by a C[ompositional] M[eaning] T[heory]."  "A 
story by Schiffer is helpful here."
"Ralph came upon a race of creatures  which he thought comprised a 
previously unencountered biological species, and he  introduced the word  
'shmog'  
to designate members of that species."  ""A thing shall be called a 'shmog'," 
Ralph said, "just in case it belongs to  the species of those creatures."  
"Unbeknown to him, however, shmoghood IS  doghood." "Ralph had stumbled not 
upon a new species but a new race of dogs, and  thus the property that 
'shmog' has been introduced as standing for is none other  than doghood."
"But 'shmog' and 'dog' will have to be synonymous for Fodor's  theory since 
they will both stand in the relation that Fodor offers to the same  
property, the property of being a dog." It is sad that Grice focussed on the  
'shaggy', in retrospect -- i.e. on the D-correlate, rather than the R-correlate 
 
[Jones's shmog?], in retrospect, that is. In his sixth (almost there!) 
William  James Memorial Lecture, Grice educated his audience:

"Suppose that for U  (utterer), the following two correlations hold: i. 
Grice's dog is an  R-correlate of "Plato"
ii. Any hairy-coated thing is a D-correlate of  "shaggy". "Given that U has 
the initial procedures that he has, we can infer  that U has the following 
resultantprocedure, to wit:
RP: To utter the  indicative version of a predication of Beta on "Fido" if 
U wants A (Addressee)  to think U to think Grice's dog to be one of a 
particular set of D-correlates of  Beta. Given RP and (ii) we can infer that U 
has: RP2: to utter the indicative  version of a predication of "shaggy" on 
"Fido" if U wants A to think U to think  Grice's dog is one of the set of 
hairy-coated things (i.e. is hairy-coated). And  given the information from the 
linguist that "Plato is shaggy" is the indicative  version of a predication of 
"shaggy" on "Fido" (assumed),
we can infer U to  have:
RP3: To utter "Fido is shaggy" if U wants A to think U to think that  
Grice's dog is hairy coated. And RP4 is an interpretant of "For U, "Fido is  
shaggy" means 'Grice's dog is hairy-coated'". I now provide a definiens which  
may be adequate for adjectival X (e.g."shaggy"): For U, X (adjectival) means  
'...' iff U has this procedure: to utter a psi-correlated predication of X 
on  ALPHA if (for some Addressee) U wants A to believe a particular  
Referentially-correlate of Alpha to be ..." (where the two lacunae represented  
by 
dots are identically completed). Any specific procedure of the form 
mentioned  in the defininens can be shown to be a resultant procedure: if U has 
(2) 
it is  inferable that he has the procedure of uttering a psi-correlated 
predication of  "shaggy" on Alpha if for some A U wants A to belief a 
particular 
 referential-correlate of Alpha to be one of the set of hairy-coated 
things, that  is, that for U, "shaggy" means "hairy-coated"". More formally: By 
uttering  V, U has correlated "shaggy" with (and only with) each hairy-coated 
thing iff  There is a Reference such that U effected by V that there is an x 
such that  R(shaggy, x) iff x belongs to y (y is a hairy-coated thing) andU 
uttered V in  order that U effect by V that there is an x..." And this is 
insufficient as it  stands".
(p.133).

For certainly, why wouldn't "one want to know that  my dog is shaggy unless 
she wants to beautify it?" INTERLUDE: A shaggy dog story  is defined by the 
OED as "a lengthy tediously detailed story of an  inconsequential series of 
events, more amusing to the teller than to his  audience, or amusing only 
by its pointlessness; also shaggy dog yarn, etc  --  1945 D. Low in N.Y. 
Times Mag. 4 Feb. 40/1 -- The logical lunacy of  `Shaggy Dog'. -- 1946 Coll. 
Shaggy Dog Stories facing p. 1 -- Stories of the  Shaggy Dog variety are 
essentially tales to be told rather than read. -- 1947  Beat Apr. 6/3 -- Here's 
one of my favourite `shaggy dog' stories. -- 1952 A. R.  K. Barnard in A. Red
man Somewhat `Shaggy' 4 -- The comparatively recent type of  story-the 
`Shaggy Dog' yarn. -- 1952 Koestler Arrow in Blue i. viii. 68 -- The  people of 
Budapest have a peculiar shaggy-dog kind of humour. -- 1958 Listener  16 Oct. 
623/1 -- It was a shaggy-dog story about a small-town worthy who shams  
madness to avoid paying bills. -- 1972 P. Ruell Red Christmas xi. 102 -- He  
seemed to be in the middle of an autobiographical shaggy-dog story.

I  have elsewhere, indeed in most fora I have been able to push Grice (as 
if he  needed my pushing, but I enjoy doing it -- I push other things too, 
like  wheelbarrows, if that´s the word, not when I´m selling cockles in Dublin 
(which  I don´t, but when I garden at the Villa Speranza). Anyway, a recent 
query  elsewhere was looking for experimental evidence of the 
´procedural´thing about  things. The interesting thing, if I may be redundant, 
is that 
the query appeared  to be from someone who is well aware of the literature of 
what is called L2 --  not your mother tongue, but not your father tongue, 
either. Instead, the querier  was looking for evidence _in_ the mother tongue. 
This is confusing. The mother  tongue while etymologically IS my mother´s 
tongue, I guess she (my mother) will  have NO idea. So I take ¨mother 
tongue¨to mean, by metonymy, my tongue. And what  does "procedure" mean in my 
tongue? It was fun to have the most active of my  sexual years -- the ten of 
years -- involved in the Gricean analysis of  propositional attitudes and 
communicative reasoning alla whatever. Not only was  great at bars, A: Do you 
come 
here often? B: Only in the mating season.
--  and cfr. my The content of content -- but I was able to _think_. For 
ten years I  was looking -- true I was writing my PhD thesis too, and needed a 
LOT of field  work, for the right word for various things, and discovered 
it all, one dark  night, in Grice´s

A Shaggy Dog Story Grice tells, then, in WoW:vi.  It involves a dog which 
is shaggy. For Grice, the shaggy dog story is a way of  illuminating us 
against the evil (he then thought, and vice versa) influence of  Chomsky. For 
Grice, there are two types of procedures: basic and resultant. It  is _both_ 
basic procedures, as he calls them, which are involved in the Meaning  of 
¨Words¨. This is not to say that this is "procedural words". For the basic  
procedures, for Grice, involve the incorporation of content.
Notably, what is  the content of "dog"? What is the content of "shaggy", 
when we say "The dog is  shaggy."
We need a basic procedure for ¨dog¨, and we need a basic procedure  for 
"shaggy". Both fall within the same nominal category, while Grice takes  
provisions to deal with "shaggy" as not "the shaggy one", but "shaggy" qua  
adjective. However, the procedures are similar. Here Grice applies work by  
Strawson and himself on reference, in "Individuals" by Strawson and elsewhere.  
¨The dog¨involves the referential bit of the utterance; "is shaggy" involves 
the  predicational bit. It´s these functional categories that tell us then 
what a  noun versus an adjective is. I won´t inverse the terms here, since 
"The sh*g is  doggy" makes so much sense, and according to my friend, J. M. 
Geary, much MORE  sense than the original sentence used by Grice. WoW indeed.

So this  leaves us what another "use" of "procedural" which I find easy to 
digest, or  understand. Grice seemed to have problems with some "procedural" 
procedures. I  would think that what is elsewhere, or in some elsewheres, 
called "procedural"  is merely esultant procedure in Grice. For consider, 
"¨The dog is shaggy and the  sh*g is doggy". Here the resultant procedure 
involves "and". And, for some  reason, "and" was the first resultant procedure, 
as analysed by Grice that  struck Chomsky. Anyone familiar with his 1966 
Theory of Syntax will find a  reference in the name index which is bound to 
amuse him: Grice, A. P. That _is_  our Grice. The "A" possibly meaning, 
"Aristotle" -- and "P", Plato? Anyway, for  Grice, "and" involves various 
procedures, all resultant. The first is the  "&" thing. "The dog is shaggy & 
the sh*g 
is doggy", where "&" is  defined in terms of the truth-value table. The 
second resultant procedure is  more complicated, for it involves a FLOUT of the 
maxim, Be orderly. So, the  following dialogue may ensue:
A: The dog was shaggy and the sh*g was  doggy.
B: Are you saying you found it out in that order?
A: No. Matter of  fact, I did feel the sh*g was being doggy
well before I percieved that it  _was_ a shaggy dog.
B: Still, I don´t see what kind of sh*g you were  expecting
from a dog other than the one that struck you so
deeply.
A:  Dunno. Is this the Griceian in me?
Now, experiments, made on oneself are --  painful. I am going to experiment 
with myself, whose mother tongue involves  "shaggy" and "dog" and proceed 
to see if I see a distinction between "shaggy",  "dog", and "and". And I 
don´t! I mean, "and" can become a noun. ¨His ands bore  me, especially as 
follwing "try", "try and"¨". Here "and" is "mentioned", not  used, but in 
metalogic, "p & q" may become the metadiscourse at some higher  level, where 
"and" 
is the focus of our attention, and hence it achieves  "content". After all, 
Grice WILL talk of something like a Fregean SENSE [this  should interest 
McEvoy, since he (McEvoy) senses sense senses] iin Reply to  Richards, for 
things like "not", "and", and in WoW:ii, he refers to the  "meaning" of things 
like "to" and "of" as being just as tricky as "or". But then  I am not an 
innocent informant. Chapman has a good one on this when she  mentions, in her 
bio of Grice (Macmillan, 2006) that Grice would use Tim´s and  Karen´s (his 
children) playmates as naive informants, for things like "Nothing  can be 
green and red all over". What kind of procedures, basic and resultant  does 
this 
"judgement" of synthesis a priori involve? Years, later, and in  Lancaster, 
too, Nigel Morley-Bunker reported a same experiment -- cited by G. R.  
Sampson in his book in Experimental Linguistics. People don´t _know_ what they  
are talking, is mainly Sampson´s conservative point. T. Wharton who teaches 
at  Sussex should approach Sampson and get the real answer about this! I 
first  learned of Morley Bunker via Sampson´s more philosophical "Making Sense" 
 (Clarendon), which amused me, since deals with the complete protocol for 
the  experiment. Something like that I assume is the way to go with the 
original  query. And in any case, going through it, has allowed the Grice Club 
to 
expand  on distintions made by this mastermind the club is dedicated to.

Grice on  shaggy (five-step semi-inferential sequence) après WoW:364  ---- 
Since  Grice (WoW: 364) is careful to use 'feature' (which he distinguishes 
in WoW:vi  from 'item') it's best to state his semi-inferential sequence 
thus (his  shaggy-dog story): STEP 1: It is, speaking extensionally, general 
practice  (merely) -- which can be with myself --, to treat 'shaggy' as 
signifying  hairy-coated. Or 'runt' to mean 'undersized person'.  STEP 2: It 
is, 
now  speaking, with Carnap, INTENSIONALLY too, general boring practice to 
treat  'shaggy' to signify (or mean) 'hairy coated' (or 'hairy-coated' to mean 
shaggy).  STEP 3: It is generally, as a matter of fact, rather 'de iure', 
acceptED that it  is LEGITIMATE (sc.) acceptABLE) to treat 'shaggy' as 
signfiing or to signify  'hairy-coated'. STEP 4: It is legitimate (ceteris 
paribus 
acceptable) to deem  and treat 'shaggy' to signify or as signifying 
'hairy-coated'. STEP 5: "'shaggy'  signifies 'hairy-coated'". Then there's 
Grice on "
α ∈ β" (WoW:VI:133n1) -- repr.  in Searle and scaring Chomsky
One may wonder: how are the concepts to  which  the uppercase words refer 
actually represented?
I pointed out  that Grice will NOT engage in this silly practice, and use 
rather, if he must  ('must' understood in the preterite, here -- no such 
trick in English), by  variables, using the more distinguished phi and khi, or 
alpha, alpha', beta, and  beta', as in the footnote referred to in title 
which scared Chomsky when he read  it in Searle, "The philosophy of language" 
(Oxford, 1971) and which him fodder  for his boring 3rd lecture against Grice 
at Oxford (The Locke  Lectures).

α ∈ β

So Grice wants to just stick with any 'feature'  that belongs to an 'item'. 
We have four features in his long shaggy-dog  story:
the dog that Jones calls 'Fido' -- alpha
the dog that Jones owns  ----------- alpha'
shaggy ---------------------------- beta
hairy-coated  ---------------------- beta'
When we think of co-extensionality of features,  we mean that all items 
which have feature beta, also have feature beta' (All  shaggy things are 
hairy-coated things). We are concerned with Grice on 'beta',  here because for 
the 
sub-mechanism of referring the logic is slightly different  from the more 
basic sub-mechanism of 'predicating'. So what that 'infamous'  footnote that 
does display Grice as the extensionalist he once was reads: "The  definiens 
suggested for explicit correlations is, I think, insufficient as it  
stands." To see that Grice managed Harvard University Press to have this as a  
footnote is miraculous. He goes on in same self footnote:
"I would not wish  to say that if A deliberately detaches B from a party, 
he has thereby  correlated
himself with B, nor that a lecturer who ensures that just ONE  blackboard 
is visible to EACH member of his audience (and to no one else) has  thereby 
explicitly correlated the blackboard with EACH member of the audience,  even 
though in each case the analogue of the suggested definiens is satisfied."  
His ability to bring in the most disparate illustrations is just genial. It  
brings the whole abstract field he is plowing into something that even the  
dullest student or reader should understand. It's like he is saying: 'No 
way you  can defend yourself by saying that I did use convoluted examples.' He 
goes on:  "To have explicitly correlated X
with EACH MEMBER [i.e. each ITEM  that is a member. JLS] of a set K, not 
only must I have intentionally  effected that a particular relation R holds 
between X and all those (and   only those) items which belong to
K [feature. JLS], but also my purpose or  end in setting up this  
relationship must have been to perform
an act."  Imagine if you wanted to say this -- of such an importance -- in 
just a  footnote! Grice goes on:
"to perform an act as a result of which there will  be some relation or 
other which holds between X and 
all those (and only  those) things [or items. JLS] which belong to K." And 
here is the important  bit, where he plays Zermelo-Fraenkel: "To the 
definiens, then, we would ADD,  within the scope of the initial quantifier, the 
following clause." And what does  the clause look like? It looks LIKE this. In 
fact it IS this: & U's purpose  in effecting that (x) (......) [six dots. 
JLS] is that
 
 
 (ER') (z)(R' 'shaggy'z <-> z ∈ y (sc. y is  hairy-coated)).

Crystal clear, right? To understand him fully we need to expand on the NEED 
 of this footnote, but I bring it to the forum because it's the ONLY place  
(surely for abbreviatiory purposes) that Grice cares to use that rather 
infamous  concept, ∈, that he will later criticise when he sees 
"Extensionalism" as a bête  noire. Of course he KNEW he wasn't an 
extensionalist (enemy of 
intenSions) even  then, because he IS using at least intenTional terms (like 
'end' or 'purpose'),  and he is well aware of the problems of quantifying 
in: "I [did say] at one  point that intenSionality seems to be embedded in 
the very foundations of  the  theory of language". Grice allows, though, that 
"it may be possible to  derive ... the intenSional concepts which I have 
been using from more primitive  EXTENSIONAL concepts." (WoW:137). So, the 
"R-correlates" (rather than  "D-correlates") he has in mind is for the 
'subject' 
line of the canonical: Fido  is shaggy, rather than Whiskers is shaggy -- 
where "Fido" has as a R-correlate,  what Grice has as "Jones's dog", rather 
than "Smith's cat" (Grice uses "Smith's  cat" elsewhere in this essay, 
WoW:VI). He does not spend much time at all on  alpha-correlates, since he is 
more 
notoriously going with Frege in D-correlates  then, or beta-type correlates. 
I.e. extension of the predicate 'shaggy' itself.  Here he is treading the 
ground of compositional semantics. If Tarski was into  disquotational: "Snow 
is white" iff snow is white. Grice wants to go inside the  'propositional 
complex' and want to say that "shaggy" means 'hairy-coated'. It  is via 
D-correlates that the thing is achieved. He allows for an extensionalist  
reading 
in set-theoretical terms, and an intensionalist reading in terms of the  
'property of hairy-coatedness', and so on. Now, it is interesting that for 
Grice  those D-correlates feature WITHIN a 'phrastic' as it were. "Tactful" is 
another  predicate he considers (along with "shaggy" -- if not applied to  
'dog'). Suppose that we define 'tactful' as considerate.
So, he wants to  say that:
"Smith is tactful", "Smith, be factful!", "that Smith be tactful!",  etc. 
These utterancesintroduce the same D-correlate. But obviously there is a  
variancy there. In the 'assertion' (for which he uses Frege's symbol,  
judgement-cum-content, /-), 'tactful' is predicated of Smith. Also in "Smith, 
be  
tactful" but not within the scope of an 'assertion' operator, but what Grice  
would symbolise, naively, as "!", and where, notably, the psychological 
attitude  involved will be a 'desire' rather than a 'belief'. In the analysis 
of 
 D-correlata themselves, Grice uses variables for psychological attidudes 
(psi),  variables for moods (*) and the basic notion of a neutral 
'intention',  rather.

Interestingly, Grice was concerned with Positivism as it was then  known in 
Oxford back in the 1930s. Recall that Ayer was more or less of Grice's  
generation (Ayer born 1911, Grice, born 1913). So, while Grice will be of 
course  identified, post-war, 1945, with Austin), back in the 1930s, when Grice 
was a  student at Corpus Christi, he knew what was going on. There was a lot 
of  activity (positivist and other) with Ayer, Hart, Austin, McNabb, and 
others,  meeting at Hampshire's rooms in All Souls, as I recall -- or perpahs 
Hart's  rooms. In any case, Grice always reminded that he never was invited 
to those  'seminal' advances of Oxford philosophy (vide Berlin, "Austin and 
the early  origins of Oxford philosophy") because he had been born on the 
wrong side of the  tracks. Corpus Christi catered for "Midlands" types like 
Grice. Whereas All  Souls, and Hart in particular gathered around him a more 
'pretentious' group.  Ah, in any case, as I think M. Arnold once said, and he 
knew: "Only the poor  learn in Oxford". Or something. It is good to review 
Grice's earliest  publications for traces of 'positivism', or 'empiricism' of 
a radical type. His  analysis of "I"-sentences ("I was hit by a cricket 
ball") look empiricist enough  (in terms of mnemic states, _Mind_, 1941), and 
he has unpublished contemporary  stuff on "Negation and privation" (negation 
in terms of 'ignorance') and  "Intention and disposition". But soon enough 
Grice took all his introspections  for valid, and never questioned again 
their validity. Positivism as a creed  ceased to be a dogma, as it were.

And so on.