"She was poor but she was honest", said Grice -- but Toulmin disagreed

J has recently reminded of the grand oeuvre by this Cambridge philosopher (originally) Stephen Edelston Toulmin, who died a couple of years ago.

When he did (die) I re-read all the obituaries -- AND "Uses of Argument". I may have dropped the odd note in the Club. One page of "Uses of Argument" (shouldn't the title be "Uses of ArgumentS", since he is proclaiming a pluralism?] I focused was when he discusses:

logical form

and he has Strawson and Grice in mind with things like

She was poor, but she was honest
and her parents were the same
till she met a city fellar
and she lost her 'onest name.

----

We are never told who "she" was. It seems the predicate analysis is

FEMALEx

As if we shouldn't care what her 'proper name' (versus 'pronominal' definite description) was. Actually, _hers_ is a long story, which I memorised once by reading "Other men's diamonds" or something. A book on poetry by an Indian exile, or something (A British military type who lived idly in India).

---

As J notes:

"but" should read as "and"

But the idea is that there is an attending second-order logical subform to that:

"She was poor"
AND
"She was honest"
AND
"I want to contrast her poverty with her honesty" and say "but".


Toulmin indeed mentions that 'but' SHOULD count as a 'connective'. He finds it 'obscene' that Grice has dediced to eliminate it. Grice's elimination came fullest in "Causal theory of perception", as per online edition, NOT the one Grice reprinted in his WoW.

Toulmin mentions ANOTHER feature:

"Most poor girls ARE NOT honest".

"Most".

This he pairs with 'but'. He notes that if 'but' the logicians think they can deal without, ditto for 'most'.

This is an oversimplification. What 'most' requires, at most, is a threshold. This was studied by followers of Geach's pleoretetic logic: Altham, notably ("The logic of plurarity"). They are Cambridge people and Grice would ignore them.

But it MAKES sense to have 'most' as a defined quantifier.

Grice and the Griceans have rather focused on 'numerical quantifiers':

She, this one-and-only girl, was poor and she was honest"

Or:

"I met three girls which were honest AND poor".

'three girls'
'two girls'
'one girl'

----

This is like 'most' but more 'definite'.

Quine analyses -- he is a religious type --

"The twelve apostles were drunk" (the Last Supper).

in "Methods of Logic"

He wants to formalise that as

(12x) DRUNKx

and I agree. But Quine spends like, er, what is it, 3 pages? on how (12x) has to be understood qua quantifier. He draws from Frege.

When Boulinck wrote his PhD on "Gricean numbers" he noted that:

"My ball itches"

implicates that one ball has been removed from the speaker. In more careful Gricean terms,

"my ball"

indicates that the utterer cannot have Two balls. Else he would have uttered:

"One of my balls itches".

Unfortunately, some female linguists have objected to this uniqueness-based analysis of posessive definite descriptions. But what do they know?

If I say I (or someone) has two balls, I am also saying that I (or that someone) has one ball. For how can you have two balls if you don't have "at least" one ball?

This had Atlas and Horn suggesting that:

"I have a ball"

has, as per logical form,

"I have AT LEAST a ball"

Atlas added,

"plus a closure clause, 'and at MOST one ball'"

This is the meaning of 'the' ball.

'One ball'.

"The three stooges" complicates things: here we have 'the' which entails uniqueness, and yet, to contradict that, it's _three_. We have a combo of an iota operator and a numeric quantifier.

---

Viva la forma logica!