"Probably" as sentence adverb

--- by JLS
----- for the GC

OSHER DOCTOROW has posted an interesting entry on "Probably Grice". Grice considers the sentence adverb function of "Probably".

He compares it, rather artificially, with "desirability". That is, he compares "probability" with 'desirability'. He wants to say, apres Davidson, that 'probability' works for 'doxastic' or belief-contexts, whereas 'desirability' works for 'practical' contexts. The conclusion of an inductive inference is "Probably, c", where "c" is the conclusion. The conclusion of a PRACTICAL inference is, "Desirably (?), c".

But I should need to work more on that.

From:

www.cs.cmu.edu/afs/cs.cmu.edu/project/cmt-55/lti/Courses/7...

1. Probably he can rely on my support.

2. He probably can rely on my support.

3. He can probably rely on my support.

But:

4. ?He can rely probably on my support.

5. ?He can rely on my support probably.

Why?

--- The author of that document claims that that is a consequence of 'probably' being a SENTENCE adverb rather than your common-or-garden verbal phrase adverb:

For the case of a verbal phrase adverb, like 'completely', it's the 1 and 2 that sound odd:

1. *Completely he can rely on my support.

2. *He completely can rely on my support.

3. He can completely rely on my support.

4. He can rely completely on my support.

5. He can rely on my support completely.

---

As the author puts it:

"A sentence adverb, like 'probably', must be immediately
dominated by a node labeled Sentence. On the other hand a common-or-garden Verbal phrase adverb, like 'completely' must be immediately dominated by a node labeled Verbal Phrase."

-- and yes, Grice DID trees! (as Chapman found them in the Grice Papers at Bancroft, and reported in her bio of Grice, "Grice" (Macmillan, 2006).

Glad to learn O. Doctorow presented the thing in 1981 to the Philosophy Dept. at Berkeley.