Regarding:
   "The unpainted house stood on the hill, seeming deserted."
   "The cowboy walked down the street, trailing a roll of toilet paper."

I think the commas are generally needed in such sentences, there being a
pronounced pause before the participles that needs to be signaled to the
reader.

On the other hand, there is a potential for ambiguity.  For example, read the
following sentences, the first with a pause at the comma, then the second
without a pause:

   "The candidate walked down the street, shaking voters' hands." [pause at
comma]
   "The candidate walked down the street shaking voters' hands." [no pause at
comma]

I read the first sentence as roughly equivalent in meaning to "The candidate
walked down the street, and she was shaking voter's hands while she did so."
I read the second sentence, without the pause, as equivalent to "The candidate
walked down the street so as to shake voters' hands."   The participle in the
first modifies the noun "candidate."  The participle in the second seems to be
adverbial and modifies the act of walking.

Dick Veit


Michael Kischner wrote:

> "The unpainted house stood on the hill, seeming deserted."
>
> "The cowboy walked down the street, trailing a roll of toilet paper."
>
> The participle phrase at the end of each of the two sentences above is
> nonrestrictive.  Nonrestrictive participle phrases are set off by commas.
> Therefore, the two participle phrases are set off by commas.  But
> something in me wants to strike those commas.  I gues my reasoning would
> be that in these cases the participle phrases are set off by intervening
> words and are therefore not in danger of seeming restrictive as they would
> if they were right next to the nouns they modify.  (Whether they really
> modify those nouns is another question; they seem so adverbial.)