Michael:

"
Steele uses the stress numbers only to indicate that, in an iambic line, not
every stressed syllable is equally stressed and that not every unstressed
syllable is is equally unstressed. There are two chapters on metrical
variation and substitution. As he says in a post to Eratoshere ( a public
forum so I think it's fine to quote it here):

"We’re accustomed to fourth-generation vers-libristes who imagine that to
write in meter is to forsake personal rhythm. But another tendency--one to
which we who are interested in meter are perhaps sometimes prone--is to
forget that there are all sorts of rhythmical potentials within meter. One
does not have to write ka BOOM, ka BOOM, ka BOOM, etc"
"

I can partly accept that (and my post was both bad-tempered and
off-the-cuff). But it's surely a truism that metrical stress and speech
stress aren't identical -- indeed (as I think Steele implies) rhythm is the
very result of the tension between the two, otherwise you get today (as
Candice pointed out) or in the 1550-1570s (when things went +really+ to the
bad) the good old de DUM de DUM de DUM. The question is how you represent a
metrical analysis of a poem. Accepting the difference between (natural)
speech stress and metrical stress (ictus), it's possible to present the
metre of almost any English poem written in the syllable-accent tradition,
in a way that makes sense, with only two signs. I'm inclined to see even
the use of the half stress marker -- X = unstressed, / = full stress, \ =
half stress -- as a bit of a copout.

When it comes to the Tragger-Smith four-stress business (and I think David
Bicumshaw might disagree with me on this), the metrical result is horrendous
overcomplication, based on an idea (that English has four major degrees of
stress) that was chucked out the window long ago.

I say this with a degree of personal venom, as trying to get my head round
scansion via Tragger-Smith (in the sixties) did my brain in. The crucial
text here is, I think, Wimsatt's "The Concept of Meter: An Essay in
Abstraction". Wimsatt makes the central point that (metrical) stress is a
matter of contrast. Once that's in place, you don't +need+ the
four-degrees-of-stress stuff.

But I'll try and come back to this tomorrow, when I've had a chance to read
Steele's article, and follow up on the places you suggest.

For which, much thanks. (I have to -- guiltily -- confess I was partly
teasing you into just such a reaction. Saves me finding things out for
myself. Sorry -- blame it on the time of night [or morning] when I was
writing.)

Although I +do+ still feel that Steele gives a rule-driven rather than
ear-driven approach to meter.

Robin

"
The full discussion is here:
http://www.ablemuse.com/erato/ubbhtml/Forum19/HTML/000038.html
BTW, the site is so heavily java-scripted and otherwise over-designed that
it is nearly impossible to use without a broadband connection and a fast
machine.

But I do think that Steele, along with some of the other New Formalists, is
too iambi-centric, as I have already indicated back-channel to Robin.
"