[Jonathan James Henderson]

Some very good posts so far (as I expected), so thanks everyone for contributing, but I do want to explicitly address Rose's interesting points:

Originally Posted by Rose Novick

A few truths about meter:

• it's made up. it's fake. you can't have firm and fixed opinions about Correct And Proper Scansion without being a crank.

Very true. Meter is an attempt to abstract the most common patterns of natural speech and make them into fixed patterns for musical/rhetorical effect. It's artificial by its very nature, but all art is made up of some amount of artifice.

Originally Posted by Rose Novick

meter is most useful (I find; ymmv) when it is kept minimal, sharply separated from rhythm. when too much of natural rhythm is stuffed into meter, the flexibility of meter is lost.

I don't entirely follow here; perhaps some examples would help. It seems perhaps you're saying that meter should be a spice that's thrown in rather than the main dish from which variations or "natural speech" deviates from? This is kind of how I think of Whitman. We know Whitman for innovating the use of free verse, but there are many lines in Whitman that scan perfectly and stand out by contrast from the rest. The famous "When I Heard the Learn'd Astronomer" ends on a line of perfect pentameter after a string of Whitman's trademark anaphora.

Originally Posted by Rose Novick

it follows from this that, in reading a poem, one should read its rhythm but not its meter. its meter is a ghost that haunts it.

In that old thread I linked to I talked about "the ghost of meter" haunting lines in other rhythms as well, though I'm not sure I would agree that we shouldn't read meter at all. It really depends on the poem. Some poems are meant to be read in meter, other in different rhythms with meter as a background. I've experimented with both and both have different potentialities for expression and poetic effect.

Originally Posted by Rose Novick

corollary #1: any metrical system that recognizes more than two degrees of stress (stressed, unstressed) is baroque and unhelpful.

Generally agree, but my one caveat is that multiple degrees are probably more accurate for mapping natural stress patterns, and I think keeping in mind that there's a broader range than two degrees also helps us in thinking about how to categorize those two degrees when there's ambiguity. Perhaps an example later.

Originally Posted by Rose Novick

corollary #2: relative stress is much more important than absolute stress, because relative stress points to the abstract pattern underlying the line, while absolute stress points to the variation within that pattern.

Absolutely, but, continuing from the above, relative stress can have more than one relation ranging from the foot it's a part of to what follows it to a more general pattern.

Originally Posted by Rose Novick

corollary #3: with exceedingly rare exceptions, spondees do not exist in English language poetry. relative stress rules nearly always resolve them into either trochees or iambs. one consistent class of exceptions is the {u u / /} pattern, which I prefer to think of as a special two-beat foot, the "double iamb".

This I don't agree with. Unless someone can point me to a scientific study that says people never put roughly equal stress on two subsequent syllables I'm going to trust my ears that people do.

Originally Posted by Rose Novick

mastery of meter is then the ability to create variation within this consistent abstract pattern. some of this comes from outright substitutions; much of it comes from playing with relative stress. again, precisely this sort of play is lost when the distance between meter and rhythm is shortened.

Agree.

OK, here's the most illustrative example I can think of that I also used in the previous thread, though I'm going to take a slightly different tact here (CAUTION! What follows is a grade-A example of meterbating. Consider this a warning that what follows is pedantic and honestly worth nobody's time including mine):

Bright star, Would I were steadfast as thou art

I honestly don't know of a single line of IP that could be naturally read in a greater variety of ways. Here's just some of them:

Bright star, Would I were steadfast as thou art
Bright star, Would I were steadfast as thou art
Bright star, Would I were steadfast as thou art

I think you mix and match variations on these, such as keeping the opening a spondee, but making reading "steadfast" and "though art" as natural trochees, or keeping the IP reading but not stressing "as". This is one line where I'd love to get 1000 random people to read to study the different ways it would be stressed. It's also the first line of a poem, so we can't even be clued in by any prior context. We could clue ourselves in by reading the rest of the poem, but that kinda feels like cheating; how should we read it on a first reading? Just follow the traditional "rules" of IP? But, as I've said, we know poets used substitutions, even if some were more stricter about when and what kinds, though I think those restrictions loosened up more in the 19th century compared to the 18th and 17th. Further, if we read this in IP, with or without substituting a trochaic inversion of the first/second foot or a spondee for the first foot, I think it reads really awkwardly to stress "as" as opposed to either "fast" or "thou."

This is a good example of how context helps to determine stress because humans naturally want to stress more important words in sentences. In grammatical hierarchy, nouns and verbs are usually first, with adjectives and adverbs second, then prepositions, then articles and conjunctions. However, this hierarchy runs up against English being a stress-timed language in which stresses should be (roughly) equal distance from each other. Often times those two paradigms conflict or present ambiguities, and this is a great example. Here, "as" is an adverb, which might take stress after following an unstressed syllable (like "fast" could potentially be), but given that it's followed by a pronoun it almost never would when spoken naturally. But if we speak it naturally that leads to a metrical problem as it was never acceptable to end a line of IP on a trochaic inversion, which would mean that last foot would be a spondee.

If we scan the first three feet we run into similar-but-different issues. If we make "Bright star" a trochee, especially after the caesura of the comma, it's difficult to argue that the verb "would" wouldn't take stress; but then we begin the poem with two trochees, which makes us want to completely ruin the IP by using a weird trochee/dactylic hybrid meter (second reading above). That works for "natural speech," but not for meter. It would seem on metrical principles we could rule that out. The grammatical hierarchy doesn't help either because we have an adjective/noun/verb/noun/verb. The only one of these that one would naturally demote would be the adjective, but it's at the beginning, and I defy anyone to say "bright star" and tell me with a straight face they don't sound silly. Further, it would seem "stead(fast)" has to be stressed. So, what do we have for sure? That "bright" and "stead" must be stressed; that leaves "star, would I were." If we stress "would" there we run into the problem of ruining the meter completely, as I said above, so it seems "would" shouldn't be stressed. However, if "would" shouldn't be stressed it would seem "star" should be just due to stress-timing paradigm; and when you put a caesura after "star" and the meter tells us we shouldn't stress "would" that leaves putting the stress on "star" and then "I" since there's nothing left.

The above is a long justification for why my preferred reading here is actually the one with the three spondees. Besides the above, there's an artistic/poetic reason that I analyzed in the old thread, so to pretentiously quote myself:

Originally Posted by Jonathan James Henderson

The major thing I--and I think most--would notice here are the three pairs of consecutive stressed syllables: "Bright Star," "stedfast," "thou art." Once one picks those out I think you also notice a strong semantic correlation between them; it's almost the poem in little right there. In fact, inverse it and you have "Thou art stedfast Bright Star." So what's left over? Well, you have "would I were" between the 1st and 3rd spondee and "as" stuck between the 3rd and 4th. I think it makes sense to group these together to form an amphimacer and catalectic unstressed foot.

What's the use of this? Well, besides pointing out the semantic relation between the spondee set, I think it contrasts well with the two "leftovers:" "would I were" places the stressed "I" between an "island" of unstressed syllables while poor "as" is left out on its own. The differing rhythms of these two groupings contrasts wonderfully with the strong, consistent spondee pairs, seeming to enact the very thing it's describing: the speaker, like the words associated with him, are not as 'stedfast' as the 'bright star / thou art' subject.

My condolences to anyone who actually read all that. Have a smile: