Critiquing the Critic: Dan Schneider Responds

Add to FacebookAdd to DiggAdd to Del.icio.usAdd to StumbleuponAdd to RedditAdd to BlinklistAdd to TwitterAdd to TechnoratiAdd to Yahoo BuzzAdd to Newsvine

To be honest, my first reaction is to be flattered.

That said, I still find his initial essay ludicrous and stuffed with fallacious arguments.  He made many points in response to my own assertions (he lambasted  Carlo Parcelli), but most of them are tangential to a definition of meter.  For example, he points out that I got the title of his essay wrong, true, and that there are typos in my posts, also true.  (There are also typos in his response, but does anyone really care?) He accuses me of sending him a possibly virus ridden hate-E-Mail which I don’t remember and which he, conveniently, can’t produce. (I’m calling that one, false.) He also takes issue with how I characterized his arguments. I don’t blame him, but I stand by those characterizations. However, none of this has anything to do with meter itself.

On to his assertions concerning meter.

In the entirety of his response, he provides only two (2) specimens to support his arguments.

In answer to my rhetorical question, ‘…what metrist has ever asserted that meter is composed of just two discrete stresses and that, furthermore, these two stresses are precisely the same no matter the context?’, Dan writes the following:

I will now disprove such by using two definitive texts. The first is from Webster’s Universal Unabridged Dictionary (1964). In reference to meter (meaning poetic metrics, no other usages of the term):

1. (a) rhythm in verse; measured, patterned arrangement of syllables, primarily according to stress and length; (b) the specific rhythm as determined by the prevailing foot and the number of feet in the line; as iambic meter; (c) the specific rhythmic pattern of a stanza as determined by the kind and number of lines.

I don’t see how Webster’s helps Dan’s case. Notice that Webster’s does not assert that meter is composed of two discrete stresses or that they are the same no matter the context.  Dan’s original assertion was that:

“meter is the theory (claiming origin by several cultures) that spoken language consists of 2 primary vocalizations of a sound- i.e.- stressed & unstressed.”

And this definition, as a very general one, isn’t necessarily wrong. But he then calls that definition into question by writing that:

In fact the dualistic notion of mere stressed & unstressed sounds is- in practice by its many proponents- almost always so loose as to be meaningless anyway, as metrics should really redefine its definitions as greater & lower stress(es) (with a plenum of in-betweens), since (obviously) a truly unstressed syllable would be silent.

In other words, (according to Dan) the  “plenum” of stresses available in an accentual language contradicts the notion of “2 primary vocalizations”. But it only contradicts if one assumes that the “2 primary vocalizations” can’t be relative (or widely vary in relation to each other). Schneider’s argument only holds water if the “2 primary vocalizations” are discrete and always the same. But, as I wrote, no metrist, to my knowledge, has ever asserted the same (only, ironically, Dan Schneider). All “theories” of meter recognize that stress is relative and therefore recognize a “plenum” of stresses. They recognize that English is an accentual language, and that within the language’s “plenum” of stresses, one stress will always be relatively strong and one will always be relatively weak.

Webster’s definition in no way bolsters Dan’s contention that meter doesn’t exist. Nowhere does Webster’s definition limit meter to two discrete stresses which are always the same. The Webster’s definition  rightly asserts  that meter is a pattern of stresses (English) or lengths (Latin).

What is especially curious about Dan’s example is that Webster’s defines meter the way I do(!) and, most importantly, doesn’t question its very existence.

On to Schneider’s next example:

The oldest and most important device of Verse form, meter selects one phonological feature of lang. (stress, pitch, length) and reduces it several levels or degrees in ordinary speech (3 or 4 levels of stress; high, mid, and low pitch; various durations) to a simple binary opposition (‘stress’ vs. ‘unstress’; ‘level’ vs. ‘inflected’ pitch; ‘long’ vs. ‘short’) which may be generalized as ‘marked’ vs. ‘unmarked’.

This is from the New Princeton Encyclopedia of Poetry and Poetics. Dan rightly mentions that Princeton’s overview covers several pages. However, he glosses over the implications of this concession by writing:

This is very important to note, because from the start of my essay through its end, I am the person arguing that meter is a reductio ad absurdum, it is not real, and it reduces human speech to a false binary opposition. Princeton proves I’m right on that score, and says so in black and white.

(Never mind that Dan’s own example from Webster’s contradicts his claim that meter is “a reductio ad absurdum” – which is to say, it doesn’t exist.) Well, OK Dan, but, as you intimated, Princeton says a lot more in black and white. It also writes:

The traditional view had always been that m. is an arbitrary pattern imposed on words — that, as Gurney put it, “metrical rhythm is imposed upon, not latent in, sppech” (1880). It seems indubitable that meter is in some sense a filter or grid superimposed on langauge. But 20th century linguistics has shown convincingly that many aspects of poetic form are merely extensions of natural processes already at work in language itself.

One page later, and after much exegesis to support this contention, Princeton closes the section by writing:

But modern metrics also holds that strong syllables outside ictus are “demoted” and weaker syllables under ictus “promoted” under the influence of the meter. Promotion of weak syllables under non-ictus weights and slows the line, adding power. Demotion of stresses under ictus gives a quicker and lighter line. This is not a purely metrical mechanism, it shadows normal phonological process by which alternation of weak and stress, and strong and stronger, is effected atomically in polysyllables.

Apparently Dan either couldn’t be bothered to read this far or conveniently chose to ignore this portion. Princeton, in fact, not only disagrees with Dan but recognizes the binary stress pattern of the English language as a “normal phonological process”. And, by the way, did I mention it does so in black and white? Not only that, but Princeton rightly points out, as I have, that 20th Century linguistics has shown convincingly that many aspects of poetry are “extensions of natural processes already at work in language itself.” The next time Dan claims to be a man of science, take it with a grain of salt.

Dan then goes on, at some length, railing at my characterizations of his argument. None of which, curiously, supports his claim that meter doesn’t exist. He repeatedly refers back to  Websters and Princeton, neither of which support his argument.  Among other things, he writes:

This is really amazing. First, VP spends the bulk of his essay claiming that my claim that meter is a fallacy is wrong, then he cites a study (naturally, the links do not work)…

I just checked the links. They work.

Without, apparently, reading them, he both dismisses and reinterprets the science (which, did I mention, he didn’t read).

More importantly, Dan never counters the example of an artist like Eminem. As I wrote above, Rap is a “thumping example” of accentual and accentual syllabic verse.

Dan quotes Princeton out of context, ignores science, and glosses over 8 Mile. He then closes:

As I implied in the piece VP quotes, I was a mediocre formalist. Note the past tense. I am a great poet, formally and in free verse. There are poems of mine that scan perfectly, according to metric nonsense, but not because I was following metric dictates, but because any well musicked poem will, given the reductive aims of meter, scan well. It’s what is in them that matters.

So, according to Dan, meter doesn’t exist but, by gosh(!), when he wants to, he writes meter with genius!

Not that all meter isn’t “nonsense” (but his poem scans perfectly). He’s not following metrical dictates  (it’s just that a “well musicked poem” does the same thing), and not that it’s not nonsense (but it scans well). Never in the annals of “seminal” essays has a more self-contradictory paragraph been written.

jester

I guess that’s what happens when you try to have your cake and eat it too. At the very least, readers shouldn’t be taking advice from a man who claims meter doesn’t exist, then hurriedly, as an afterthought, asserts that he nevertheless writes meter with genius. Makes you wonder who the idiot really is, doesn’t it?

By the way Dan, I prefer – Fool.

In a play like King Lear, he’s the only one who lives.

About these ads

One response

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 518 other followers

%d bloggers like this: