From stevenw@hma.att.ne.jp Sun Sep 26 00:37:22 1999 Received: from mxu1.u.washington.edu (mxu1.u.washington.edu [140.142.32.8]) by lists.u.washington.edu (8.9.3+UW99.09/8.9.3+UW99.08) with ESMTP id AAA67200 for ; Sun, 26 Sep 1999 00:37:20 -0700 Received: from hma.att.ne.jp (hma.att.ne.jp [165.76.202.5]) by mxu1.u.washington.edu (8.9.3+UW99.09/8.9.3+UW99.08) with ESMTP id AAA26194 for ; Sun, 26 Sep 1999 00:37:19 -0700 Received: from hamapc07.sizcol1.u-shizuoka-ken.ac.jp (33.pool0.hamamatsu.att.ne.jp [165.76.202.48]) by hma.att.ne.jp (8.8.8+Spin/3.6W-CONS(11/18/98)) id QAA13939; Sun, 26 Sep 1999 16:37:16 +0900 (JST) Message-Id: <199909260737.QAA13939@hma.att.ne.jp> From: "Steven J. Willett" To: classics@u.washington.edu Date: Sun, 26 Sep 1999 16:37:58 +0900 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: Re: Pindar & Short-term Memory Reply-to: stevenw@hma.att.ne.jp X-pmrqc: 1 X-mailer: Pegasus Mail for Win32 (v3.12a) From: Steven J. Willett Thanks to George, Helma and Anne for their suggestions. One of course counts metra in Greek, but that's not possible in English metered poetry, which for all practical purposes only has two forms: iambic and trochaic, neither of which in any length can be considered made of feet or metra. My author's point was that short-term memory, which he regards as basically an acoustic system, is limited to 7 +/- 2 items (capable of recoding into 7 +/- 2 chunks), but English meter, with all its variations, cannot be recoded, so the longest verse line that can be perceived as a rhythmic unit without an obligatory break is about 10 syllables long. The reason stems from his theory of poetic rhythm: he claims that poetic rhythm, as opposed to meter, is the result of the reader's performance skill in reciting poetry so that the metrical pattern and the actual linguistic stress pattern are *simultaneously* accessible to awareness. There is no problem doing this with lines that lack variation. But as soon as we have to recite initial / x x / and / / /, or internal / / /, / / x x, x x / / and the like it becomes a major problem. To render a verse line perceptible as a rhythmic whole, the reciter must manipulate his vocal resources in such a way that the verse line can be completed before its beginning fades out of short-term memory. The means by which the voice can make the abstract metrical pattern and the deviant lexical stress pattern simultaneously available to awareness are such things as grouping, prolongation, late peaking of accent and others (Anne can probably guess who he is by now). My author therefore denies that the only two ways to read poetry are, as C. S. Lewis phrased them, the minstrel's way, who emphasizes the meter over the actual lexical stress patterns, and the actor's way, who emphasizes the lexical stress patterns against the meter. This middle way is supported by a lot of very complex phonological analysis of actual readings. My difficulty is with his theory, not his analysis of performance techniques, which is objectively grounded. I've spent a day in reading about short-term memory, and discovered quite a bit about it that the author either ignored or didn't know existed. 1. Short-term memory is now widely called "working memory" after a model created by A. D. Baddeley and G. Hitch in their classic work _Working Memory_ in Bower (ed.) _The Psychology of Learning and Motivation_, vol. 8 (The London Academic Press, 1974). There they propose that working memory consists of three subsystems: a. the phonological or articulatory loop (AL), which is capable of holding 7 +/- chunks of information or as much information of any type as can be verbally rehearsed in about 2 seconds, b. The visuo-spatial sketchpad (VSSP) whose capacity is still uncleart and c. the executive memory (EM) that controls a. and b. 2. The Baddeley and Hitch model has displaced the older STM model. Considerable new research also indicates that the capacity of working memory varies considerably by individual. Those with less use optimization techniques to exploit it to the full. 3. Subject involvement with or focus on a task seems to increase working memory. 4. Much more sophisticated methods are being used by the NIMH to examine how the brain processes working memory when faced with complex psycho-motor tasks (like piloting a modern jet aircraft) when various sensory inputs overlap temporally. This work is far in advance of what Miller and B&H have done with simple counting and recitation tests. It suggests that working memory is much more sophisticated and flexible than the early mechanically limited model. All this implies that working memory is quite capable of handing not only longer sequences than the iambic pentameter line but also the eight metra upper limit that Maas notes. A Greek chorus and its audience knew in advance that the periods would be grouped into chunks of something--e, D, E and the like--so they would have focused closely on them, which appears to increase memory capacity. And we know in fact that rhythm can act as a focus for attention. Even if we accept the limitation of 7 +/- 2 for working memory, it should have been able to comprehend the longest possible Greek period. Someone wrote me privately about the very long rhythms of classical Indian ragas, which also seem to violate the limit of working memory. Does anyone know if these are articulated into segments, or rhythmical chunks, that would aid working memory? Anne raises the question of whether a stanza has some theoretical limit on the number of words or phrases. I would rather consider a possible limit on lines or periods. Stanzas connected internally by rhyme can be very long without breaking apart or becoming bloated. The longest regular stanza in Wordsworth--who is easily the greatest master of the form in English--runs to 16 lines. He created it for his late and much under-appreciated "On the Power of Sound." It holds together quite nicely and is perhaps the most intricate structure ever devised by an English poet. And that includes Spenser's haunting "Epithalamion," whose stanzas are 18 lines long. In "Lycidas," the flower catalog stanza runs from line 131 to line 164. On the whole, I suspect that a regular stanza much over 20 rhymed lines would not work. If one accepts the colometry of Gentili for Pindar, rather than the modern Boeckhian lineation, then his stanzas can be pretty long. Personally, I'm increasingly sympathetic to Gentili's view, heretical as that may be. That sympathy is partly based on a recent translation I did of Pythian 4 using his colometry. A Greek chorister would have had the music, if not the dance, to help tie everything together. The lyrics were learned as song, not prose. We have to use expedients. When I recite Pindar, for example, I first chant the meter of a strophe emphasizing the quantitative difference between u/- and keeping the metra clearly distinct. Dah = long, de = short. The start of P. 4 then becomes (dah de dah) dah (dah de de dah de de dah) ||. I do that repeatedly until I've internalized the rhythm, then I recite the Greek focusing on the metra. The interaction of syntax with metra and period then becomes much clearer-- at least for me. I might add that I was taught Morse Code using just this method. It works. ================================================ Steven J. Willett University of Shizuoka, Hamamatsu Campus Voice: (053) 457-4514, Fax: (053) 4555 Japan university email: steven@sizcol1.u-shizuoka-ken.ac.jp Japan home email: stevenw@hma.att.ne.jp US email: sjwillett@earthlink.net .