238 Advances in spoken discourse analysis
the syllables we have designated 0, something that we can achieve in this
case by appeal to the distinction between protected and unprotected vowels.
Both -ized u- and -pon a- have one vowel which is usually reduced to something
like / / and one which is never so reduced. On this basis, we can recognize
the additional degree of ‘weight’ we need and re-analyse the line as
| 20 | 10 | 10 | 30 | 30 | 10 |A similar treatment of the proposed reading of the Auden stanza is carried out
on the assumption that it comprises mainly a combination of iambs and anapaests.
This is only one among a number of ways in which this stanza—and indeed
the whole poem—could be mapped into the ‘feet’ of traditional metrical
theory. I do not wish to advance arguments in favour of doing it in this way
rather than another. The point is rather that, whichever way we divide up a
line into feet, reasons can be found in the reading I have suggested for
identifying one syllable within the foot as the ‘weightiest’, and thus for
providing a firm basis for the ‘stressed/unstressed’ distinction.
In saying this, I may well seem to be flying in the face of a widely held
view about the nature of the metrical basis of a poem. Crystal sums up this
view as follows:
There is almost total agreement that metre, however defined, should not
be identified with the psycho-physical analysis of utterance, as displayed
in the reading of a text. Metre is held to be an abstraction, in some sense,
and is not to be identified with performance.
(1975:105)Many people have found it hard to accept the notion of an underlying psychological
pattern of stresses that owe their definition to no regularities in the spoken
version of the poem. And accepting it seems to be a necessary step towards
going along with the further claim that a poem, as realized in speech, does
not usually coincide exactly with this underlying pattern; it is because the
e