[Richard Meyer]

Any discussion of scansion usually raises interesting and differing opinions. This particular thread prompted me to poke through my bookcase and retrieve a text from my university years: Sound and Sense: An Introduction to Poetry by Laurence Perrine (Harcourt Brace, Third Edition, 1969). This classic poetry handbook was originally published in 1956 and has gone through many editions. I don't know if it's still in print, but for decades it served as a bible for students undertaking a serious study of poetry.

Here are a few excerpts from Chapter 12, "Rhythm and Meter":

Scansion is at best a gross way of describing the rhythmical quality of a poem. It depends on classifying all syllables into either accented or unaccented categories and on ignoring the sometimes considerable difference between degrees of accent. Whether we call a syllable accented or unaccented depends, moreover, on its degree of accent relative to the syllables on either side of it.

Scansion is not an altogether exact science. Within certain limits we may say that a certain scansion is right or wrong, but beyond these limits there is legitimate room for personal interpretation and disagreement between qualified readers.

Finally—and this is the most important generalization of all—perfect regularity of meter is no criterion of merit.

Perrine's remarks about scansion are consistent with those I've read by other eminent scholars in a variety of texts, and we would be wise to heed their insights when composing metrical poetry lest our verses become petrified.

Richard