WHEN THE faithful gather to argue, a persistent issue is whether poets should write metrically or not.

About a year ago, Diane Wakoski wrote in The American Book Review that a new conservatism is emerging in American poetry in the form of a mindless return to traditional meters. Her view is that believers in the Walt Whitman heritage have some fighting to do, because proponents of "form," many of whom know very little about it, are urging it on young writers who find "rules" a comforting shelter from anxiety.

More recently, in The New Criterion, Brad Leithauser writes, from a more conservative stance, that the free-verse revolution is approaching exhaustion. Much of his argument is based on the difficulty of creating variations on a rhythmic pattern, when, as is the case with free verse, the underlying pattern is not predictable.

This is a recent example of a dichotomy that has long characterized American poetry, ever since it was realized that American writers must use something very close to the mother tongue in establishing their independence of it. Some poets, like William Cullen Bryant, were satisfied with English; others, like Whitman, saw a need for a uniquely American idiom. It is a vast oversimplification, but a useful one, to say that the major split among American poets has descended from that disagreement.

Dividing all American poets into traditionalists and pioneers, however, presents difficulties; Emily Dickinson is probably the greatest of the poets who are hard to discuss in these terms, but there are many. To treat Robert Lowell this way, it is necessary to break his career into phases; he himself could not see why, if he could write both free and metrical verse, he should give up one for the sake of the other. I can't either. But being unable to write metrically is easier than choosing not to.