For a poet who is attempting what many consider to be the most radical of recent experiments with language--the Dream Songs--John Berryman's poetic is ruthlessly functional. "The ordinary modern reader is sound asleep...Eliot understood this very well, in the Waste Land; it's necessary to kick him, otherwise he won't perform, and if he doesn't perform there's no poem. Because a poem is a reciprocal kind of action between the writer and the reader. No reader, no poem. It's like unperformed music, Bach scores lying in manuscript for hundreds of years. That's what a poem unread is like...Not too good...So you get the reader awake, and once he's awake you make sure he stays awake. And you tell him all of things he doesn't want to hear. But you tell them to him in a very dulcet voice, so that on the whole he's pleased with them."

Remarks like this are always startling when you're used to hearing delicate technical circumlocutions from poets, discussions of style at the expense of content. We have very thoroughly withdrawn from the theory of messages, Berryman no less than most; but the man is fully as anxious to see people grasp what the poem is about as he is to alarm and confuse them with unusual language. He began writing, he says, "as a burning trivial disciple of the great Irish poet William Butler Yeats," but Yeats "could not teach me to sound like myself (whatever that was) or tell me what to write about." What drove him away from Yeats, through periods of Eliot and Auden, and finally into the ambiguous arms of Anne Bradstreet, was in part, perhaps, a violent dissatisfaction with having nothing to say. "When I finally woke up to the fact that I was involved in a long poem, one of my first thoughts was: Narrative. Let's have narrative, and at least one dominant personality, and no fragmentation--in short, let's have something spectacularly not the 'Waste Land,' the best long poem of the age..."

This fully articulate dissatisfaction with the limits of the lyric form led him to adopt a rather insignificant historical figure and build an imaginary relationship with her. "Why did I choose to write about this boring high-minded Puritan woman who may have been our first American poet but is not a good one?" In a way, he says, "she chose me." "Your deep subjects, I'm talking about attempts at major poetry, not lyrics or meditative poems, they come and take hold of you... The point is to throw as much light as possible on her, and apparently at some point the decision was made to throw light on the twentieth-century poet, and to let this be explored in a dialogue."

The result is a long narrative masque in darkness between Anne and the poet. After the first four stanzas, the poet interrupts her monologue in three places. Finally she dies, and stays with him as the imaginary presence from the American past we have know her to be all along; then he turns to face the very real and serious world of the twentieth century. The poem leads us through her child-marriage to Simon Bradstreet, her crossing on the Arbella in 1630, her writing, the birth of her first child, her bout with smallpox, her religious difficulties, the expulsion of Anne Hutchinson from the colony, and her later life, all over-shadowed by the image of an angry God and quivering with the rhythms of guilt and insecurity. What emerges is a cerebral, passionate, deeply religious and thoroughly female Anne Bradstreet--innocent and worldly, orthodox and impatient with orthodoxy's drab practitioners, in love with her husband and with more than that, quietly violent in her sexual self-expression--the "always heretic" that is the poet in any language at any time. Whether this persona bears more than a verbal similarity to her prototype is a question better not asked, for Berryman's Homage is even less like history than the average historical novel.

The poem is no wearying topical exercise in verse-biography. The stanza-form, Berryman's invention, is difficult; the language is astonishing.

--It is Spring's New England. Pussy willows wedge

up in the wet. Milky crestings, fringed

yellow in heaven, eyed

by the melting hand-in-hand or mere

desirers single, heavy-footed, rapt,

make surge poor human hearts. Venus is trapt--

the hefty pike shifts. sheer--

in Orion blazing. Warblings, odours, nudge to an edge--

The choice of words, the dislocated syntax, the archaisms like "trapt" or the frequent use of accents--all show a taste for the bitter, explosive, tactile qualities of words that few poets demonstrate in greater intensity than Dylan Thomas. Occasionally the language slides off into bluster, or mist:

faintings black, rigour, chilling, brown

parching, back, brain burning, the grey pocks

itch, a manic stench

of pustules snapping...

--Be kind you, to one unchained eager far & wild...

But Berryman gets away with rhetorical excesses that would be preposterous in any other modern poet. Why? For one thing, the fact that Bradstreet is a long poem checkmates most of the poetics we have carefully engineered to deal with short lyric poems. Rhetoric is permissable when you're speaking in persona and pointing to something very great and very vague: the past. Demands of economy and tactile immediacy are more than satisfied by Berryman's aggressive, spiky, studied choice of words.

Constantly you feel that Berryman is daring to say something oceanic, then returning to the concrete with a thump or a blast. And the man is absolute master of his materials, which points less toward facility in the use of stanzas, rhymes, and meters--like Auden's--than toward an utter control over all the possible sounds and meanings of each word.

--Wan dolls in indigo on gold: refrain

my western lust. I am drowning in this past.

I lose sight of you

who mistress me from air. Unbraced

in delirium of the grand depths, giving away

haunters what kept me, I breathe solid spray...

But the old question of causes--formal and final--arises. Many critics have reacted unfavorably to the whole project. Why do it at all? Like a Tiffany vase without a mouth, what's it for? Speaking of conception as well as of language, Stanly Kunitz remarks "Berryman is tempted to inflate what he cannot subjugate." The effort to conquer an old emminence grise from the American past may be thought of as a false one, a spurious gesture of research toward a subject that is just not real.

As far as am concerned this is the only objection that can be raised to the poem. To write about yourself extrinsically and comically, as in the Dream Songs, is one thing: to sublimate your urge for self-projection in the posture you take toward a deliberately irrelevant subject is something else. Berryman is a little like Max Beckmann in his habits of constant self-depiction (which differs from self-revelation in that the latter is usually true), for running through Bradstreet is the image of the twentieth century poet in a tense pose of self-indulgence. But the worst that can be said of the poem is that it errs slightly in the direction of a naive, mannered Romanticism.

Two answers can be made. First, every lyric poem is rationalized to some extent on its model. All sorts of syntactic oddities are permitted in sonnets, for example, because they fulfill certain expectations, while the aesthetic order of a poem like Lycidas, random as it may seem today, is regular in terms of its formal cause: the genre. A poem on a "deep" subject--a poem as catholic in its intent as Paradise Lost--has no one model, but uses and subsumes many. Berryman had no model for the Homage to Mistress Bradstreet. Harte Crane's The Bridge ("a set of lyrics") was out; Edward Arlington Robinson and Robert Browning were uncongenial ("I admire them, but I dislike them"). Eliot's Prufrock and Waste Land are disjunctive and impressionistic where Berryman's effort is continuous and expressionistic.

In the abrupt, resonant dialogue that forms the midsection of Bradstreet, Berryman was influenced by Anna Karenina and by Saul Bellow's novel Augie March, which he had just read in manuscript: "very ambitious, totally unlike most modern novels. It threw me the feeling that if I appeared to go outside the ordinary sort of business, that would be all right." The absence of any clear poetic precedent forces the reader to make a major revision of his conventional expectations.

The second answer runs deeper. Talk about Berryman keeps coming back to subject-matter, to content rather than form, to purposes rather than techniques. Just as the subject of Bradstreet, in the deepest sense, is Bradstreet, the Dream Songs are "about" Henry by God, and if Berryman's public descriptions of Henry are cagey, he is no more willing to divert the audience with coy adversions to his own skills or state of mind. For a long time poetry in this country has been working with an arsenal of familiar tools--"effects," "devices" and "meanings"--all largely technical considerations. Such a situation has often led to the spectacle of hordes of young men writing very well about very little. The reader is only encouraged to read the work, and locate all the sources of his response there.

Berryman's work simply cannot be read this way. None of the great creative violations of convention in literary history can be. Literature redeems itself by going the limit, by taking the same sort of risks that Berryman found in Augie March in the form of "an inquisitiveness, let's call it that, so extreme that it becomes a way of life, a tempting, a touching of all the boundaries..." No one, of course, can take risks who doesn't know the rules, but what is perhaps most impressive about John Berryman is his unwillingness to define expertise in purely technical terms. In a review of Hugh MacDiarmid he once wrote "a poet is to prove that he is not squeamish, as a poet (his private attitudes being nothing), by being absolutely responsible for his material and its psychological and spiritual employment, while technically he is absolutely independent of both. Flourishes will not do at all." Check.