To say that John Berryman's poetry is controversial is to state the obvious. Few poets—and none that I can think of since Pound—have aroused such varying and often violent responses. A. Alvarez has written that one either loves or loathes Berryman's work, and that seems to be the case. This should not be a matter of great surprise. For one thing, Berryman's work—especially The Dream Songs—is difficult, difficult in a way that most recent poetry is not. At a time when American poetry is moving away from the Eliotic modern, Berryman's work seems to be a throwback, if a self-conscious one, to an earlier age, bringing poetry, to paraphrase Williams's remarks on The Waste Land, back into the classroom. (p. 3)

[There is] a tendency in much of Berryman's mature work which in the best light might be called reticence, in the worst light gamesmanship—a certain toying with the reader. Berryman spoke in interviews of his "trade secrets," and implied that there were secret meanings in The Dream Songs that would challenge scholars for years. His work is highly allusive, and he obviously enjoyed this aspect of the difficulty of his work, as lines from The Dream Songs indicate:

           I feel the end is near
           & strong of my large work, which will appear,
           and baffle everybody.

This is partially a way of poking fun at the literary establishment, a blague directed at the academic community with which Berryman enjoyed a distinct love-hate relationship (compare his portrait of the MLA convention in 77 Dream Songs). More frequently, though, the allusiveness of Berryman's work is intended to serve a serious function. Berryman was an enormously well-read man, and in his work, especially The Dream Songs, he used that knowledge both to create a world and to give the effect of a man ransacking our culture in an attempt to find a way of living in that world. This is not a new use of allusion, obviously; it is firmly in the modernist tradition. The effect of both the playful and the serious uses of allusion, however, has been harmful in some cases, irritating some readers and obscuring more important aspects of his work. The Dream Songs is not an intellectual game, but Berryman managed to convince several critics that it was. "After working out the double crostics," wrote Mary Curran of 77 Dream Songs, "there is only the feeling of dissipation and boredom."

Let us add a further element of difficulty before going on: style. The words "Berryman's style" call up a host of adjectives: idiosyncratic, eccentric, quirky, willed, perverse, in ascending order of disapproval. Style: a use of language that so impresses itself upon the mind that subject and meaning are all but lost. Randall Jarrell's early criticism, in a review of The Dispossessed, was prescient: "Doing things in a style all its own sometimes seems the primary object of the poem, and its subject gets a rather spasmodic and fragmentary treatment." The quotation, selectively applied to Berryman's work, even his major work, is accurate—most accurate when the verse is weakest, for the assumption behind the criticism is that style is not intended to hide meaning but to express it, an assumption with which I agree. (pp. 3-4)

What is centrally important about The Dream Songs is that [to quote Martin Dodsworth in The Survival of Poetry] "despite all appearances, these poems are not hermetic in intention; they seek to expose the poet's most remote level of consciousness, to present the reader with the kind of experiences Berryman believes and feels to be behind the superficies of life, but which we are ordinarily content to ignore." This emphasis on experience in the poem proves it to be anti-Symbolist, despite its surface resemblance to a work in the Symbolist tradition…. The poem appears to be modernist, but the style is not explicable by ordinary means. Therefore, the style must serve some other, extraordinary purpose. In [Peter] Dale's view, it serves to obscure a lack of substance. [See CLC, Vol. 2.] In Dodsworth's view, the language of the poem is intended to emphasize the importance of experience. What Mr. Dodsworth calls "the superfluous oddity" of Berryman's style "is calculated, I think, on the supposition that what is superfluous to the conventional demands of literature may establish a means of personal expression and communication with the reader, something that cannot, as yet, be contaminated by the student's desire to incorporate all literature into some vast and inoffensive system of conventions."

"Superfluous oddity," in the example that Mr. Dodsworth discusses (dream song 33), refers principally to the use of "ah" and "ha," especially their emphatic use as rhyme words. It is not necessary to posit a new poetic to come to terms with this oddity, however. In song 33, the ahs and has are the outbreak of Henry's own powerful emotions aroused by the story of Alexander and Clitus; both the violence and the form it takes of chopping are aspects of Henry's own character that he can barely control, as song 29 makes clear. They are thus not superfluous here; they are rather an indication of the enormous difficulty Henry is having in telling the story. The oddity comes principally from not reading the song as part of a larger whole. I must hasten to add here that I don't think every sample of distorted or mannered language can be justified, or seen as adding to the poem. When such distortions cannot be justified, it seems to me, the language has failed; I do not think that it is necessary (or reasonable) to make of these failures a success. Berryman was a daring poet who pushed the language to its limits and sometimes beyond the limits. (pp. 5-6)

I do not think The Dream Songs is a radical break with the Symbolist tradition any more than I think it is a radical break from Berryman's earlier poetry. There are some important differences between Berryman's work and the work of the Symbolist and modernist poets who wrote before him, and there are some crucial differences between his major work and his earlier work; but there are also important similarities, ones that should not be ignored. Berryman was, in fact, always a modernist poet, although at different stages in his career different aspects of that modernism were of primary importance to him. The history of his career may indeed be seen as an attempt, or a series of attempts, to discover which aspects of what we have long called the tradition were appropriate to his own vision. And the two things which come immediately to mind, for they were the two areas that were of greatest importance to Berryman's work, are language and the image of the poet.

The origins of Berryman's style have been the subject of a great deal of speculation, largely in terms of a search for a presiding influence. It is not unreasonable to look for such an influence. Berryman's early work was so heavily indebted to the work of Yeats and Auden, and Berryman was always such a bookish poet, that we can expect to find important stylistic influences on his major work as well. Hopkins is most frequently mentioned in this context, although Berryman denied that Hopkins had much influence on him. Cummings has also been mentioned, as have Dylan Thomas, Crane (Hart and Stephen), and Stevens. I would add Milton and especially … Tristan Corbière, whose work Berryman was reading during the time of his own greatest stylistic development in the mid-1940s. (pp. 6-7)

Berryman's style evolved not just through the reading and influence of a few other poets, but in response to an aesthetic implicitly accepted, but never explicitly expressed, the aesthetic of Symbolism, with its insistence on creating a new language that, in Rimbaud's words, "would be of the soul, for the soul, containing everything, smells, sounds, colors; thought latching onto thought and pulling." In this context, I mean by "aesthetic of Symbolism" not the idealistic foundation for that aesthetic, but the necessity it imposes of expanding the language, and the technical liberation that it implies. There is a certain amount of the idealist-aesthete in Berryman, especially … in his early work, but in his major work he is, as Mr. Dodsworth quite correctly points out, concerned above all with living in the world. The language that he found, which we could well describe as containing everything—even some things which some Symbolists would have excluded—served as a means of describing and living with the absurdity and pain of that world.

Berryman's style—but not his interest in style—changed greatly over the course of his career; his thematic concerns changed to a lesser extent. His subject was, as he described Pound's subject matter, "the life of the modern poet," and virtually everything he wrote revolved around that theme. Berryman's image of the poet, though, changed through the years. The poet figure of the early poems is usually sensitive, committed, alienated, and above all aloof…. (pp. 7-8)

There is much in this image of the poet that would remain as a part of Berryman's later work. Berryman's later poet figures—especially, of course, Henry—tell our story, indeed live our story, but remain rejected by the rest of us, apart from us, poètes maudits who find it increasingly difficult to find anything that makes the suffering worthwhile.

Berryman's style, as I have noted, derived from the Symbolist aesthetic, with the very important difference that it was intended to create a language that would enable us to express and thus survive, perhaps even triumph over, the absurdity and pain of our lives. Berryman's suffering poet figure is the result of a similar transformation of an element of the Symbolist tradition. He undergoes the suffering that Rimbaud said the poet should undergo, hoping to find not so much the "unknown" or some ideal realm but a strategy for living in a world filled with all-too-real difficulties…. The attempts to deal with [these difficulties], the emergency adaptations of the Symbolist notions of the poet and of language recorded in Berryman's work—especially Homage and The Dream Songs—are what make Berryman's poetry memorable. The problem is living in the world, a problem that became in Berryman's work increasingly insistent, increasingly intense. (pp. 9-10)

[If] one were looking for Berryman's "confessional verse," [Berryman's Sonnets] would be the place to start. It is clearly unreasonable to assume … that Berryman entered the post-Life Studies "confessional movement" with 77 Dream Songs in 1964. On the contrary—throughout his career, from the earliest Yeats imitations on, Berryman experimented with various ways of turning personal experience into art. At the same time he was constantly aware of the need for his poems to "leap into myth," to move out from the personal to the general. And in the Sonnets, which is one of Berryman's most personal works, there is a similar attempt to "leap into myth." The original problem, as he makes clear in the preface, was not the affair, but the more general problem of "whether wickedness was soluble in art." In the Sonnets he attempted, through adopting an old convention, to put his specific and particular character in a general situation—to portray, as he portrays in his other work, the life of the representative man, the modern poet who bears the burdens of his culture.

Nevertheless, Berryman's Sonnets cannot be said to be a completely successful attempt, precisely because these poems do not "hunt the whole / House through," at least not very convincingly. When Berryman restricts himself, as he does in most of the Sonnets, to "fidelity-infidelity" and the ramifications of that theme, mostly Freudian, in the culture, the poems are generally effective. But Berryman is not successful—as he is in his later work—in expanding this central theme to include broader considerations. Sonnets which attempt to relate his particular situation to political or metaphysical themes are generally flat and unconvincing. This is not because the form of the sonnet sequence is not flexible enough to handle a variety of themes—it is—but because Berryman was discovering themes that he didn't yet know how to handle. (pp. 42-3)

Homage to Mistress Bradstreet is in many ways Berryman's central work, the breakthrough that fulfills earlier promises of genius and makes new promises for the future. This is true of both the language of the poem and its thematic elements. In Mistress Bradstreet Berryman brought together several concerns that he had touched on in his earlier work and which would be of increasing importance in his later work: loss, rebellion and submission, the importance of the family. He also brought up, through the voice of Mistress Bradstreet, that difficult word which he had mentioned briefly and to no apparent purpose in the Sonnets, but which he would use frequently in later works: God. Mistress Bradstreet is an ambitious poem, but it attempts less than The Dream Songs and, because it is more coherent, is more successful. (pp. 47-8)

It is tempting to say that Homage to Mistress Bradstreet is Berryman's Sonnets writ large, but this would be a bit misleading. Let us say, rather, that in the Sonnets Berryman discovered an important theme, adultery, which was closely related to his major thematic concern, the relation between the poet and his culture, but—perhaps because he was too close to the actual experience—Berryman didn't quite know what to do with it. Berryman's Sonnets is an attempt to generalize his experience but, finally, not a successful one. Berryman was unable to include everything he wanted to say. What he needed was a form that would allow both the lyric voice of the sonnet sequence and the scope in subject matter of the epic. What he needed, indeed, is what every ambitious poet since Wordsworth has needed—a form that doesn't exist. Homage to Mistress Bradstreet and The Dream Songs are Berryman's two attempts to create a narrative form that would include both history and the personal voice. In Homage to Mistress Bradstreet he faced the problem of form by yoking two forms together and creating what might be called a love poem containing history, a poem in which the muse and heroine are one…. (p. 48)

[Anne Bradstreet's] only claim to Berryman's interest would seem to be that she was the first American poet, and thus the stuff that myths are made of. But Berryman, unlike Hart Crane, whose Bridge has often been compared to Mistress Bradstreet, is not primarily concerned with creating American myth. Mistress Bradstreet is concerned with more particular relations—relations between individuals and between individuals and society. (p. 49)

As the [writing of The Dream Songs] grew and the years wore on, the line between Berryman's poem and Berryman's life became at times a very fine one. At times, indeed, individual songs simply became vehicles for the expression of Berryman's peeves about minor aspects of his life. If Berryman had trouble with the mail, for example, Henry complained about it. Trivial experiences can make good poetry, of course, but frequently in The Dream Songs—especially in the latter sections of the poem—they do not; they remain records of trivial experiences. There is, it seems to me, no question that this reduction of Henry and his experiences harms the poem. At the same time let me emphasize that this doesn't happen so often that The Dream Songs dwindles into the expression of a series of merely personal grumblings. (p. 61)

From evidence in the poem it seems quite clear that Henry is intended to be a representative figure. He is yet another of Berryman's poet-heroes, taking "immortal risks" and bearing "in the fading night our general guilt." Berryman's subject is the life of the modern poet, and that subject is not only personal but also national and metaphysical; it is absolutely necessary that Henry be a broad enough character to be able to contain all these concerns. The obvious precedent here is Song of Myself, with the important difference that in The Dream Songs the poet is united with the rest of us not through a transcendental vision, but through suffering. In song 242, which begins "About that 'me,'" Berryman describes a woman visiting Henry's office and breaking into tears. He cries with her and at the end of the song says "I am her."

Saying this in one song does not make it true for all the songs, of course, but it clearly states Berryman's intention. When this characterization of Henry is successful, we are given a picture of a comic poet-hero, taking upon himself our suffering, and bodying it forth in song, saying, with Jeremiah in Lamentations, "I am their musick." When it does not work, as in "Henry's Mail," mentioned above, we are merely given a picture of Berryman. (pp. 61-2)

The Dream Songs is different in form from Homage to Mistress Bradstreet, but in theme it is similar. In many respects The Dream Songs is of longer and more complex treatment of the themes Berryman had explored in the earlier poem. Like Mistress Bradstreet Henry rebels against his environment, his family, himself, and God. Like Mistress Bradstreet Henry ultimately finds in his family the values necessary for survival in and, indeed, triumph over, a hostile world. At the center of Henry's world, though, is an element that figures only peripherally in the earlier poem—loss…. (p. 63)

To a certain extent, the conditions of the modern world are the cause of much of the despair of The Dream Songs. The world of the poem is not very different from the world of "The Dispossessed," in which "no hero rides. The race / is done." This fear that the end of the race is imminent, that we are at the end of a long decline, and that "Man has undertaken the top job of all, / son fin,"… occurs periodically in the poem. It is viewed in relation to the other deaths and losses of the poem, fulfilling the pattern. But if Henry's difficulties are more intense than they would have been in the past, at bottom his problems are not new, nor are they merely the product of twentieth-century life…. The problem is less twentieth-century life than it is life itself, the attempt to live right in a world that makes it very difficult…. (p. 65)

The title of The Dream Songs allows Berryman a great deal of linguistic freedom…. The dreamer is an inveterate punner and player with language. But "good poets love bad puns," and we do not really need the title to describe (or defend) the language. The title refers in the first place to dreams as we have learned about them from Freud—the welling up of repressed material in a disguised fashion in sleep. Dreams are caused by a conflict that is of great importance in The Dream Songs, the conflict between instinctual needs and the needs of the culture. Indeed, dreams are the battlegrounds of that conflict, for the instinctual material which would conflict with the needs of the culture is allowed in dreams a harmless release. So the title refers to both the purpose and the method of dreams. (p. 79)

Randomness, although it is certainly present in The Dream Songs, is more the lack of structure than a structural principle. At the same time, as tantalizing as the thought might be, there seems to be no tightly knit but hidden or "secret" structure to The Dream Songs….

I think that this could safely be said about the other books in the volume and about the volume as a whole. It has a beginning and an end and a general development from beginning to end; it is loosely unified by theme, subject matter, imagery, and, of course, Henry, but many individual songs could be juggled around without affecting that loose unity. (p. 81)

The Dream Songs is not a completely successful poem. Homage to Mistress Bradstreet, more compact, more coherent, is a more successful work. There is not enough narrative in The Dream Songs to hold such a long work together, and what narrative there is is often confused. The methods of organization which are essentially lyric—recurrent images and language—are not developed consistently enough; there is just enough development to force us to read the poem as a whole, but not enough for us to see it as a totally coherent whole. The poem was too available to the daily events of Berryman's life, perhaps. The attempt to turn his life into myth succeeds best when Berryman abstracts himself from Henry, by putting him in a fictive environment—up a tree, or dead, or even simply bored—or by treating him ironically or humorously—showing us Henry hungering after a woman in a restaurant, or suffering the effects of a "truly first-class drill" in a dentist's chair. Unfortunately, this attempt does not always succeed. But despite its faults, the poem remains a considerable achievement; Henry, his dream world, and his language remain in the mind after reading, indelible. (pp. 85-6)

[Love & Fame] is structurally more satisfying than Berryman's major long poem, The Dream Songs, for there are few obtrusions into the narrative, and the story told is fairly clear. That story is not very different from the stories told in Berryman's other works. The book describes the development of a poetic sensibility struggling against both outer and inner obstacles. In his earlier work, though, much greater attention is paid to the outer obstacles—the mad culture the poet finds himself a part of—than is paid here. In Love & Fame there are some poems about American society, but what is most important about that society is Berryman's isolation from it. (pp. 88-9)

Love & Fame is far from being the aesthetic botch that most reviewers found it to be. But this is not to say that it is Berryman's finest work. There is a problem with this "wiping out" strategy, for after the sensibility of the early poems is wiped out, there is very little left. There is very little stylistic attraction in the first two sections. Had the young Berryman been presented as a slightly more despicable character, the volume would have been improved. One will return to the unpleasant Berryman of "Damned" as one returns to some of Browning's monologues, for the monstrousness and the unattractive vitality of the poem; but most of the early poems present a sensibility without vitality. Young Berryman is not evil enough to attract, and, once or twice read, the anecdotes lose much of their power. The poems are necessary—the second half of the book depends upon them—but they are too often simply straw men, put up in order to be knocked down. The structure of the volume is satisfying, but the early components of that structure, by and large, are not.

In Delusions, Etc., published posthumously but seen through proof by Berryman, there is a thematic unity of a sort, but no clear overarching structure. When poems fail, and quite a few do, they do so on their own, without any narrative compensation. It is true that section 4, the only section with a title, is called "Scherzo," but I would hesitate to claim a symphonic or sonata structure for the book. Individual sections are fairly well unified, but beyond that the book is more of a collection than were Berryman's earlier volumes. The full title of the book is Delusions, Etc. of John Berryman, but the delusions are not Berryman's alone. The chief delusion is man's thinking—the pride he takes in his rational faculties. Berryman's work had always been romanticist, and here that romanticism is quite blatant, but curiously mixed: at times Berryman sounds like Emerson, at other times like Melville.

"I don't try to reconcile anything," Berryman had approvingly quoted Ralph Hodgson, "this is a damned strange world"; and that is how Henry had ended, looking to Ralph Hodgson and not trying to reconcile "things and the soul." But the conflict that he had ended for Henry did not end for himself. Berryman did try to reconcile things: reason and faith, science and religion, the existence of evil and the notion of a benevolent God, and Delusions, Etc. is largely a record of that attempt. Hence, the Emersonian-Melvillian contrast: in some poems Berryman does seem to have reconciled these things, in others he recognizes, with pain, that he cannot. (p. 94)

There are … some very weak poems in this collection. The book has been compared favorably with Love & Fame, in both reviews and criticism, but I must believe that this is due less to the poetry than to the attitude Berryman seemed to strike in each volume. In Love & Fame Berryman appeared to be a bragging oaf and the appearance hid the poetry. In Delusions, Etc. Berryman is "properly humble," and this too seems to have hidden the poetry, for the best one can say of the volume is that it is very uneven. The chief failure of the book, moreover, is not one of sensibility, but one of language—it is not that Berryman is too "confessional" or that his conversion is not convincing, but that his artistic experiments too frequently do not succeed. (p. 98)

There is a distinction that must be made, even in these late, personal poems, between poet and man, between maker and subject. Berryman the poet has not been subsumed by Berryman the man. In all of these poems the emotion is worked, given form and order. Berryman was never a poet of "raw expression" (indeed, critics linking him with the Beats have overlooked this obvious fact), and even in Delusions, Etc. one must allow for the essential distinctions between the personality of the poet-in-the-poem and the personality of the poet. They are not completely discontinuous, but neither are they identical, and it is dangerous to argue across the barrier between them.

One fairly popular conception of the dangers inherent in confessional verse is that the poet, by seeking out the unsettling and self-destructive in himself, succeeds, not in controlling those feelings and putting them to the service of his art, but in allowing those feelings to control him. At least part of the responsibility for the rash of recent poet suicides, then, is to be laid both to the type of material the poets were concerned with and to the ultimate failure of their art. This might be true in certain cases, and one may look at Delusions, Etc. in this manner, but such a view would seem to me to be a great oversimplification. One must be very careful of the post hoc fallacy here. The relation between life and art—or at least Berryman's life and Berryman's art—is too complex to fit such a matrix. I cannot prove this, of course. To do so it would be necessary to conduct what scientists call a thought experiment—working out in one's mind (or in computer simulation) all of the results of all of the causes of both the life and the art. Nevertheless, I hope I have indicated enough evidence of the complexity of the relationships between Berryman's life and his art—even in these late poems—to call such a simple explanation into doubt.

There are both failures and successes of art in these late works. Despite the successes, however, both Delusions, Etc. and Love & Fame are overshadowed by the accomplishments of Homage to Mistress Bradstreet and The Dream Songs. It is hard to see how it could be otherwise. But I cannot argue from the weaknesses of many of these poems to a failure of life, or even a failure of talent, as many critics have—there are poems enough here to attest to the continuance of Berryman's gift. Moreover, what problems there are here are what might be termed positive problems—attempts to do with the language more than the language will successfully allow. Thus his failures here are not failures of character or even the failures of a dried-up talent, but those of the experimentalist that Berryman was until the end, pushing the language around at times quite roughly, to use it in ways in which it had not been used before. (pp. 99-100)

Gary Q. Arpin, in his The Poetry of John Berryman (copyright © 1978 by Kennikat Press Corp.; reprinted by permission of Kennikat Press Corp.), Kennikat, 1978.