Translating Rilke’s poetry presents formidable difficulties. There are grammatical elisions and omissions which can make the language appear abstract when it isn’t, and imagery which appears symbolic or realistic but may, in fact, be both simultaneously. There is constant movement between statement and image (and, particularly in the Elegies with their greater length, strings of images – in some ways Rilke is as inveterate an imagist as Mandelstam). There is also the issue of the poet’s voice, by turns meditative and anguished, mocking and rapt, and, behind all this, of elegiac tone in both the usual sense and the sense special to German poetry. In the Sonnets to Orpheus, there are the concentrations and compressions that inevitably stem from using the sonnet form.

There have been many attempts at translation into English – William Gass’s survey deals with 15 versions of the Elegies, for example – but none of them approaches the effect of Rilke in German. As translators have built on the efforts of their predecessors, translations have developed from the muffled abstractions of Leishman’s and Spender’s 1939 Elegies, through the more direct versions of Robert Bly and David Young, to translations of all Rilke’s significant poetry by Stephen Mitchell (1985) and Graham Good (2004) which are accurate, clear and, until now, the best that could be imagined. Yet compared with the German original, even these versions lack whole dimensions of meaning and resonance. I had begun to think that Rilke is, in some profound sense, untranslatable, but now realise that the problem is that all the modern translations are by academics or by poets like Bly and Galway Kinnell who, in their own work, prefer direct demotic utterance to other kinds of writing.

This realisation follows the simultaneous but unrelated publication of translations of Duino Elegies and Sonnets to Orpheus by two major British poets. In very different ways, these versions transform our understanding of Rilke in English and one of them is unlikely to be bettered for very many years. Both Crucefix and Paterson bring long experience of writing poetry which responds to complex movements of the human spirit. In addition to great formal skill, this gives them a fine-tuned awareness of the potentialities of language, of how choices of particular words and phrases affect understanding.

To begin with the problem of apparent abstraction which has bedevilled previous translations. The First Elegy evidently begins with the poet crying out to the angels for help – “Who, if I cried out, would hear me among the ranks / of the angels?” – but realising that, if one embraced him, he would “wither”. This is followed, in the poem’s second sentence, with an apparent abstraction: “Denn das Schöne is nichts / als des Schrecklichen Anfang”. Previous translations render this as versions of “For Beauty is nothing but the beginning of Terror” and the reader is left wondering, whose beauty? The line reads as a generalisation about beauty, like Keats’ “Beauty is Truth”, and the intensity of the first lines is disrupted.

Martyn Crucefix understands that the beauty is the angels’ – “For their beauty / is really nothing but the first stirrings of a terror / we’re just able to endure” – which leads to the poem’s abrupt third sentence: “Every angel is terrifying”. We understand without distraction that the poem’s opening line isn’t a cry for help, but a cry of terror at the thought of massed ranks of angels. Rilke establishes from the start that angels, which represent all religious entities, are no help to mankind and that religious faith kills something vital in the human spirit. As he puts it later, if an archangel took a step towards us, “we would be beaten to death / by our own high-beating heart” (II, 11 – 12). His angels are exterminators.

Crucefix has understood that Rilke is very rarely abstract and this makes his translation both clear and resonant. Again on the poem’s first page, Rilke says that even the dumb creatures know that we humans are not at home “in der gedeuteten Welt” which Mitchell and Good translate as “in our interpreted world”. What, the reader wonders, is an interpreted world? who interprets it? Crucefix has “in our interpretations of the world” which both makes perfect sense and captures Rilke’s essential point that all experience is subjective.

Two other examples must suffice. Rilke says we must work ourselves free of those whom we love and “endure as the arrow endures the tensed bowstring, / becomes something more than itself in the leap / of release” (1, 63 – 64). He follow this with “Denn Bleiben is nirgends” which Good translates literally as “For staying is nowhere” and Mitchell as “For there is no place where we can remain”. Crucefix has “For our point of rest is nowhere” which expresses Rilke’s view that all experience is provisional, however much we desire permanence, while glancing at Zeno’s paradox of the arrow that cannot move.

And in the vision of life-after-death in the Tenth Elegy, the dead man must climb alone “in die Berge des Ur- Leids”. Previous versions translate this as “into the mountains of primal pain”, leaving the reader to wonder what “primal pain” could possibly be. Crucefi x has “into the mountains of original grief” which, with its echo of original sin, expresses the condition of grief to which mankind is born – which, for example, Hopkins recognised in Margaret’s tears over Goldengrove unleaving.

Crucefix has brought greater poetic resources to bear than any previous translator of the Elegies and the result is faithful to Rilke’s intentions both at the level of textual meaning line-by-line and at the levels of image and, crucially, tone. Crucefix recognises that Rilke is sometimes sardonic and matches this in tone and movement:

I cannot bear these half-filled human masks.
Better have a puppet. At least it is full.
I can put up with a puppet’s stuffed limbs, wires, the appearance of a face… (IV, 27 – 30)

and, in the fairground of human existence:

There is – for adults only –
Something special to see: how money multiplies!
in the raw! not just for entertainment! money’s
genitalia! the lot! the business! uncut – educational
and it will improve your performance… (X, 31 – 35)

Crucefix manages the transition from this to the solemn description of the young man’s journey into death without a false note, as also the movement at the end of the Fifth Elegy from the description of Death as a showy Parisian milliner to the rapt description of lovers making love on a carpet, athletically like circus performers, before an audience of “the silent and innumerable dead”.

The emphasis on death-in-life throughout the Elegies brings me to the question of elegiac tone. The poems are in the German tradition of elegy as practiced by Goethe, Schiller, Hölderlin and others – not about a dead person but rather a meditation on how the present is exceeded by the past in culture, spirituality or morality. These issues did not interest Rilke greatly; by the time of writing the Elegies and Sonnets to Orpheus he was, like Chekov’s Masha, in mourning for his life but not entirely seriously. Crucefix renders Rilke’s elegiac tone by close attention to cadence. This shows itself in individual lines which are both beautiful in themselves and elegiac in movement; for example, the mother’s protectiveness towards her child – “There was no creak your smile could not explain” (III, 48) – and the ironic image of religion’s lack of meaning:

and the church alongside it, bought ready-made,
clean, closed, disappointing as a Post Office on Sunday. (X, 22 – 23)

Crucefix’s skill with cadence also shows itself over longer passages, for example at the end of the Eighth Elegy where the image of the bat’s fl ight is rendered delicately – “As if frightened of itself, it must hurtle / through the air the way a crack goes / through a teacup – so a bat’s track / streaks through the porcelain of evening”. The repetition of “crack” and “track” and the assonance of “streaks” are close to Rilke’s tone and allow the reader to perceive the fragile beauty of the evening, in a way that Mitchell’s “zigzags” and Good’s insistent “bat’s track / fracturing” do not. In Crucefix this leads to a rendering of the last line – “so we live on, forever bidding goodbye” – with its echo of Keats. Simple as such lines are, no previous translator has approached their plangent clarity.

The Sonnets to Orpheus present different formal challenges – 55 sonnets with somewhat varying line lengths but mostly full rhyme. Paterson varies line length and uses half rhyme to accommodate meaning, which presents no problem to the modern ear. He also gives each sonnet a title, as has Willis Barnstone in his 2004 version though Barnstone gives the German text with its numbering in parallel. The effect of the titles, like titling the book Orpheus, is evidently to reinforce Paterson’s point – made on the title page and discussed at length by Paterson at the end of the book – that this is a version, not a translation. This should alert readers to the fact that the book is not a close translation of Rilke’s Sonnets

which is indeed the case.

In fact, Paterson’s versions vary enormously in their fidelity to Rilke. Some like Horse (1, XX) are close translations of the sense of the original and have a vigour that previous versions lack, although even here “his ticking mane” is a bizarre rendering of “wie schlug seiner Mähne Gelock / an den Hals” (‘how his mane beat against his neck’). ‘Schlagen’ can be used for a clock beating time, but in no meaningful sense does a horse’s mane tick. Others, like the final sonnet, render Rilke’s tone and much of his meaning persuasively throughout.

In general, however, each sonnet includes renderings of parts of Rilke’s text which are accurate or at least close in meaning and other parts which are so inaccurate as not to represent Rilke closely or at all. The proportions vary sonnet by sonnet. Take Dog (1, XVI), for example; the original has no title and includes no mention of dog or dogginess, so that the poem’s meaning becomes clear only at the end when Rilke refers to his master’s hand and describes the creature as “Esau in his pelt”. Paterson renders the opening “Du, mein Freund” as “My dumb friend”; invents a reference to barking in line 7 and to cursing in line 8; then omits all mention of “my master’s hand”. The effect is semi-Rilke.

As a final example, consider Sonnet 2, XVI, which Paterson titles The Stream. This is an important sonnet referring to mankind’s relationship with Orpheus the singing god and Paterson’s version begins far more effectively than any other:

God is the place that always heals over,
however often we tear it. We are so
jagged, as we always have to know

The next line is “aber er is heiter und verteilt” which Mitchell and Barnstone translate as “scattered and serene” and Good as “diffused and at peace”, implying ironically that the god has been torn so often that he is used to being in pieces and is quite calm about it. Paterson ducks the difficulty of this and invents “but withholding both his favour and disfavour” which merely prepares for the god’s indifference described in the second quatrain. Here, Paterson invents two lines which have no parallel in Rilke: “standing motionless to face the rift / our each enquiry opens in his realm”. One notes the clumsiness of “our each enquiry” and the antiquated word “realm”. Quite often when Paterson is inventing Rilke, he lapses into untypically artificial or old-fashioned language. I like to think that this is a symptom of subconscious guilt or at least unease.

The first tercet of the sonnet expresses Rilke’s meaning clearly, with the “gehörten Quelle” (‘heard spring’) becoming “that low hissing is the stream”. However, the second tercet imagines us praying to keep the hissing near us. What Rilke actually writes is “Uns wird nur das Lärmen angeboten” which Good translates precisely as “We are offered the sound instead”, that is, the sound only, not the drink. Again, Paterson – who writes of his scientific materialism in an Afterword to the sonnets – resorts to unfelt religious language (“We pray”) when inventing.

I wondered why Paterson had spent so much effort in producing inaccurate versions when accurate ones would not have taken any more time. He admits that he is not a linguist, but the various versions, including M Herter Norton’s which he says he used chiefly, would have given him a fair guide to meaning. The answer is given frankly in his Afterword: “My main motivation in making this version was selfish. I wanted to make a rhymed English version for my own use… meaning one I could memorise and carry round in my head. Over the last twenty-five years I’ve undergone a long and at times painful conversion to scientific materialism… Accomplishing this gave me some satisfaction, but it left the room terribly quiet and empty. I then sought some text I might get in my head as a vade mecum, whereby I could simply remember what I now held to be most true… It was a surprise to find that the best candidate was a poem I already knew pretty well.”

One respects the honesty of this. Orpheus is not Rilke but Paterson’s version of Rilke, made for his own purposes. But this admission leads to a valuable consideration of how the Sonnets, like all Rilke’s writing, resist religious interpretation and cannot be read meaningfully as ‘spiritual literature’. And this leads in turn to a profound discussion of Rilke’s focus in the Sonnets

on Orpheus, the god of death and singing: “To sing as a human is not to sing as birds sing; as birds sing, humans talk. For a human, to sing is to do something unique and with no analogue in other species. It is to unite the discrete quanta of passing time through music and lyric”. Paterson has understood, more clearly than most other commentators, that Rilke saw artistic creation as the only stay available to mankind against time-the destroyer and dedicated his life to the art he could perform – poetry. Which is why Rilke is an extreme case of the poets’ poet, an avatar with which every serious poet has eventually to come to terms.

Paterson’s Afterword is a permanent contribution to Rilkean studies and will, I suspect, be read long after his versions of the Sonnets. While Paterson uses Rilke, Crucefix has put his talents at Rilke’s service and produced a translation of the Elegies which makes all previous ones clumsy and partial. With the German text in parallel and a useful commentary on each elegy, it deserves to become the standard English edition worldwide. One hopes that Enitharmon, a small London publisher, negotiates a deal with one of the American university presses to make this happen.