As the text of Holy Scripture, the written Word of God and the ultimate source of authority, St Jerome's translation of the Bible was enshrined for centuries not as an authorised version but as the pristine word itself. Even in the 1950s, when I attended the pre-conciliar Latin Mass every morning of the school year, the beginning of the Last Gospel sounded like the first note of God's tuning fork. 'In principio erat Verbum et Verbum erat apud Deum, et Deus erat Verbum. Hoc erat in principio apud Deum.' I soon knew enough Latin to be thoroughly at home with this, with the result that the English translation ended up having less immediately persuasive power than what I then took to be the original. 'In the beginning was the Word, and the Word was with God, and the Word was God' —venerable as this English sounded, it came across as secondary. For whatever reason —maybe because of the numinous force which Latin then possessed as the medium of the liturgy and of the Church's magisterium, maybe because of some older need for a magic language that would altogether open and close the world— for whatever reason, the Gospel heard in my own tongue sounded smaller, whereas St Jerome's version came forth like those orb-sized words of pre-Babel speech imagined by Wallace Stevens at the end of his poem 'The Idea of Order at Key West':

Words of the fragrant portals, dimly-starred,
And of ourselves and of our origins,
In ghostlier demarcations, keener sounds.

To put it another way, when I stood up in those days at the end of the Mass and followed the priest's spoken Latin in my missal, I was becoming aware of the arbitrariness of English as a linguistic system. My Catholic education had succeeded to that extent, at any rate: the primal rightness of the Church's language had been established as a fact of the aural life. 'Adeste fideles' would henceforth win out over 'O come all ye faithful'. 'De profundis ad te clamavi, Domine, Domine exaudi orationem meam' would be prior to 'Out of the depths I have cried to you, O Lord, Lord hear my prayer'. I left St Columb's College, in other words, a perfect construct of that pre-Vatican II culture, and I remain reluctant totally to deconstruct myself. This reluctance, however, has little to do with the confession of faith. It arises from a source commonly acknowledged by myth and meditated upon by linguistic philosophers, and it is shared by all poets. Octavio Paz, in his essay on 'Reading and Contemplation', stated the case like this:

It has always been believed that the relation between sound and meaning appertained not only to the natural order but also to the supernatural; they were inseparable and the tie that joined them was indissoluble. This idea presents itself spontaneously to the understanding... and is extremely difficult to dislodge. I confess that it is only with great antipathy that I accept (provisionally) the fact that the relation between sound and meaning, as Ferdinand de Saussure and his disciples maintain, is the result of an arbitrary convention. My misgivings are natural: poetry is born of the age-old magic belief in the identity of the word and what it names.

Paz is talking here (in Helen Lane's translation) about the poet's longing for the utterly persuasive word, the word where the spirit 'unappeased and peregrine' can finally rest in peace, where the link between inevitability of sound and plenitude of sense is indissoluble. The diversity of tongues, however, broke this link. Contact with other tribes, trade with other nations, invasion by bigger empires, conversion to other faiths, education in other cultures —progressively and capably, from the cave to the computer age, human beings kept evolving as creatures of language, meeting the new and integrating it, but never without experiencing every time a vestigial tremor, a repetition of that first shock of hearing the other. Hence the writer's longing can be understood as a nostalgia for the original undifferentiated linguistic home.

In the above account of my St Columb's College experience, for example, I am really representing English as a form of diaspora, a diminished life lived in exile from a fullness available in the old country of the Latin. And when put like this, it parallels another famous account of an Irish student experiencing a similar case of linguistic exile or exclusion or, if you want to bring it down a peg or two, cultural cringe. At any rate, my sense of the hierarchical distinction that applied between the sounds of the Latin and of English obviously relates to the distinction Stephen Dedalus intuits between the English spoken by an English-born Jesuit and that of his own Dublin, or rather Drumcondra, vernacular. The Joyce passage is prompted by Stephen's momentary feeling that his own speech is being demeaned. The Dean of Studies is bemused by his use of the Dublin term 'tundish' instead of the standard English 'funnel', and as far as Stephen is concerned, the bemusement downgrades not only his speech but also his nation. He felt, Joyce writes,

with a smart of dejection that the man to whom he was speaking was a countryman of Ben Jonson. He thought: — The language in which we are speaking is his before it is mine. How different are the words 'home', 'Christ', 'ale', 'master', on his lips and on mine! I cannot speak or write these words without unrest of spirit. His language, so familiar and so foreign, will always be for me an acquired speech. I have not made or accepted its words. My voice holds them at bay. My soul frets in the shadow of his language.

At the end of Joyce's novel, Stephen Dedalus is cured of his 'fret' when he looks up the word 'tundish' in his dictionary and discovers that it is not an Irish provincialism, as the Dean of Studies had implied, but an English word, and, as he notes subsequently in his diary, 'good old blunt English too'. Adding, post-colonially ahead of the time, 'Damn the dean of studies and his funnel. What did he come here for to teach us his own language or to learn it from us. Damn him one way or another.' By finding that his Dublin vernacular is related to the old English base, Stephen discovers that his own linguistic rights to English are, as it were, prenatal. He may not be the true-born English man, but he is the new-born English speaker. And at this moment, he is also born as a writer, liberated from subject-people status, freed of the language question to become part of the language issue. He realises that his vemacular possessions are buried treasures, that his own word-hoard is the artistic equivalent of a gold hoard. Or to put it another way, the flicker of illumination in the word 'tundish' is an intimation of those ghostlier demarcations and keener sounds of the ideal poetic order.

I suppose I am trying to find a way of talking about the liminal situation of the literary translator, the one standing at the frontier of a resonant original, in awe of its primacy, utterly persuaded, and yet called upon to utter a different yet equally persuasive version of it in his or her own words. This, after all, is what St Jerome achieved. What I once responded to as the pristine word was in fact translation. It belonged, so to speak, to another persuasion, different from the Greek and Aramaic originals, and yet it had gone on to live a fullness of linguistic life beyond them. And as such, it stands as an ideal for the literary translator. As does Stephen Dedalus' transition from linguistic fret to linguistic offensive, since the literary translator often suffers from what might be called the funnel complex —a self-doubt induced by the supreme undoubtingness of the original— and must overcome it through an assertion of his own tundish-based entitlement. The literary translator might therefore take as a motto something that Austin Clarke once said to Robert Frost. When Frost asked the Irish poet about his writing procedures, Clarke replied, 'I load myself with golden chains and try to escape'. The 'golden chains' bit is easy enough to recognise. It could be the need to match the 'terza rima' of Dante's 'Divine' 'Comedy', including the difficulty of matching in English the feminine rhymes of the Italian; or it could be the challenge to repeat the mixture of formal light-footedness and vernacular gumption that characterises Pushkin's Russian; or going the other way, out of English, it could be the simple-seeming fetters of Emily Dickinson's quatrains, sometimes as domestic as a garden trellis, at others as charged as a transformer. Faced with such a bondage, the translator may well remain tongue-tied, shy of the job of making over the original. What the translator needs then is for a move to happen similar to Stephen's move from being daunted to being undaunted. In the introduction to my version of Beowulf, I trace the origin of my own translator-boldness to the word 'thole', which formed a part of my buried word-hoard. This word possessed the kind of fetter-easing power which I eventually found described with such sweet accuracy by the Beowulf poet himself, in his lyric simile of a spring thaw:

It was a wonderful thing,
the way it all melted as ice melts
when the Father eases the fetters off the frost
and unravels the water-copes, He who wields
power over time and tide: He is the true Lord.

I don't mean to imply by this that the line-by-line experience of translating Beowulf was as supple and deliquescent as this passage would suggest. Far from it: the fret-work of the day-to-day encounter with the alliterating Anglo-Saxon lines was edgier and more intricate. Like all work, of course, it generated its own pleasures and resolutions as it proceeded, but it could not proceed at all until it had found a way to get started, an excitement that would amount to an entitlement, and this came when I realised that the old-fashioned usage of the word 'thole' in the country speech I grew up with gave me a prenatal link to the Anglo-Saxon verb 'ðolian'. That became, if you like, the thole-pin and the secret fulcrum of the whole enterprise.

Still, entitlement or creative excitement is one thing; 'techne', the can-do factor, is something else. Luckily, however, the origins of whatever 'techne' I possess were also located in Old English, specifically in the neo-Anglo-Saxon of Gerard Manley Hopkins. Hopkins had a definite fetter-melting effect on my mid-Ulster tongue when I was an undergraduate, and all I want to do here is to represent the reality of that effect by a new parable. What I would now say is this: the lessness of English in relation to the Latin that I had experienced as a schoolboy was banished when I later read the opening stanzas of Hopkins's poem 'The Wreck of the Deutschland'. Suddenly the world was tuned to a new and irrefutable note. The liturgical power, the ontological base-touching, the body-bracing sheerness of sound in Hopkins' apostrophe to God the Father, had the force of a primal incantation, one which reinforced what Octavio Paz called 'the age-old magic belief in the identity of the word and what it names'. This English was not a set of arbitrary signs but given, oracular, from the beginning:

Thou mastering me
God! giver of breath and bread;
World's strand, sway of the sea;
Lord of living and dead;
Thou has bound bones and veins in me,
Fastened me flesh,
And after almost unmade, what with dread,
Thy doing: and dost thou touch me afresh?
Over again I feel thy finger and find thee.

One result of this encounter with Hopkins's sprung rhythms was the eventual composition of a number of poems that tuned the abrupt, staccato movement of my first Northern lrish speech to the old, alliterative measure of the Anglo-Saxon. And what happened twenty-odd years later when I was invited by the publisher of the Norton Anthology of English Literature to try my hand at translating Beowulf was essentially a repetition of that early process. To quote briefly from my introduction:

While I had no great expertise in Old English, I had a strong desire to get back to the first stratum of the language and to 'assay the hoard' (line 2509). This was during the middle years of the 1980s, when I had begun a regular teaching job at Harvard and was opening my ear to the unmoored speech of some contemporary American poetry. Saying yes to the Beowulf commission would be (I argued with myself) a kind of aural antidote, a way of ensuring that my linguistic anchor would stay lodged on the Anglo-Saxon sea-floor. So I undertook it.

The translation that Norton commissioned was intended to replace a scholarly prose version by E. Talbot Donaldson, an authority on the poem and one of the editors of the Norton Anthology, where his translation has heretofore appeared. The publishers therefore wanted to be sure that my work would not depart too far from the line-by-line meaning established by generations of editors and commentators, so in order to keep their minds at rest and me on my toes, they appointed a reader who was a kind of minder. Once I had completed five or six hundred lines, I sent them to the New York office and the office sent them on to this man whom I did not then know except as a name at the bottom of a letter. His brief was to keep me from my own mistakes and to point out what might be considered my oversteppings, and I was lucky that he combined a deep knowledge of Beowulf's language and meanings with a real feel for what might be permitted if not altogether required in a new translation.

Here's an example of what went on. In the Anglo-Saxon, there is a famous description of the mountain tarn where Grendel and his mother dwell. It contains the following lines about how it is so deep and so forbidding that a hunted deer will allow itself to be torn apart by the hounds rather than enter it:

Ðær mæg nihta gehwæm    niðówundor seon,
fyr on flode.    No þæs frod leofað
gumena bearna    þaet þone grund wite.
Ðeah þe  hæðstapa   hundum geswenced,
heorot hornum trum   holtwudu sece,
feorran geflymed,   ær he feorh seleð,
aldor on ofre,   aer he in wille
hafelan hydan.
(lines 1365-72)

My first version went as follows:

At night there, something uncanny happens:
the water burns. And the water is bottomless.
Nobody alive has ever fathomed it.
There too the heather-stepper halts:
the hart in flight from pursuing hounds
will face up to them with firm-set horns
and die in the wood rather than dive
beneath its surface.

My reader was prepared to let me off for letting myself off the job of alliteration in a line like 'Nobody alive has ever fathomed it', but alliteration could not quite compensate for what I had written in the previous line: 'the water burns. And the water is bottomless'. His note read: '"bottomless." Well, the water is so deep that no one has ever fathomed the bottom: "it" in the next line must refer to some bottom.' And then came his reminder about the kenning for the hart:

Heather-stepper. [Is the literal sense not] heath-stepper? Again, a short line —especially if heather becomes heath. Possibly re-arrange? 'There the hart halts, the heath-stepper / hard-pressed in flight by pursuing hounds...' I like how the 'firm-set horns' become functional in the translation. In the original they are simply an attribute of the hart.

My letter back on this occasion seems to me to be worth quoting at length, since it reveals all the contradictory commands which the literary translator will feel called upon to obey. It so happens I had not answered directly until this moment because my instinct was to defer consideration of the comments until I was so deeply engaged with the work that I could not be put off my stride —not that stride was exactly the word for the pace I was going at. 'I am writing now,' I said,

on impulse, having got to that moment in the text where I call the heath-stepper the heather-stepper and where I want to persist with my perverse rendering... In general, however, I have made the revisions on the lines you suggest... glad to be convinced in so many places that what is called for is a more literal rendering, more of a word-for-word match with the original. Hence, this morning I've tried to rewrite the bit where Hrothgar describes the prowlers on the moor; and have got to the 'bottomless' water of the mere, which you rightly point out goes beyond what the original warrants.

At this point, I might have pleaded out that 'bottomless' occurs as the last word in an early poem of my own, but even if I did not adduce the chapter and verse, I went on to say that for me 'bottomless' was a word

with mere-y suggestions, since as a child I was always being warned away from bog pools in our district — because they had 'no bottom to them'. So I was prepared to transgress, and paused for a while before coming round to a different rendering. In general, after my pauses, I have come round. I dropped 'thole' early on; dropped 'wallstead' but then retained it; foresee a reluctance to drop 'gap of danger'; must warn you about the appearance of seanachaí (an Irish/Hiberno-English word for a professional storyteller); and so on... And to come to this morning's heatherings and thitherings, so to speak — heather is a word deeply within my own first speech whereas heath is [to my first ear] very much a literary word — Danelaw-y, English, and admittedly very thrilling in its deep [King] Lear-y wildness. I have nothing against literary words, of course, and in fact I thrill to the depth-charges in 'heath', back through Hopkins's 'wiry heath-packs' and Brontë's Heathcliff, and so on —but somehow, 'heather-stepper' has more spring in its step for me and since me heath is contained in the heather, I want to hold on to it.

I said earlier that I wanted my anchor to be lodged on the Anglo-Saxon sea-floor, down in the consonantal rock, but I had a second mooring down in the old soft vowel-bog of the local speech. I was honour-bound to the feel and sense of the original, but at the same time could not desert whatever it is in my ear that makes me sound convincing to myself.

At a seminar on translation last year, Efim Etkind quoted Samuel Marshak, a great translator of poetry from English into Russian. 'Poetry is impossible to translate,' Marshak declared. 'Each time it is an exception to the rule.' So I would not claim that there is anything exceptional about the work I have done, just that it constantly manifests the impossibility factor. Consider, for example, the problem of translating a line that occurs earlier in the heath-stepper passage. The original tells us that the country people have seen two 'micle mearcstapan moras healdan / ellorgæstas' (lines 1348-9) —two big border-steppers keeping to the moors, spirits from elsewhere. In the final version, I rendered this as 'two such creatures / prowling the moors, huge marauders / from some other world'. I liked the menace and stealth in the word 'prowling'. It seemed to catch what was shadowy and ghostly about these 'ellorgæstas', as well as what was wild-beastish and brute-dangerous. An earlier version, however, had rendered them as two such creatures, 'ranging the moors, huge marauders / from some other world', because 'ranging' alliterated with the stressed syllable in 'marauding', and I was wanting to keep as much as possible to the four-stress pattern and the alliterative requirements of the Anglo-Saxon line. Nevertheless, in spite of the falconish sweep of the verb 'range' and the glamour it still retains four and a half centuries after Sir Thomas Wyatt used it in his poem 'They flee from me' (where his former lovers 'range / Busily seeking with a continual change') —in spite of all this, I went for the unalliterating 'prowl' because it contained more darkness and danger. On the other hand, when it came to telling where the monsters dwelt, on 'windswept crags / and treacherous keshes, where cold streams / pour clown the mountain', the word 'kesh', meaning a causeway or log bridge, presented itself uncontradictably, combining as it did the local and the alliterative, the drag of the golden chain and the fret-free exhilaration of having slipped me leash.

I suppose I could well have ended up signing my envelopes to my word-warden 'Tundish Rules OK', not as a triumphalist slogan but simply as a poet's caveat. At any rate, I came to restore the word 'thole' in my rendering of some of the early lines of the poem, and even though I finally substituted 'reciter' for seanachaí, I did hold on to 'gap of danger'. And I risked oddity for the sake of fidelity to an older idiom and emphasis when I translated the gnomic praise of Scyld Scefing, 'þaet waes god cyning' (line 11), as 'That was one good king'. 'He was a good king' or 'That was a good king' is the usual rendering, but in this case the Ulster vernacular use of 'one' in order to distinguish a person seemed more than justified, since it echoes nicely the Latin usage of ille to indicate somebody set apart and acknowledged as the possessor of certain qualities. And I used a similar bit of Ulster-speak in another place, translating the phrase 'waes seo þeod tilu' when it was applied to the Geats (at line 1250) as 'They were a right people'.

In the end, the point is this: 'words of the fragrant portals' come glimmering up out of the merest puddles, from the very capillary roots of consciousness in forgotten or half-remembered sensation. 'Literary' does not mean 'lofty'. The proper translation —'proper' in the Latin sense of belonging, belonging recognisably to the original and to the oeuvre' of the translator— exists half-way between a crib and an appropriation. For the truth is, as Eliot Weinberger has said, one of the big motives for translating at all is in order to write vicariously, and for that vicariousness to be complete, the writing as to include those quickenings and homecomings that accompany successful original composition. What keeps the translator in a state of near (but never quite complete) fulfilment is this tension between the impulse to use the work in its first language as a stimulus and the obligation to give it a fair hearing in the second. And because of the strong pedagogical function served by the Norton Anthology, I was more than usually subject to that tension. Indeed, there could be no better illustration of the fact of the tension itself than the footnotes in the new volume. At certain points, it is the very translation that has to be translated for the benefit of the worldwide audience of English-speakers to whom the anthology is directed.

Given that audience, some such light editing is certainly called for. My favourite instance refers to a word used to render the poet's laconic description of what was going on in Heorot Hall when Grendel and Beowulf were in the throes of their combat. 'Dryht-sele dynede,' says the Anglo-Saxon; 'Denum eallum wearð, / ceaste-buendum, cenra gehwylcum, / eorlum ealuscerwen' (lines 767-9). 'The lord's hall resounded; to all the Danes, the dwellers in the castle, to each brave one, it was a warriors' ale-sharing.' My version went like this:

And now the timbers trembled and sang,
a hall-session that harrowed every Dane
inside the stockade: stumbling in a fury,
the two contenders crashed through the
building. (767-70)

And the footnote in the anthology is to read: 'In Hiberno-English the word 'session' ('seisián' in Irish) can mean a gathering where musicians and singers perform for their own enjoyment. (Translator's note).' Enough. The examples could be multiplied. What they all go to prove is that one of the most acute dilemmas faced by a contemporary poet is one that is also shared by the contemporary literary translator. Moreover, since this dilemma is explicated so vividly in Ted Hughes's extraordinary essay on 'Myths, Metres, Rhythms', I want to turn to that in order to conclude. The essay is essentially a meditation on the middle state of the writer, poised between his own idiolect and the vast sound-wave and sewage-wash of the language's total availability, but, being the poet he was, Ted Hughes eschewed the use of a technical term like idiolect. His critical prose, in spite of recent insinuations to the contrary, was never at odds with the flourishing of his creative impulse, so in this instance he typically invented one of his animal parables.

One gazelle flicks its tail —and the tail flick goes from gazelle to gazelle right through the herd, while they keep their heads down, nonchalantly feeding. To the individual gazelle it must feel like a communal brief prayer, meaning: while we all exist as one gazelle, I exist as full strength gazelle, immortal gazelle.

In fact, this parable could equally well be a parable for the indissolubility of individual consciousness, shared language and cosmic at-homeness that we suppose existed in the world before Babel. Hughes employs it, however, to illustrate the way the non-standard language of any sub-group functions as a means of communicating and conserving 'the voltage of the whole group's awareness and energy'. This shared inner language is also, however, a badge of the group's eccentricity when they come to speak it as part of society's lingua franca. 'From the point of view of the lingua franca,' Hughes writes, 'the solidarity system and mythology of any sub-group tends to appear parochial, old-fashioned, limited and limiting — to be indulged, if at all, only as local colour.' On the other hand, from the point of view of the sub-group, 'the lingua franca appears shallow, arbitrary, empty, degraded and degrading, even destructive, if not altogether meaningless'.

As ever with Ted Hughes, there is a great cogency in this writing, and it does not lessen when he goes on to outline its implications for the writer.

Setting aside just how any writer resolves or fails to resolve this dilemma, the fact remains that each modern literary works has to take its place on a continuum between some sub-group's (author's) system of shared understandings... and the most inclusive, ideally global wave-length of a multicultural lingua franca.

Whether me writer intends it or not, is even conscious of it or not, by the very act of bringing the work to linguistic focus they fix it at some point on that continuum. Just where that point is only becomes clear after publication.

So, in a sense, there is nothing more to say. The translator, like the writer in Hughes's parable, 'can only grope along, transmitting what are intended to be meaningful signals, the most meaningful possible'. If the work is successful, the flash of the right word-choice should create a tremor that makes readers feel they exist as 'full strength' members of the language-group. A voltage should travel up the line, from the hoard into the herd, a sensation of being tied into an extensive and self-fortifying network, a far-reaching system of pulse and beat and heft and hold. The individual translator of Beowulf shoulders the burden of the past and tries to launch it into the swim of the present. My own metaphor for this process comes from a poem called 'The Settle Bed', about a big unwieldy piece of rural furniture, as solid and clinker-built as a Viking longship. The settle in question came into my possession after a cousin of my father's left it to me in her will, mine to keep, to have and hold and bequeath in turn —so the following lines seemed an appropriate epigraph for the translation:

and now this is 'an inheritance' —
Upright, rudimentary, unshiftably planked
In the long ago, yet willable forward

Again and again and again.

To the literary translator, this attempt to will a thing forward is the raison d'être of the whole business and belongs, as such, with the larger effort of poetry itself, especially if poetry is conceived of as Czeslaw Milosz has conceived of it, 'a dividend from what you know and what you are'. Put in a different way, this means that our language pays tribute to itself when tribute is exacted from it; it suggests that our value to ourselves as individuals or as a group or even as a species can be re-estimated and increased by dwelling upon the sum total of the experience stored in our word-hoard. Our fret as investors in ourselves can, if you like, be allayed when poetry recirculates the language's hidden wealth, a recirculation that is not only etymologically renovating, but psychologically and phenomenologically so as well.