Chapter 1 this storie/The World may reade in me (on translation) Shakespeare is a very good writer. This is not news. We know how clever he is, how versatile, how compassionate, how insightful, how constantly contemporary. But while this book might make a familiar case for Shakespeare''s brilliance, it will simultaneously argue that he can be appreciated for all his qualities even if we don''t hear a single one of his words. That Shakespeare with every word changed can still be great, and can remain Shakespeare. And that this possibility - that a play might be capable of surviving a full text transplant, from those 25,000 old words into these 25,000 different, entirely new ones, and surviving it somehow intact - isn''t simply a factor of his pretty good storylines, or his perceptiveness about general human experience, or the universality of his ''themes''. I work as a literary translator, which means that I am daily faced with that task: to replace the words in a book - all of them - in such a way that when I''m done, it is nonetheless still itself. There''s that old idea that within a seven-year cycle, all the cells in our bodies are replaced.

People who understand these things better than I do tell me that this is not exactly true - but nearly. A substantial number of the thirty-six trillion cells in my body (give or take) will have been formed in the last few years, yet I still remain essentially the same person. Over a period of thirty-plus years, Niels Brunse replaced almost all the component cells in the works of William Shakespeare, but his As You Like It is still essentially As You Like It; his Rosalind is still essentially Rosalind. So one of this book''s questions will be about what that ''essentially'' means, and what preserving that essence of expression entails. I am not naturally given to talking about the resilience of some transcendent soul while the gross cladding around it changes; I''m more interested in the mechanisms that go into physically regenerating this thing in such a way that, despite its being made up of new constituent parts, the differences are not what matters. I often encounter doubts as to whether a great writer like Shakespeare can ''survive'' translation - indeed, a very clever friend who knows Shakespeare far better than I do said, on hearing that I was writing this book, ''But Shakespeare can''t be translated, isn''t that the whole point?'' The assumption being, I suppose, that translation is merely a destructive process - or a process, at best, of damage limitation. And it is destructive, of course, but it''s creative, too. And in good translation, Shakespeare does not merely survive, he thrives.

Shakespeare''s ability to connect with actors and audiences beyond his own language and place - and indeed outside his own time - is predicated on a general agreement that there is something universal about his writing. The plays resonate because everybody falls in love, and power invariably corrupts, and death comes to us all. But it is not, of course, as simple as ''everyone gets sad sometimes, so a play about someone getting sad will resonate with everyone everywhere across all time''. In Shakespeare, people get sad with precision . King Lear dies and Desdemona dies and Falstaff dies and their experiences of death are entirely specific, and different from each other. Nobody dies generically. Cordelia and her intransigent father don''t have the same relationship as Celia and her intransigent father or Hermia and hers. And this precision of character and experience is located in its expression.

When I argue that Shakespeare is actually very good (I know!), my argument is rooted in his management of the very smallest units of writing - words, syllables, pauses - and not simply that he''s managed correctly to sequence four basic narrative building-block events into a story that makes people cry at the end. Precision is a word you''ll see a lot in this book - translation is about change, but that change obviously cannot be indiscriminate. Start swapping out our cells carelessly, or entirely at random, and we quickly become some other person altogther, or, I don''t know, a lobster or an avocado. (Translation is too often described through metaphors, incidentally, and this replacing-cells one is as imperfect as any other; so for the most part, this book will work by example, rather than abstraction.) Underlying this focus on precision in our translators'' work is a simple enough idea: language is not merely a delivery mechanism for basic dictionary-limited meaning. If you speak a word, what that word ''means'' is not the only thing that''s happening. It''s rare that you can paraphrase a bit of Shakespeare - dress the basic meaning in alternative words - without diminishing it. He puts the language under pressure, makes every part of it work hard.

A line in a Shakespeare play is custom-designed to have an effect - activating the actor who speaks it (because expression here is indissoluble from stagecraft), the audience listening. Replace the line with an approximation (maybe a couple of syllables shorter, an alternative collection of consonants, a differently placed breath), and the effect will change. Drastically? Well, no. I don''t believe that getting a couple of words wrong in Hamlet would bring the whole edifice clattering down. But these things are cumulative. The whole play is nothing but one syllable after another, and even small alterations will threaten the integrity of the structure if you make enough of them. The version of Kiss Me, Kate I saw last year in London opened with director Fred Graham giving his actors notes on their rehearsal of Shakespeare''s The Taming of the Shrew. One of the actors wasn''t quite on top of his lines, but protested that he had managed at least to convey their gist.

Fred''s response is at the heart of this book: People do not come to a Shakespeare play for the gist. And, crucially, a translation needs the same precision as the source text. Making a show in which, well, at least the boy''s still called Romeo and she Juliet, or near enough . and rival families, and there''s a nurse . That''s it. Oh, and some people definitely die at the end? I feel that is probably not enough. In a good piece of writing, every word might have a dozen different functions - the better the writing, the more precise, and perhaps the more subtle, will be the roles each word fulfils. It will operate syntactically, it will convey information, it will have a distinctive shape, texture and sound, resonating with other words, precisely influencing the rhythm of a sentence.

No word operates without context. So the translator is trying to find words in his own language that deliver all of those dozen things, not to mention preserving several simultaneous formal features of the writing (a fixed verse pattern, sounds that rhyme when spoken aloud on a stage, etc.). Impossible? No matter. Shakespeare has done quite well over these past 400 years, and his success is by no means limited to the Anglosphere. But discounting the language, one might ask: what''s left, apart from some variably robust plots? Well, of course, I don''t think we should discount the language; in translation, at least in good translation, the language is there, all the things that the language is doing will survive. A translator - a good translator - will work very hard to ensure that. Great literature survives as still great, not despite translation but because of it.

There''s a much-paraphrased line from the American poet Robert Frost about poetry being ''what gets lost in translation''. We translators are inclined to roll our eyes whenever it''s deployed. As much as anything, this book is about why we do. Dog , the noun, is chien, perro, Hund. Easy. That''s the basic information. But the word has other qualities. It also rhymes with log, can be used as a verb (as Shakespeare does), it''s a monosyllable, and can be a nice pun if there''s a nearby God.

Suddenly chien, perro, and Hund don''t quite cut it. (My Oxford English Dictionary devotes half a dozen densely packed pages to the word ''dog'' and its derivatives. The majority of what''s covered there will not overlap with what you might imagine as the ''direct'' translation of the word into another language.) The more sophisticated, complex, layered and playful a text''s language, the more a translator must struggle to find a replacement. Inevitably, that equivalent - a word in the new language with all the same dozen properties as the original word - cannot exist. So what''s a translator to do? And Shakespeare, of course, is about as sophisticated, complex, layered, playful, and - again - as insistently precise as they come. Look at these lines not from Hamlet : Being or not being? One can''t help wonder! To be, or to be not - thus goes the matter. Existence or its absence? You tell me.

Here''s the question: should we be, or not? A well-tuned translator would produce something different for each of these lines, despite the fact that the basic meaning they all convey is the same. What if Shakespeare had written This is the question rather than That? Or That is my question? What''s the difference between to be and being ? Well, the rhythm of the line for starters - that''s no accident, either. Because translators don''t traffic in meaning alone. Recreating meaning i.