Gapped text

It was a single typed sheet, dated a fortnight before his death, and it was not an ending at all. It was a list of instructions, addressed to no one in particular, about what should become of the manuscript if he did not live to finish it. The last item read, in its entirety: ‘Do not finish it.’ Judit watched me read it twice. ‘He never showed this to his editor,’ she said. ‘I think he was embarrassed.’

Remarkably few reviewers mentioned the appendix, and those who did treated it as a scholarly curiosity, of interest chiefly to specialists. But I have had more letters about this book than about any other I have translated, from readers who say that its last page altered how they read everything before it. One woman wrote that she had wept not at the ending but at the instruction not to write one.

Judit’s answer, when it came, was unequivocal. The English edition should stop exactly where her husband had stopped, in the middle of a sentence, and anyone who wanted more could learn Hungarian. She had already written to my publisher to say as much, in terms that left little room for negotiation, and she expected me to support her position in whatever discussion followed.

Her reply was characteristically brief. She did not mind, she said, what the English edition contained, provided it also contained the truth. That phrase, which I first took to be evasive, turned out to be the most practical advice anyone gave me. It did not tell me which ending to use. It told me what I could not leave out.

The typescript was less a draft than an argument he had been conducting with himself. Whole paragraphs had been struck through and then, in a different ink, reinstated; marginal notes contradicted one another across the same page; a character called Ilona became, two-thirds of the way through, a character called Edit, with no hint as to whether this was a change of mind or a different woman altogether. The exercise books were worse. They held what appeared to be the ending, in three incompatible versions.

I am conscious of the irony that the English edition, like the Hungarian, does fill them, at least until the appendix begins. I have made my peace with that, more or less. A novel is a public object, and its readers have rights too, including the right to an ending, however provisional. What they do not have, I think, is the right to be deceived about where that ending came from.

That edition appeared in the autumn, and the reviews were respectful, even warm. Several critics singled out the final chapter as the most moving thing Varga had ever written. I read it with mixed feelings, since I knew it to be a splice: the first half of one exercise-book version grafted, with considerable skill, onto the second half of another. It was, I had to concede, better than either.

Part of the answer lay in the nature of the job itself. A translator is already, by definition, a kind of forger: every sentence I produce is one Varga never wrote, offered to readers under his name. That uncomfortable fact is bearable only because there is an original against which the forgery can, in principle, be checked. Here there was no such original, or rather there were three, and choosing between them felt less like translation than ventriloquism.