It is no small irony that for a writer so punctilious about his texts – almost notoriously so for their performance – Samuel Beckett's work has been subject to so much inept editing and so many publication blunders that he could lament to his official biographer, James Knowlson, ‘my texts are in a terrible mess’. The innumerable printing errors introduced into early editions of his work (the edition of Watt published jointly by Collection Merlin and Olympia Press in 1953 and reprinted then both by John Calder in the UK and Grove Press in the US being perhaps the most egregious) have still never been fully corrected, although some progress has been made with recent editions (Grove Press, 2006; Faber & Faber 2011). On 13 August 1992, John Banville, then Literary Editor of the Irish Times, could lament in an essay for the New York Review of Books, ‘It is time now for all of Beckett's works […] to be properly edited and published in definitive and accurate editions in order that future readers be allowed to see them for the unique testaments that they are’ (20, emphasis added). One could hardly agree more – but Banville's call for something like textual purity may simply be a longing for a ‘paradise lost’, since textual problems are more easily recognised and ridiculed than remedied.

A spate of letters to the Times Literary Supplement as Dream of Fair to Middling Women was published is a case in point. What should have been a cause for celebration, the appearance of Samuel Beckett's long-suppressed first novel of 1932, has instead fuelled a textual controversy and led to a clash of egos. Although Beckett wrote only one Dream of Fair to Middling Women, two separate and competing editions of it, with more than a few typographical differences between them, remain in print. In his letter to the Times Literary Supplement on 16 July 1993, Eoin O'Brien, co-editor of Dream, dissociated himself from the second edition, although he remains listed as its editor: ‘Both the US (Arcade) and UK (Calder) 1993 editions of this work have been printed without taking into account the necessary corrections I, and my co-editor, Edith Fournier, made to the proofs of the re-set text.’