Which edition of George Orwell’s most famous novel is on your shelf and why does it matter?
The first British and American editions of Orwell’s great satirical novel, Nineteen Eighty-Four, published in June 1949, conclude with the protagonist, Winston Smith, intellectually and physically broken by an oppressive totalitarian regime. This is symbolised by the disturbing image of Winston tracing 2+2=5 with his finger in the dust on the table at the Chestnut Tree Café as he plays a solitary game of chess. It is terribly bleak. Orwell, who was extremely ill while drafting the novel, died shortly after it was published from pulmonary tuberculosis.
Michael Radford cast the late John Hurt as Winston for his superb film adaptation of the novel (released in 1984). Hurt, whose physical appearance was not dissimilar to Orwell’s, gives a memorable performance. Please view the final scene at the Chestnut Tree Café for thirty seconds or so before reading on.
Radford had apparently taken the liberty of having Winston tracing 2+2 = rather than adding the 5. For those familiar with the novel, this is a significant textual variation which completely changes the meaning conveyed about Winston’s own mental freedom and the omniscience of the regime that controls Oceania. It gives the viewer hope (and is not the only one of Orwell’s novels to have a challenging ending significantly changed for the big screen to make it less bleak).
However, if one studies the publishing history of Nineteen Eighty-Four in the UK, it is not the case that Radford changed the ending at all. His script was echoing how every edition of the novel was printed between 1950 and 1987 (except one).
A Brief Publishing History
Several years ago Dennis Glover, author of The Last Man in Europe, alerted aficionados of George Orwell’s writing that the deeply disturbing conclusion to Nineteen Eighty-Four may have been changed by the author as he lay dying. Glover, while browsing through a bookshop, found that a rare second impression, published in March 1950, just two months after Orwell’s death, no longer included the “5”. Glover speculated that Orwell may have instructed his publisher to make a textual change to this second printing of the novel providing some hope that Winston is still capable of independent thought and able to resist both Big Brother and the Party by writing 2+2 =.
It is unlikely we will ever know for certain why the first two UK printings had this significant textual difference. Oddly, the third impression returned the 2+2=5 but every subsequent printing in the UK after that had 2+2 = until 1987. Glover explains it well in his introduction to the Black Inc. edition of the novel:
To recap: the ‘5’ was present in the first impression of the first edition (June 1949), was omitted, probably intentionally, in the second impression (March 1950), then reappeared in the third impression (August 1950), only to be removed again in the re-setting of the book for the second edition (December 1950). It is just possible that the second omission of the ‘5’ was an error resulting from the typesetters of the second edition incorrectly proofing their new layout against the second impression, rather than against the first impression. But it is more likely that Secker & Warburg made a positive editorial decision to omit the ‘5’, consistent with the overwhelming likelihood that the ‘5’ had been purposely removed from the second impression of the first edition, possibly following Orwell’s instructions.
In 1987, ‘definitive’ editions of George Orwell’s standard works, edited by Peter Davison, were published. From this year onwards, all British editions of Nineteen Eighty-Four concluded with 2+2=5 for the first time since the third impression published in 1950, with one exception. Bernard Crick’s scholarly, annotated edition of Nineteen Eighty-Four (Clarendon Press, 1984) has 2+2=5. Crick included an extremely interesting footnote:
The Uniform Edition, despite “2 + 2 = 5” clearly in both Orwell’s typescript and the proof, printed “2 + 2 = “. The American edition always had the “5”. I now follow Peter Davison’s text, despite a sneaking feeling that the empty space may have been a last-minute correction by Orwell to signify not that the regime clearly says that two plus two is five, and that Winston accepts it; but that he is now open to take anything they say, and they are completely arbitrary. This is more in accord with O’Brien’s line of argument. However, with no record of a correction, the presumption must be in favour of a printer’s error.
Crick did not know that the “5” disappeared in the March 1950 printing, just two months after Orwell’s death, making it more believable that the author had requested the correction. However, rightly, Crick (being Crick) knows that the evidence supports that an error has been made not the more fanciful speculation.
The main sources for understanding the publishing history of Nineteen Eighty-Four are by Ian Willison (1953/2012), Frederic Warburg (1973), Peter Davison (1987/1998/2013), Gillian Fenwick (1998) and DJ Taylor (2013/2019). When one compares their commentary as many questions arise as answers. There are many errors.
The other important sources are the various printings of the novel that sit on my bookshelf. I have more than thirty-five different copies of Nineteen Eighty-Four, many of these in languages other than English. Dennis Glover kindly found a second impression of the novel for me last year at a rare booksellers in Cambridge and I have managed to collect all UK impressions of Nineteen Eighty-Four printed between 1949-1955 by Secker & Warburg.
Frederic Warburg, who took a gamble and also published Orwell’s Animal Farm, when it had been rejected by T.S. Eliot on behalf of Faber and many left wing publishers for political reasons, is a good place to start. Warburg explains in his 1973 autobiography that:
On June 8 we published 1984 in an edition of 25,575 copies. By the end of October we had sold 22,700 copies. By March 1950 a second printing of 5,570 copies was ready, in April a third printing of 5,200 copies, in November a fourth printing of 5,000 was ready. These lasted till April 1951, when a fifth printing of 7,570 copies became available. These five printings totalled 49,917 copies. In the next 15 years to 1966, a further 21,400 copies were sold. Between 1966 and 1971 the sales averaged about 2,000 copies a year, bringing the total number sold to 81,250 copies of the hard-cover edition alone.
The paperback edition was published by Penguin and has sold 804,000 copies by 1972, and is now selling at the rate of 150,000 copies a year. Together with 10,000 to 15,000 copies sold in special editions, the sale must now have reached one million copies, give or take a few thousand, an extraordinary figure for a novel that is not designed to please nor all that easy to understand.
Fenwick makes these points about the Secker & Warburg editions of Nineteen Eighty-Four:
– first printing 8 June 1949 was 25,575 copies
– second printing (impression) or what Fenwick calls the “Reissue of First Edition” March 1950 had 5,570 copies
– third printing (impression) or “Second and third Reissues of First Edition” April 1950 (5,150 copies*)
– fourth printing November 1950 (5000 copies)
– Reset Second edition December 1950 (4975 copies)
This does not match with the actual detail in the copies I own. What seems odd is that the ‘reset’ (printed December 1950) only has thee other printings above it on the copyright page when it should have four according to Fenwick. Do you get what I mean? Fenwick is very specific with dates and print runs but she is far from infallible. Her detail does echo Warburg’s though. Both must be wrong.
Read the following pages (135-36) of Fenwick’s bibliography to see the detail:
In summary: if one reads Warburg’s autobiography and Fenwick, an interesting anomaly arises while looking at the various copyright pages in the editions between 1949-55. There are in fact four printings, according to these two sources, prior to the “reset” in December 1950 but if you look closely you will see that this is not represented on the copyright page. If one looks at the copyright page of the 8th printing (which is a film tie-in) there are only three printings above the reset too.
It is interesting that Fenwick reports that that 3,150 of that “second reissue” or third printing (April 1950) became a “Star Editions” (August 1950) stamped “TO BE SOLD ON THE CONTINENT OF EUROPE ONLY” with a “purplish red advertisement for Star Editions” on the back flap. The Orwell Archive copy that Fenwick mentions cannot be located by the archivists and the only copy I could find to purchase is from 1951 (not mentioned in Fenwick). I suspect an error and possibly there is not a 1950 edition which may explain the other error. Either way, my copy has 2+2= and the copyright page reveals the same order of printings. *UPDATE (19/6/2020) A rare bookseller contacted me who had come into possession of the third printing mentioned above. Fenwick is correct. As expected, it has 2+2=5 on page 290.
Davison writes (1998) correctly about the publication history of the novel saying:
Nineteen Eighty-Four was published by Secker & Warburg on 8 June 1949. It was published five days later by Harcourt, Brace and Company in New York. Secker & Warburg printed 26,575 copies for the first edition; a second impression, of 5,570 copies, was issued in March 1950, and a third impression, of 5,150 copies, in August 1950. A second edition, entirely reset, was ordered in December 1950.
but demurs in his note for Taylor’s, The Annotated Edition (2013) with his claim about the ‘serious flaw’ considering two printings prior to 1951 had dropped the ‘5’.
A serious flaw occurred in the 1951 printing of the Seeker & Warburg text. The ‘5’ in the formula ‘2+2=5 (page 334) dropped out of the printer’s forme. All English editions thereafter, including the special 1984 editions prepared by Secker & Warburg and Penguin Books, have repeated this error. Lack of the ‘5’ negates Orwell’s point that Winston has submitted without reservation to Big Brother and thus there can only be hope in the proles (as Winston wrote, page 80), not in the likes of Winston – the relative intellectual.
David Taylor writes (2019):
Secker had printed a first edition of 25,575 copies, followed this up with a second impression of 5,570 copies, and then ordered a further 5,150. In America, Harcourt, Brace and Company began with an initial print run of 20,000 copies, and two follow-up impressions of 10,000 each. The US Book of the Month Club edition, which appeared in July, sold 190,000 copies in a little over eighteen months. By the early autumn there were Spanish, Spanish, Swedish, Japanese, French, and Danish versions in the offing. A German translation was published in the recently established intellectual monthly Der Monat. Meanwhile, American enthusiasm for the novel produced, in quick succession, an NBC broadcast, a Reader’s Digest condensed version, and proposals for a Broadway play. Orwell had no objection, ‘although I should not have thought it lent itself to stage treatment’. On the other hand, he presciently remarked, ‘I should think it ought to be filmable’.
and goes on to say:
There is at least a possibility that some of these glimmers may pulse through Winston Smith. In the original UK Secker & Warburg edition, as Winston sits listlessly playing chess in the novel’s final chapter, he traces in the dust of the table-top ‘2 + 2 =’. As the marked-up typescripts, the corrected proofs, and all the US editions contain the equation in its completed form (‘2 + 2 = 5’), this is usually assumed to have been a printer’s error. As Peter Davison remarks in his authoritative edition of the text, ‘Lack of the “5” negates Orwell’s point that Winston has submitted without reservation to Big Brother’.
It is worth noting that the German journal Taylor mentions, Der Monat, was founded by Melvin J. Lasky, an American journalist and intellectual on the anti-Communist left. This serialised version of 1984 concludes with 2+2=5.
Fenwick’s bibliography differs from information found in Warburg, Davison, Taylor and Glover. I am still trying to uncover the finer detail of where they are all mistaken or have made careless factual errors.
What do you know about the publishing history that is not mentioned above?
Most, not all, editions of Nineteen Eighty-Four published in languages other than English or outside the UK conclude with Winston writing 2+2=5 in the dust. Editions (mostly first printings) that have been checked (by either Chris Angel or myself) and have 2+2=5 include American; Argentinian; Austrian; Canadian; Dutch; Esperanto; French; Hebrew; Japanese; Spanish; Swedish and Turkish publications. The first legal Hungarian edition, published in 1989, does have 2+2=.
One interesting variation, found in German editions of Nineteen Eighty-Four, is 2×2=5. In fact, I am yet to find a German edition published prior to 1987 (except for the serialised version of 1984 in Der Monat) that has anything other than this variation. However, I recently downloaded an ebook for my Kindle in German that was published in 2009 and it has 2+2=.
A final thought…
The novel was first adapted for television in 1953 by CBS in New York and screened live as part of the Studio One anthology. There is a very simple reason for why it concludes with Winston writing 2+2=5. American editions of the novel have never ended any other way.
Orwell’s second wife, Sonia Brownell, felt strongly that the novel needed to end with 2+2=5 for textual integrity to be maintained. I am not so sure. It is likely the appendix is signalling hope from a future age where it appears that Big Brother has been defeated. The victory can possibly be attributed not to a resistance group but language itself; Standard English is again the norm. It does make sense then that Winston is still capable of thoughtcrime at the end of the novel. Crick’s suggestion, that the “5” may have been dropped – as Winston “is now open to take anything they say” – also makes sense. You may find it interesting to watch Brownell argue her case in an interview with Bernard Crick filmed in 1965.
What do you think? Which works best, 2+2= or 2+2=5? There is something deliciously ironic in the impermanence of text considering Winston’s job at the Ministry of Truth.
Thanks to Dennis Glover for alerting me to the detail of this textual controversy and to Chris Angel for his observations regarding foreign editions of the novel (and checking his copies so diligently plus supplying photographic evidence). David Ryan’s detailed understanding of the film and televisions adaptations is also greatly appreciated.
Fenwick, Gillian, George Orwell: a Bibliography, Oak Knoll Press & St. Paul’s Bibliographies, 1998
Glover, Dennis, The Last Man in Europe – A Novel, Black Inc., 2017
Nineteen Eighty-Four, Secker and Warburg, June 1949 (26,575 printed)
Nineteen Eighty-Four, Harcourt, Brace and Company, June 1949 (20,000 copies printed)
Nineteen Eighty-Four, Secker and Warburg, March 1950 (5570 copies)
1984, New York: Signet, 1950 (first American paperback edition Signet #798)
Nineteen Eighty-Four, Secker and Warburg, 1950 (fourth UK impression “reset” with Orwell’s death on the back cover in December 1950 with 4975 copies printed)
Nineteen Eighty-Four, Secker and Warburg, 1951 (NB According to Fenwick, 5150 copies copies printed of this third impression in UK of which 3150 were bound up for the Star Editions in 1950 but this is a 1951 edition. I think Fenwick is wrong but cannot check as the Orwell Archive copy is missing?)
Nineteen Eighty-Four, Secker and Warburg, 1951 (Reprinted July 1951 and usually sold as a ‘fifth impression’ with 4975 copies printed)
Nineteen Eighty-Four, London: Penguin, 1954 (first UK paperback edition)
Nineteen Eighty-Four, Secker & Warburg, 1955 (Uniform Edition)
Nineteen Eighty-Four, Secker and Warburg, 1955 (eighth UK impression with 1956 film tie-in cover)
Nineteen Eighty-Four, Oxford: Clarendon Press, 1984 (with an introduction and annotations by Bernard Crick)
Nineteen Eighty-Four – The Annotated Edition (2013) Penguin Classics (edited by DJ Taylor)
Nineteen Eighty-four, Black Inc, 2017,
Neunzehnhundertvierundachtzig, Zürich: Diana Verlag, 1950
Neunzehnhundertvierundachtzig, Zürich: Alfons Bürger Verlag, 1950. 231 pp. Translated by Kurt Wagenseil, cover art, Kurt Hilscher
Neunzehnhundertvierundachtzig, Zürich: Diana Verlag, 1964
1984, Frankfurt: Büchergilde Gutenberg, 1984
1984, Ullstein, 2009
1984, Európa Könyvkiadó, 1989 (First legal Hungarian publication of Nineteen Eighty-Four)
1984, Barcelona: Destino, 1952 (First Spanish edition)
1984 (Ediciones Kraft), Buenos Aires: Guillermo Kraft LTDA, 1954 (An Argentinian edition of Nineteen Eighty-Four translated into Spanish)
1984, Işık Kitapları, 1958 (First Turkish edition)
Mil naŭcent okdek kvar ĉe, Mondial, 2012, 288pp (Esperanto)
1984, Gallimard, Paris, 1950 (First French trade edition)
Ryan, David, George Orwell on Screen, MacFarland & Company Inc., 2018. 247pp.
Taylor, DJ, On Nineteen Eighty-Four: A Biography, Harry N. Abrams, 2019. 256pp.
Warburg, Frederic, All Authors Are Equal, St Martin’s Press, 1973
Willison, I.R., George Orwell: some materials for a bibliography, 1953
Willison, I.R., The history of the book, author bibliography, and literary biography: The case of George Orwell, Publishing History, 2012