In January 2008, Quercus published in the United Kingdom the first English-language edition of Stieg Larsson’s Män som hatar kvinnor (Norstedts Förlag, 2005) as The Girl with the Dragon Tattoo, translated by Reg Keeland. That name is a pseudonym. The translator was Steven T. Murray (1943–2018), an American who translated from Swedish, Danish, Norwegian and German, and who had previously published translations under his own name, including Henning Mankell’s Sidetracked (1999), which won the UK’s Gold Dagger Award in 2001. The pseudonym was not imposed on him. It was a choice he made, and the reasons he gave for it are the subject of this essay.
The story that gets told about Reg Keeland is usually a story about a translator being erased. That framing is half right and half wrong. Murray was not forbidden his own name. He withdrew it. The distinction matters because it changes what kind of act the pseudonym was: not a silencing but a protest, and a very specific one, aimed at a very specific editorial practice.
What Murray said
In a 2010 interview with Joanne Matzenbacher, reprinted by SouthWest Writers, Murray explained the general principle first: translators should get credit for their work, which means having their names on the cover or title page and being mentioned in publicity. Then he described the exception:
But if the editor or publisher makes major changes to the text without the translator’s permission, so that the final English version no longer reflects our work, then we sometimes (reluctantly) choose to use a pseudonym.
And then he applied it to the Millennium books:
In the case of the Millennium books, I wasn’t given enough time to go over the final editing, and I also didn’t agree with many of the changes that had been made.
This is the whole of the documented case. There is no evidence that Murray was legally barred from using his name, and no evidence that the publisher demanded the pseudonym. The sources attribute the choice to Murray. What the sources do not tell us is who made any particular edit, or how many edits there were. Murray’s account is one translator’s account, given in an interview, and it should be read as such.
The sentence he quoted
Murray gave one example, and it is worth reading slowly. In a scene where Blomkvist is at his sister’s house and she asks how he is doing, Larsson wrote, in Murray’s rendering, “I feel like a sack of shit.” The published English text, according to Murray, reads: “He told her he felt as low as he had in life.”
Set the two sentences side by side and the loss is not a loss of meaning. Both say the man is miserable. The loss is a loss of register. “I feel like a sack of shit” is direct speech, first person, blunt, self-lacerating, and slightly funny in the way that men being honest about their own wreckage are sometimes funny. “He told her he felt as low as he had in life” is indirect speech, third person, grammatically smooth, inoffensive, and generic. It reports the feeling instead of letting the reader hear it. It also moves the sentence one step away from Blomkvist: we are no longer in the room with him; we are being told about the room.
A reader can test this without knowing a word of Swedish. Ask which sentence sounds like a man talking to his sister. The first does. The second sounds like a narrator summarizing a man talking to his sister. That is a real difference, and it is the kind of difference that accumulates across five hundred pages into a voice.
What the sources do not establish is who made this change. Murray says it was “arbitrarily changed”; he does not name the editor. Christopher MacLehose, the publisher who acquired the books for Quercus and its MacLehose Press imprint, has described his own extensive re-editing of the translation, but no retrieved source attributes this specific line to him. The honest reading is: a translator quoted a rendering pair as evidence of editorial intervention he did not agree with, and the publisher has separately confirmed that intervention occurred.
The publisher’s side
MacLehose has never hidden what he did. In a 2010 interview with Anthony Gardner for europe.org.uk/culture, he said he bought the Larsson books “on the understanding that they could be edited without interference from the translator” and devoted months of his own time to polishing them. In a 2013 interview with the Sydney Morning Herald, he stated the principle behind it:
I do think that almost every translation of a certain literary density has to be treated like an original text. If you had the author, you would make suggestions. We didn’t have the author, but that shouldn’t stop you making the sentences more interesting for the reader.
This is a coherent position, and it is not the same as saying the translation was bad. MacLehose told Gardner that the first English version was produced in only eleven months, at the instigation of a film company that wanted it in a hurry. He also told the SMH that the American-English translation he received was “clearly quite hurried.” His argument is that a translation of a dense literary text needs an editor the way an original manuscript needs an editor, and that with the author dead, the publisher becomes the only party who can play that role.
Murray’s position is equally coherent: a translation is the translator’s work, and if the publisher rewrites it without the translator’s consent, the translator should not have to sign it. The dispute is not really about accuracy. It is about authority. Who is the final author of the English sentence? MacLehose says, in effect, the publisher, when the author is unavailable and the schedule is tight. Murray says, in effect, the translator, or else no one signs.
Both men were working within the same constraint, which was time. Murray said the trilogy took about eleven months, and that each book was more than 900 pages in manuscript, “equivalent to translating six regular sized novels.” MacLehose said the schedule was set by a film company. Neither account contradicts the other. They describe the same pressure from two sides.
The title as the first translation decision
The pseudonym is the most visible sign of the disagreement, but it is not the only one. The Swedish title is Män som hatar kvinnor, literally “Men Who Hate Women.” The English title is The Girl with the Dragon Tattoo. Murray said the British publisher acquired world English rights and “changed the titles of books 1 and 3” because he “thought the books needed titles that would have greater commercial appeal.” MacLehose told Gardner that he regretted the change: “I was perfectly happy to use it, but the message came back from the bookshops, ‘Please don’t.'”
A title is a translation decision made before the reader opens to page one. “Men Who Hate Women” announces a thesis about Swedish society and about violence against women. “The Girl with the Dragon Tattoo” announces a character and an image. Both are accurate to the book in different ways, but they prime different novels. One reader picks up a social indictment; the other picks up a thriller about a young woman with a tattoo. The book contains both, but the frame is not neutral, and the frame was chosen by the publisher, not the translator, and not the author, who was dead.
What the credit line is for
The practice on this site is to name the translator, the publisher and the edition year in the opening paragraph of any piece about a translated work, and to quote at least one rendering for comparison. That practice can look like bibliographic fussiness. It is not. It is the minimum condition for reading a translated novel honestly, because the credit line is itself evidence.
Consider what a reader can learn from the front matter of the 2008 Quercus edition. The title page says “Translated by Reg Keeland.” The copyright page identifies the Swedish original and the translation copyright. A reader who knows nothing else now knows three things: the book was written in Swedish, it was translated by someone using a pseudonym, and the pseudonym has a documented reason. That is not trivia. It is the difference between reading a book as a transparent window onto Larsson’s Swedish and reading it as an English text with a history.
The rendering pair does the same work at the level of the sentence. “I feel like a sack of shit” versus “He told her he felt as low as he had in life” is not a scandal. It is a small, legible piece of evidence that the English text was edited by someone other than the translator, in ways the translator did not endorse. A reader who has seen that pair will read the rest of the book differently, alert to the smoothness of sentences that might once have been rougher, and to the possibility that the voice they are hearing is a composite.
How to read a translated novel with its credit line intact
This is a method, not a verdict. It does not require you to decide whether the English Girl with the Dragon Tattoo is good or bad. It requires only that you read the front matter before you read the novel.
First, find the translator’s name on the title page. If it is a pseudonym, ask why. Sometimes the answer is in an interview, as it is here. Sometimes it is not documented at all, and the absence is itself worth noting.
Second, note the publisher and the edition year. A translation is a dated object. The 2008 Quercus text is not the same object as a later reissue, and a US edition may differ from a UK edition. When you quote a translated novel, quote the edition you actually read.
Third, find one sentence and compare it to the original if you can, or to a second translation if one exists. You do not need Swedish to notice register. You need only to ask whether the sentence sounds like a person speaking or like a narrator reporting. That question will tell you more about the translation than any review.
Fourth, treat the title as part of the translation. Ask what the original title meant and what the English title does instead. The answer is often a small essay in publishing history.
What the pseudonym was
Reg Keeland is not a man who was erased. He is a name a translator chose to publish under when he decided that the text published under his own name would no longer be his. That is a strange and specific act, and it is worth preserving in the record exactly as Murray described it: reluctantly, as a last resort, when the editor or publisher makes major changes without permission. The name on the cover is the one piece of credit a publisher cannot silently rewrite. Withdrawing it is the translator’s remaining lever.
The English Girl with the Dragon Tattoo is a real book, and it has readers who love it. Nothing here argues that they shouldn’t. What the record shows is that the book they love is a composite, made under time pressure, edited by a publisher who believed a translation should be treated like an original text, and signed by a translator who believed it should be treated like a translation. Both beliefs are legible in the front matter and in the sentences. Reading with the credit line intact does not resolve the disagreement. It lets you see it, which is the most a reader can ask.
Questions readers ask
Was Steven T. Murray legally prohibited from using his own name?
No retrieved source supports that. Murray described the pseudonym as a voluntary choice, made reluctantly, when a publisher’s changes meant the published text no longer reflected his work.
Did the publisher demand the pseudonym?
The retrieved sources attribute the choice to Murray, not to Quercus or MacLehose Press. MacLehose has confirmed that the translation was re-edited, but not that he required a pseudonym.
Who changed “I feel like a sack of shit” to “He told her he felt as low as he had in life”?
The sources do not say. Murray quoted the pair as an example of changes he did not agree with. MacLehose has described re-editing the translation but has not been documented as the author of that specific line.
Is the English translation inaccurate?
That is not what the evidence shows. The documented case is about register and editorial authority, not factual error. One quoted pair is not a systematic comparison, and readers should not treat it as one.
Why does the translator’s name matter if the book is good?
Because the name tells you whose English you are reading and under what conditions it was made. A translation is a dated, edited object with a history. The credit line is the shortest reliable record of that history.
Recent Comments