More unquotations from the New Yorker

« previous post | next post »

Janet Malcolm and Jonah Lehrer are not the only New Yorker writers who have been accused of fabricating quotations. A more recent case involves a piece by Jared Diamond, "Annals of Anthropology: Vengeance is Ours", 4/21/2008 (abstract on the New Yorker's web site here).  Diamond's article led to a long series of negative responses at iMediaEthics and Savage Minds, as well as a $10M libel suit (which as far as I can tell is still pending).

For a simple, clear and (I think) fair account of the facts and issues in the Diamond case, see Michael Balter, "'Vengeance' Bites back At Jared Diamond", Science 5/15/2009. Here's the start of Balter's account, giving the general background:

In April 2008, well-known biologist and author Jared Diamond penned a dramatic story in The New Yorker magazine, a violent tale of revenge and warfare in Papua New Guinea (PNG). Titled “Vengeance is Ours” and published under the banner “Annals of Anthropology,” the 8000-word article tells the story of a clan war organized by a young Papua New Guinean named Daniel Wemp to avenge the death of Wemp's uncle, Soll. In Diamond's telling, the war started in the 1990s over a pig digging up someone's garden, went on for 3 years, and resulted in the deaths of 29 people. In the end, Diamond wrote, Wemp won: His primary target, a man Diamond referred to as “Isum,” had his spine cut by an arrow and was confined to a wheelchair. Diamond juxtaposed Wemp's story with that of his own father-in-law, a Holocaust survivor who never exacted retribution for the loss of his family, to draw an overall lesson about the human need for vengeance. […]

The lawsuit, which also names as a defendant Advance Publications Inc., the owner of The New Yorker, demands at least $10 million in damages. It follows a yearlong investigation led by Rhonda Roland Shearer, an artist and the widow of evolutionary biologist Stephen Jay Gould. […]

As Balter notes, many quite basic facts are at issue:

Shearer and her colleagues, who included three researchers in PNG, claim that Diamond and The New Yorker got many important facts wrong in the original article, including the contentions that Wemp had personally organized the warfare, that Soll was his uncle, and that Mandingo had been paralyzed by an arrow. Indeed, the Stinkyjournalism.org report includes a recent photograph said to be of Mandingo standing and looking strong and healthy. […]

At least one other Papua New Guinean supports the account of Shearer's team. “Diamond's article is a confused story that names real places and persons but mixes up false, wrong, and defamatory allegations that bring into disrepute the good name of the named clans and their members,” said Mako Kuwimb, a member of Wemp's Handa clan and a PNG attorney now doing graduate work at James Cook University in Queensland, Australia. […] On 21 April of this year, Kuwimb sent The New Yorker's publisher, Lisa Hughes, a detailed, 30-page refutation of the Diamond article. Among Diamond's biggest errors, Kuwimb told Hughes, were his statements that the war he described had begun with the “pig in the garden” episode and had lasted 3 years. Kuwimb contends that the war was sparked by a gambling dispute and lasted only a few months.

The plaintiffs (Henep Isum Mandingo and Hup Daniel Wemp) have submitted an amended complaint that claims even more striking divergences between the New Yorker story and the facts, e.g.

f. Daniel Wemp was not a participant in this war at  all.  At the time of the fighting, Wemp was working some 200 miles away at the coast, in a city called Madang.  He only learned of the fighting after it was over.

g. Daniel was not injured in the fight, but received an injury some ten years earlier when he was 11 or 12 while he was watching another fight between two different clans.

h. Isum was not wounded in revenge for the death of Soll.  Soll was still alive, although wounded, then Isum was wounded.

But our concern here is not with the facts but with the words. In this case, unlike the Malcolm and Lehrer cases, there is little in the way of objective evidence about the wording of the contested quotations. Janet Malcolm's unquotations from Jeffrey Moussaieff Masson allegedly came from 40 hours of tape recorded interviews, and could also be compared to her extensive notes from those (and other?) interviews. Jonah Lehrer's unquotations from Bob Dylan allegedly came from available books, movies, interviews, and so on.  In contrast, the alleged quotations in Jared Diamond's article are based on his memories of discussions in 2001 when he employed Wemp as a driver in PNG, and on the notes he took during a 2006 interview with him.

Here are the first two paragraphs of Diamond's article:

In 1992, when Daniel Wemp was about twenty-two years old, his beloved paternal uncle Soll was killed in a battle against the neighboring Ombal clan. In the New Guinea Highlands, where Daniel and his Handa clan live, uncles and aunts play a big role in raising children, so an uncle’s death represents a much heavier blow than it might to most Americans. Daniel often did not even distinguish between his biological father and other male clansmen of his father’s generation. And Soll had been very good to Daniel, who recalled him as a tall and handsome man, destined to become a leader. Soll’s death demanded vengeance.

Daniel told me that responsibility for arranging revenge usually falls on the victim’s firstborn son or, failing that, on one of his brothers. “Soll did have a son, but he was only six years old at the time of his father’s death, much too young to organize the revenge,” Daniel said. “On the other hand, my father was felt to be too old and weak by then; the avenger should be a strong young man in his prime. So I was the one who became expected to avenge Soll.” As it turned out, it took three years, twenty-nine more killings, and the sacrifice of three hundred pigs before Daniel succeeded in discharging this responsibility.

If you have an ear for the English language, you may be struck by the depiction, in the second paragraph, of a PNG tribesman expressing himself verbally in phrases reminiscent of William F. Buckley in his prime:

"Soll did have a son, but he was only six years old at the time of his father’s death, much too young to organize the revenge. On the other hand, my father was felt to be too old and weak by then; the avenger should be a strong young man in his prime. So I was the one who became expected to avenge Soll."

According to Rhonda Shearer ("Jared Diamond's Factual Collapse", 4/21/2009), Daniel Wemp made exactly this point in a recorded conversation with New Yorker fact checker Chris Jennings on 8/21/2008 (four months after the New Yorker article was published):

In his interview with Jennings on August 21, 2008, Daniel said, “The words that I have spoken during that time I can’t actually remember what words I have spoken, all these notes that he has taken [inaudible] …but I see that the English [inaudible]…on the article is not good enough for such person like me [inaudible]. It is a perfect English which is written on the article.”

Jennings’ answered, “Okay. So, you think that the words may not sound like your own words.”

"Daniel Wemp" is quoted in a similarly elevated and writerly register throughout the New Yorker article. Thus:

When I asked Daniel how the war that claimed his uncle’s life began, he answered, “The original cause of the wars between the Handa and Ombal clans was a pig that ruined a garden.” Surprisingly to outsiders, most Highland wars start ostensibly as a dispute over either pigs or women.

Or

On the first day of the resulting fight, Daniel was wounded. “I was advancing in battle with my biological father, who was holding a shield to protect me, while I myself held the weapons,” he told me. “As my father and I went up a hill towards a stone quarry from which the Ombal enemy was throwing stones as well as spears, a stone hit my father on his leg. So I took the shield to protect my father, and I told him to go faster. That was when I was left unprotected, and an Ombal spear struck me on the back of my lower left leg.” He showed me an inch-long scar and explained apologetically, “If, in a fight, you receive a wound on your forehead, then you are considered to have done well, but if you only have a spear wound on the back of your leg, like this one of mine, then you are viewed as not having fought well.”

Is it plausible that these are really direct quotations from Daniel Wemp? As we noted back in 2009, that issue is addressed quantitatively by Douglas Biber, "Did Daniel Wemp really say that? Using corpus linguistics to evaluate the likelihood that Jared Diamond's reported quotes in The New Yorker were ever spoken", iMediaEthics 6/2/2009:

The language of conversation is dramatically different from the language of academic writing.  Some of the differences between the two are obvious to all of us, such as contractions and incomplete sentences.  However, many other grammatical differences are much more difficult to detect.

Over the last 25 years, a research approach has been developed for the empirical analysis of such grammatical characteristics. Referred to as ‘corpus linguistics’, the approach is based on the analysis of very large collections of natural texts from thousands of individual speakers and writers. Computer programs aid the analyses, which result in descriptions of grammatical features that occur frequently, features that are typical, and features that rarely occur. In addition, by comparing corpora with different kinds of texts, it is possible to contrast the grammatical characteristics that are usually found in conversation to those usually found in academic writing (or any other spoken or written varieties).

Doug applied these techniques to an analysis of the Wemp "quotations" in Diamond's New Yorker article:

For various reasons, scholars have questioned the appropriateness and accuracy of the information in this article. The question investigated with corpus analysis was more narrow: how likely, or unlikely, is it that the quotes cited in this article were actually produced in speech? That is, how likely is it that Daniel Wemp said these exact words?

In this case, the analysis began with a linguistic description of Diamond’s quotes (i.e., the quotations attributed to Daniel Wemp as his spoken words in the 4/21/08 New Yorker article). Those quotes were compared to an independent record of Daniel Wemp’s actual speech, based on verbatim transcripts of spoken interviews collected by Rhonda Roland Shearer. Quantitative analyses of the New Yorker quotes and the Daniel Wemp (DW) transcripts were carried out, to identify grammatical features that were frequent or rare.

Then, those quantitative findings were compared to previous large-scale corpus analyses of conversation and academic writing, to determine whether the New Yorker quotes were typical of the language normally used in conversation. The results were surprising: in many respects, the New Yorker quotes are much more similar to the language typically used in academic writing than to normal conversation.

Here's his tabular summary of the results (cited rates are per million words):

Grammatical feature Rate in actual conversation Rate in Diamond quotes Comparison of the two
Attributive adjectives (e.g., original, biological) c. 15,000 c. 30,000 2 times more frequent in Diamond quotes
Preposition of c. 12,000 c. 34,000 3 times more frequent in Diamond quotes
Noun post-modifier complexes (e.g., The original cause [of the wars [between the Handa and Ombal clans] ]) c. 500 c. 3,000 6 times more frequent in Diamond quotes
Noun phrases with both pre-modifiers and post-modifiers (e.g., a stone quarry from which the Ombal enemy was throwing stones) c. 500 c. 4,000 8 times more frequent in Diamond quotes
‘extraposed’ to-clauses controlled by an adjective (e.g., it’s not acceptable [to set fire to the hut]) c. 100 c. 2,000 20 times more frequent in Diamond quotes
Noun + to-clause (e.g., the opportunity [to see who really are the best marksmen]) c. 50 c. 1,200 25 times more frequent in Diamond quotes
Adjective and/but adjective (e.g., tall and handsome)

c. 20 c. 2,000 100 times more frequent in Diamond quotes
Preposition + Relative pronoun (e.g., each battle in which we succeeded in killing an Ombal) c. 20 c. 2,000 100 times more frequent in Diamond quotes

And his conclusion:

These comparisons show the magnitude of the discrepancies between the grammatical style of normal conversation contrasted with the grammatical style of the Diamond quotes. To find one of these grammatical features in a normal conversation is noteworthy. To find repeated use of this large constellation of features in actual spoken discourse, some of them occurring c. 100 times more often than in normal conversation, is extremely unlikely. In contrast, these are all features that are typical of academic writing, suggesting that they have their origin in writing rather than actual speech.

Other corpus studies (e.g., the book University Language; Biber, 2006) have shown that these same features are rare and exceptional in even academic speech, including university lectures. In contrast, what we find in the Diamond quotes is the pervasive use of a suite of grammatical constructions, which are all rare in conversation but common in formal writing. This constellation of grammatical characteristics is also strikingly different from the grammatical style of the verbatim transcripts of speech produced by DW. In sum, the analysis strongly indicates that the Diamond quotes are much more like discourse that was produced in writing, reflecting the typical grammatical features of formal academic prose, rather than verbatim representations of language that was produced in speech.

My conclusion? When you see a passage in quotation marks in a New Yorker article, you should not expect it to be a truthful representation of anything that the alleged speaker ever actually said. Rather, you should take it as the author's expression of what they want you to believe that the speaker meant. In some cases, these unquotations are "poetically true", that is, they give an insightful impression of the speaker's feelings and attitudes, although the writer knows that the words are not original. In other cases, the unquotations are an honest misrepresentation, in the sense that they're genuinely what the author understood (or at least, remembers) the speaker to have meant to say. And sometimes, the unquotations are completely fictional, in the sense that the author doesn't care at all what the alleged speaker either said or meant, but puts words in their mouth in order to advance the narrative.

This same uncertainty, unfortunately, seems to be characteristic of journalistic practice as a whole. Surprisingly, journalists often make up quotations even when recordings and transcripts of the misquoted material are publicly available. (See links here for illustrative examples.) Some publications have a theoretical commitment to accurate quotation — though as far as I know, the New Yorker does not — but it's rare to hear of a journalist suffering significant personal or reputational damage for misuse of quotation marks.



49 Comments

  1. The Ridger said,

    August 4, 2012 @ 9:31 am

    Was Wemp meant to be speaking English? If not, the quarrel may be with register in translation.

    [(myl) Yes, all relevant discussions and interviews took place in English.]

  2. Sarang said,

    August 4, 2012 @ 9:50 am

    I wonder what the norms are re: correcting grammar, etc. in interviews w/ people who are not native speakers of English, or who speak an unfamiliar dialect. Even with native speakers one assumes that interviewers rework the material in their tapes into complete sentences — and when the original speech is v. far from standard American English, presumably the only thing to do is to rewrite the sentences from scratch. The fact that Diamond's paraphrase sounds weirdly stuffy seems more a stylistic lapse than anything else. I don't see an issue here at all…

  3. Martyn Cornell said,

    August 4, 2012 @ 10:00 am

    … journalists often make up quotations even when recordings and transcripts of the misquoted material are publicly available.

    Often, Mark? Really? You've detailed some pretty egregious quote-fakery by some journalists, but you've given no evidence at all to suggest that making up quotes is widespread in the profession. In nearly 40 years in journalism, including national newspaper journalism in the UK and abroad, I can remember no incident of complaints about made-up quotes against any of the hundreds of journalists I have worked with.

    [(myl) I got interested in this issue in the summer of 2005, because I happened to notice that different papers reported a particular basketball player's particular postgame remarks in very different ways. During a period of a couple of months after that, I checked journalistic quotations from time to time against available recordings or transcripts:

    "What did Rasheed say?", 6/23/2005; "Ipsissima vox Rasheedi", 6/24/2005; "Ritual questions, ritual answers", 6/25/2005; "Down with journalists!", 6/27/2005; "Bringing journalism into the 21st century", 6/30/2005; "More comments on quotes", 7/1/2005; "Linguists beware", 7/9/2005; "Quotes from journalistic sources: unsafe at any speed", 7/9/2005; "'Quotations' with a word error rate of 40-60% and more", 7/30/2005; "This time it matters", 8/13/2005; "'Approximate' quotations can undermine readers' trust in the times", 8/27/2005.

    I never found a single case where the alleged quotations were even close to being verbatim.

    In many if not most of these cases, it was obvious that no one really cared. In some cases, though, it mattered quite a lot. I saw no evidence that more care was exercised when it mattered than when it didn't — all of the "quotations" were highly inaccurate — omitting and changing words, inserting paraphrases, or putting together into the same sentence things that were widely spearated in time. In terms of the metric used in speech recognition research, the "word error rate" was typically 50-70% or worse.

    Whether the result was a trivial misquoting of a ritual response to a ritual question in a postgame interview, or a highly consequential misrepresentation of the carefully-expressed position of a major political figure, the motivation (as far as I can see) was always the same — to make it easier for the journalist to slot into the story an alleged quotation to support the point that they wanted to make.

    So I concluded that this was the standard practice of journalists, namely to use "approximate" quotations, taken from their notes and/or memories, and edited to suit the story; and that I was naive to expect that words in quotation marks should normally correspond to what the alleged speaker actually said. And therefore I stopped checking, except in cases where the exact wording was an issue for some reason or another. But I have every reason to think that what I observed during that period was typical of journalistic practices before and since. Do you really have any evidence to the contrary?]

  4. languagehat said,

    August 4, 2012 @ 10:33 am

    You've detailed some pretty egregious quote-fakery by some journalists, but you've given no evidence at all to suggest that making up quotes is widespread in the profession.

    This is what bothers me. Mark's war on fake quotes is laudable and the journalists and publications that perpetrate them should be ashamed of themselves (and, ideally, made to stop), but his claim that the practice is "characteristic of journalistic practice as a whole" reminds me of cops who believe that pretty much everyone outside of fellow cops and their families are probably criminals, because they deal with criminals all day long and lose sight of the larger context.

    [(myl) I wish this were true. Do you really think that cases like this one are unusual?]

  5. Clayton Burns said,

    August 4, 2012 @ 10:39 am

    The Liberman comments are compelling.

    Even if not deliberate, imprecision in quoting is pervasive.

    After the Air India judgment, a story appeared that wildly misinterpreted it.

    The misadventures with Obama health care in the Supreme Court point to a readiness to get things wrong.

    But the question I have is: What kind of language issue is it?

    (I think that a good text sample for Lehrer is chapter 3 of "Imagine," "The Unconcealing," on working memory.)

    If the inability to hold a sentence and a line of argument in your mind so as to get it straight in a story is a linguistic/working memory matter, can we find exercises in Mark Ashcraft's "Cognition" that would be helpful?

    There is no inherent limitation of the human mind. You can train yourself to take down longhand complex discourse in court. If you have the power of concentration.

    If journalists are rounding off quotes and failing to capture sequences of ideas, they have untrained minds. What I am interested in is the experiments in cognition that would indicate the best training.

    In 2012, this is an ethical problem, not a cognitive one. What journalists need is not mental exercise in verbal memory, but the commitment to use voice recorders or smartphones as well as notebooks, and to transcribe and quote accurately what they record.]

  6. John McVey said,

    August 4, 2012 @ 10:58 am

    L’Affaire Lehrer (et al) brings to this mind gonzo journalism as practiced by Hunter S Thompson. At least we knew it was outrageous, subjective in the extreme. I wonder if implicit in gonzo was a recognition that journalism as practiced for many years (for ever?) was always subjective, and had aided and abetted plenty of evil. So it raised the ante. Lehrer et al aren't extreme enough — have honed the appearance of being well-behaved and trustworthy. Who in his/her right mind would trust HST!

  7. Dave Bell said,

    August 4, 2012 @ 11:26 am

    In the end, it might be a reliable account of the story that was told (there are other reasons for doubt), but it shouldn't have been dignified by quote marks. This could have been reported speech, but was presented as direct speech.

    It strikes me that Jared Diamond used the gross structures of a storyteller, but never wrote with the voices of the people he "quoted". You read a novel, and you expect the patterns of the narration to differ from the speech of the characters. And the way the characters speak is itself unreal. Their words are more akin to an actor on the stage than to the way people speak without scripts.

    There are no clear dividing lines, but this is neither anthropology nor journalism. Whether or not the tale was true, can any of us imagine talking in that manner, in our ordinary lives, without a script? At best, we might come close with a quotation.

    [(myl) One plausible theory is that Daniel Wemp told Jared Diamond a bunch of partly-exaggerated and mostly-false "war stories", and Diamond embedded these, without checking them, in a personal meditation on the nature of revenge. As far as I can see from the evidence available, this is most sympathetic-to-Diamond theory that is consistent with apparent fact.

    But even if this is what happened, it's still very hard to for me to believe that the quotes attributed to Wemp in the New Yorker article represent word-sequences that Wemp ever actually uttered.]

  8. dw said,

    August 4, 2012 @ 12:10 pm

    When I am conversing with someone, I tend to adjust my register somewhat to fit with that of my interlocutor. Is it possible that at least some of the differences between the conversations Wemp had with Shearer, and those he allegedly had with Diamond, could be caused by differences in Shearer's and Diamond's conversational styles?

  9. Philip said,

    August 4, 2012 @ 12:31 pm

    I wonder if all this suggests a need for punctuation innovation. In addition to quotation marks, perhaps we really should also have paraphrase marks. Is this an impossiible dream?

    [(myl) Good idea! We could adopt guillemets as unquotation marks: «blah blah blah.» ]

  10. Shanna said,

    August 4, 2012 @ 1:08 pm

    I think anyone who has ever been interviewed by a journalist, especially in a more informal setting, knows that the quotes are rarely 100% verbatim. I've been quoted for newspaper articles about a dozen times, and my quotes are never accurate – but they're almost always close enough. They get the gist of what I said even though they don't get my exact phrasing. I'm talking about situations in which the journalist is just taking notes, not tape recording – obviously, with a recording, the expectation for verbatim accuracy should be higher. But really, I thought it was understood practice that quotes in most journalists' pieces were rough approximations of what the person said as the journalist remembers it – they're piecing it together from hastily jotted notes, after all – and I'm not sure why this seems to be all over the news now but has been remarked upon so little in the past.

  11. Rhonda Roland Shearer said,

    August 4, 2012 @ 1:48 pm

    I am lead author of the discovery of Jared Diamond's false account for iMediaEthics of a three-year long bloody tribal war in the Southern Highlands of Papua New Guinea.

    I found Mark Liberman post agreeable on most points but also take exception to the blanket thrown over all of journalism with a general claim that fake quotations are everywhere.

    There is one error I would like to address. In the above post, Mark compares three New Yorker cases, Malcolm, Diamond and the latest, Lehrer.

    As background: Diamond's NYer article deceived readers by back-dating quotes. He claimed in the article that the long and complex Wemp quotations were said to him during long car rides where Wemp was his driver at times during 2001-2002.

    The truth came out after we challenged NYer. Editors soon admitted that the only notes Diamond had –and all the quotations–were from a SINGLE interview done in 2006. The interview was not in a car as Diamond claimed but in a Oil Search company dorm (then Wemp's employer).

    So readers have no idea when reading the NYer piece that Wemp's words are only from 2006 said in a dorm hallway–NOT 2001-2002 in a car ride –as they were mislead to believe.

    So it is misleading when Mark describes the Diamond case : "…the alleged quotations in Jared Diamond's article are based on his memories of discussions in 2001 when he employed Wemp as a driver in PNG, and on the notes he took during a 2006 interview with him."

    The long and very detailed quotes were NEVER claimed by Diamond or the NYer to be based on memory –in his original article or later when trying to defend his work –but based solely on 2006 notes from one interview.

    I would also like to address some of what the commenter have said but fear making too long of a response.

  12. Jeffry House said,

    August 4, 2012 @ 1:51 pm

    I think paraphrase marks are a great idea. My Dad was a highly respected professional journalist for an elite newspaper, for forty years. He certainly practiced something like what Diamond seems to have done. Instead of "Well, you know, that asshole rocketed out at me and I fuckn levelled his ass with a right chop" we might get "He came at me and I hit him with my right hand and he fell to the ground." Part of it was that the paper wouldn't print the racy language anyway, and part of it was a desire for clarity.

  13. Ron Rodgers said,

    August 4, 2012 @ 2:00 pm

    I am not sure why anyone is surprised regarding your conclusions about quotations and The New Yorker. Apply the same methodology to anything in New Yorker writer Joseph Mitchell's Up In the Old Hotel.

  14. Dave Bell said,

    August 4, 2012 @ 2:36 pm

    What might be needed is the quasi-quote.

    One definition is here, in the days of typewriters and duplicator stencils it was a quote-mark overstruck with a hyphen, but I think it can be done with HTML. Let me try…

    "Quasi-quotes are an obscure way of indicating paraphrases and other accurate, but not literal, quotes of what people say." The symbol in HTML can be represented as <strike>"</strike>

    I've seen the origins attributed to Jack Speer: that may be because he documented it in his Fancyclopedia

  15. Teresa Elms said,

    August 4, 2012 @ 2:54 pm

    As a long-time journalist and editor myself, I can confirm that quotations are routinely adjusted for readability and length in print when it does no damage to the perceived sense of the utterance. Writers and editors clean up pauses, stumbles, repairs, repetitions, grammatical infelicities, and erroneous word substitutions within quotations as well as without. We do it because such things are all easily and routinely adjusted through interaction in conversation, but are perceived as final-form in print.

    To do otherwise is to abdicate the interviewer's responsibility to listen. There is no easier way to destroy the reputation of an interviewee than to quote what they say word for word, error for error. As a general rule, interviewees expect to be listened to rather than recorded, and a journalist must respect that expectation if the sense of the interview is to be accurately conveyed to a reader.

    This is more than just a matter of style. It's a matter of capturing the result of an interaction, as negotiated over intervals of time. Anyone who has transcribed spoken conversations for conversation analysis is forced to acknowledge that, in online interactions, utterances are always constructed cooperatively between speaker and hearer to a greater or lesser degree. We may both be perfectly well aware of infelicities and repairs, but the utterance of record that we construct as common ground between us omits these — even if we both make jokes of the errors before we jointly acknowledge the intended sense.

    It is actually very difficult to transcribe speech literally into written form. In a conversation analysis transcript, one strives to record every pause and overlap and repair and gesture of each interlocutor, complete with prosodic inflections and timing data, precisely because all of this literal utterance production is often unheard at a conscious level and therefore constitutes precious data. It usually takes multiple replays of the same recorded phrase to get a reasonably literal rendering pruned of automatic auditory and contextual corrections contributed unconsciously by the transcriber.

    It's clear that Diamond went farther than simple cleanup of the surface utterance in his New Yorker article. But again, this is routine practice in books and in much of the more literary journalistic prose. What Diamond wrote may have conveyed the sense of a gesture, an eyebrow flash, a shrug, or a querulous intonation in the original utterance. These cannot be rendered in print any other way than by altering the words.

    As for the ethics of journalism in such cases, I would take the position that a test exists to determine whether or not a quotation is deceptive. Simply drop the quotation marks and introduce each so-called quotation as a paraphrase instead. If the sense of the segment, when rendered as paraphrase, is an accurate rendition of the incidents being reported, then an editor will feel justified in treating quotation marks as reported speech rather than tape-recorded literal utterance. In that case, we may insist on quotation marks to make the piece feel more interactive and compelling to a reader. We editors are like that. And good writers tend to edit themselves in like manner.

  16. Joe said,

    August 4, 2012 @ 2:58 pm

    At least part of the argument seems to be much ado about convention, and not necessarily about deliberately falsifying facts or words to “advance the narrative.” And I wonder if protesting (too much) journalistic convention in the same context doesn’t detract from the argument claiming indefensible practice of deliberate distortion. But of convention, and backtracking LL posts trying to figure out the assumptions, I came across what appears to be an interesting use of punctuation, if the use of CAPS can be considered punctuation.

    I wondered why the word “you” was capitalized in the LL transcription of the AFR interview (LL, June 16, 2012). Were CAPS used as a substitute for an exclamation point? Is the journalistic use of an exclamation point in a direct quote always an editorial comment? Or is it simply used to indicate the speaker raised his voice above his normal level? And the CAPS served the same purpose? Maybe this is too nitpicky, but the assumption that if we have a recording we have the truth seems a bit naïve. We might also wonder why we should think the 13-minute segment AFR posted wasn’t edited. I don’t think it was, but the point is that a verbatim transcription of the recording still isn’t necessarily the same thing as having been there. Quotes are blended into a narrative and even if accurate can distort, as the selection of an unflattering photo of a candidate might reflect editorial bias. Similarly, careful selection of quotes from an interview can also lead to distortion, even if what is quoted is accurate. And I can’t imagine any other journalist transcribing the AFR segment using the same punctuation as LL used, at which point we would have multiple narratives all claiming to be true. At least some journalistic styles, for example, would insist on correcting what would appear to be fragments in the LL transcription – even if the fragments seemed more true to the speaker’s voice. But the LL transcription, an attempt, apparently, to render a quote exact and accurate, is a good example of why journalistic conventions were created to begin with. That journalists in some cases have stretched those conventions to “advance” a narrative as opposed to just telling the story might simply mean that they are not very good journalists.

  17. Rhonda Roland Shearer said,

    August 4, 2012 @ 3:57 pm

    There is no need for a quasi quote (quote-light?) in my view as a working journalist supervising other journalists. If a quote is confusing, one convention is to paraphrase then next quote a person's words. Most interviews or comments nowadays are provided via email, not the phone anyway–so no easy fudging or mistakes here. You have it in writing.

    Would quasi-quoting paraphrases helped Diamond? I don't see how.

    This quote below, as one example, depicts Wemp as a heartless killer allegedly admitting that he sent assassins to kill Isum, his counterpart leader of the fight. They failed and Isum was only paralyzed (since proven wrong by iMediaEthics):

    Diamond reported Wemp said: “ ‘I felt that it was a matter of ‘kill or else die by suicide.’ I was prepared to die myself in that fight. I knew that, if I did die then, I would be considered a hero and would be remembered. If I had personally seen the arrow go into Isum, I would have felt emotional relief then. Unfortunately, I wasn’t actually there to see it, but, when I heard that Isum had been paralyzed, I thought, I have everything, I feel as if I am developing wings, I feel as if I am about to fly off, and I am very happy. After that battle, just as after each battle in which we succeeded in killing an Ombal, we danced and celebrated and slaughtered pigs. When you fight with thinking and finally succeed, you feel good and relieved. The revenge relieves you; now it can be your turn to help someone else get his own revenge.’ ”

    The above is not just a matter of misquotation. What Diamond claims Wemp said (Wemp denies it) makes Wemp a criminal and at risk of being arrested for a capital crime in Papua New Guinea.

    Diamond admits to Balter in the Science article mentioned in the blog post above that he never tried to contact Isum (nor police, the embassy, missionaries or experts on the Wola area –we asked). Diamond had no desire and made no actions to confirm that Isum committed criminal acts that Diamond described through quotations. Yet Isum–someone Diamond never met or spoke to– was never in a wheelchair with a cut spine even though Diamond stated he was in the Wemp quotes. (All proved untrue by iMediaEthics. We have Isum's medical files) .

    So why would an elite scholar and scientist do NO fact checking of an unpublished war and write a major "factual account" about killings and other crimes in the New Yorker? His sole defense was that he based his entire story on what his World Wildlife Fund driver with 2-years high school's account told him and he quotes. But the deception started with the fact that the quotes were not taken when or where Diamond first claimed they were spoken by Wemp.

    What is the difference is terms of trustworthiness in this case if Diamond paraphrased or quoted? Thanks for allowing me to comment.

  18. Keith M Ellis said,

    August 4, 2012 @ 4:06 pm

    I like the discussion of "paraphrase marks" because I think this is not an obscure, technical problem specific to journalism, but a more general one. Indeed, when we write informally, as opposed to formally in journalism or whatever, we're much more likely to not have records of speech we've heard and we're much more likely to paraphrase. But this is a problem because, as far as I can tell, we normally quote paraphrasings.

    Sometimes when I do this I'll preface it with "something to the effect of", as in "…he said something to the effect of 'blah blah blah'", making clear that it's a paraphrase, even though it's quoted. I don't know how else to write such a statement.

  19. a George said,

    August 4, 2012 @ 4:42 pm

    — a structure that would not break the style too much might be to write after a quotation between marks, 'or words to that effect'. But if a whole dialogue is made up it becomes unwieldy.

  20. Melissa said,

    August 4, 2012 @ 4:45 pm

    It also seems like much of the dispute when it comes to the NYer article is that it falls in the category of essay, not journalism–where one dances the line between fiction and fact. There are some (occasionally even me) who believe that all essay is in truth fiction, and this piece, with the narrative intertwined with family memoir, it seems probable.

    For instance, if I were to write an essay about the barroom stories I heard while sailing through the Bahamas, even though my memories were drunken and re-remembered, I'd want to distinguish between what someone said and the larger narrative. I would still use quote marks.

    Maybe it's a question of intent? Did Diamond intend to destroy the character of his interviewee? The most damning evidence is if he lied about the year and location of his interviews.

  21. a George said,

    August 4, 2012 @ 5:00 pm

    – I see now that it is really a matter of taste whether Keith M Ellis' word order is better than my proposal, which came later anyway. However, the real problem is that of academic fraud, and whether you consider journalistic treatment to be close to academic writing. The Diamond case appears to be a fraud, completely independent on the punctuation. A court transcript appears to be a guaranteed record of the utterances made. Those who are able to write a coherent text out of something said by persons who are not either telling a narrative (i.e. rehearsed through repetition) or educated in composing their thoughts before speaking, are mostly deviating from the literalness of the reproduction.

  22. Rhonda Roland Shearer said,

    August 4, 2012 @ 7:02 pm

    @Melissa, you wrote: "The most damning evidence is if he lied about the year and location of his interviews." There is no "if." The NYer story states that he met Wemp during 2001-2002 and these quotes were given 'on the road" during chauffeured car rides. Diamond wrote:

    “I first met Daniel half a dozen years after these events…As the driver assigned to me, Daniel picked me up an hour before dawn each day, drove me out along narrow dirt roads, waited while I jumped out every mile or so to record birdsongs, and drove me back to the oil camp in time for lunch…During our hours together on the road, we enjoyed sharing our life stories…It was in these conversations that he told me the story of his revenge.”

    Yet New Yorker fact checker Chris Jennings said "short-hand notes" from May 29, 2006 was the only interview and source of the quotes. The NYer never disputed this fact as they provided it.

  23. Ran Ari-Gur said,

    August 4, 2012 @ 9:37 pm

    If we want separate symbols for exact quotations as for approximate ones, then it might be better to introduce a new symbol for the former. (Maybe Python-style triple quotes, """[…]"""?) Historically, I don't think there was an assumption that quotation marks indicated verbatim text. Consider this sentence from Jane Austen's Northanger Abbey:

    > This was very wonderful, if it were true; and Lady Russell was in a state of very agreeable curiosity and perplexity about Mr. Elliot, already recanting the sentiment she had so lately expressed to Mary, of his being "a man whom she had no wish to see."

    Clearly the text in quotation marks is not meant as a verbatim transcription; rather, the reader is supposed to understand that Lady Russell had said something like "a man whom I have no wish to see." (In fact, that is exactly what she had said — earlier in the book:

    > "He is a man," said Lady Russell, "whom I have no wish to see. […]"

    but the same sort of clearly-not-verbatim quotation occurs elsewhere in the book, even when Austen has not provided exact text for comparison.)

    In a different vein, it's quite common in French for quotation-marks not to be closed and reopened around tags of the form "he said". For example, from Pierre Boulle's La Planète des singes:

    > « Très illustre président, dis-je en mon meilleur langage simien, c'est avec le plus grand plaisir que j'ouvrirai cette boîte ; […] »

    Obviously we are not supposed to understand that the narrator actually said "I say/said in my best ape language" to the ape.

    So, while it's pretty cool that quotation-marks have come to be thought of as indicating an exact quote, it seems that it's more a perception than a reality.

  24. Ran Ari-Gur said,

    August 4, 2012 @ 9:41 pm

    (Sorry, correction: That quote is from Jane Austen's Persuasion, not her Northanger Abbey. But the same sort of thing occurs in Northanger Abbey as well.)

  25. VL said,

    August 5, 2012 @ 12:20 am

    1. I remember reading the Diamond article and thinking there was no way the quotes were accurate, simply because of the style. But my objections to Diamond's work in general have to do with what I perceive as his intellectual glibness, so I honestly didn't give it much thought. It occurred to me, however, that perhaps Diamond was motivated by not wanting to make someone from PNG sound like a 'native' (or, not wanting to make himself sound like a colonialist). Tricky business for a white westerner, I suppose.

    2. I am somewhat perplexed by Ms. Shearer's fixation on the date and circumstances of the interview. Far more damning are all the errors that besmirch the reputation of those involved–if the meat is spoiled, it hardly seems worth complaining that the garnish on top is wilted, too.

    3. I am the most struck, however, by the comments of professional journalists, particularly Teresa Elms. She makes a compelling case. If we grant the journalistic argument, however, then what do we make of the huge to-do over Johann Hari's "editing" of quotes from interviews? He sometimes felt a particular _actual_ quotation was unclear, for all the reasons given by Ms. Elms, and went back to the interviewee's own writings to extract a sentence addressing the same point that he felt did more credit to the interviewee. Yet he was practically tarred and feathered. He is a gifted writer, a thoughtful journalist on many important causes on the left. Is it simply the case that leftie journalists cannot get away with tacitly accepted journalistic behavior, just like leftie politicians cannot get away with sex scandals that a Republican would just sail through? What gives?

  26. Nick Lamb said,

    August 5, 2012 @ 3:28 am

    Keith, quotative like might be what you need here.

    e.g. http://itre.cis.upenn.edu/~myl/languagelog/archives/003978.html

  27. Quotations and “Unquotations” in Journalism and Ethnography | Neuroanthropology said,

    August 5, 2012 @ 7:24 am

    […] More unquotations from the New Yorker […]

  28. tpr said,

    August 5, 2012 @ 8:03 am

    @Rhonda Roland Shearer,
    Strictly speaking, the passage you quote from the Diamond article doesn't say that "these quotes were given 'on the road'". It is perfectly possible that "during [their] hours together on the road… [Wemp] told [Diamond] the story of his revenge" in the context of what Diamond described as "sharing our life stories" while the actual quotes were given in the context of a formal interview that occurred much later. Perhaps there are other parts of the article that strongly suggest that the quotes themselves came from their time on the road, as opposed to that merely being an account of the first time Diamond spoke to Wemp about his story (I haven't read it so I don't know), but as it stands, I don't think you've succeeded in making your point well.

  29. jan said,

    August 5, 2012 @ 9:48 am

    What about the words of people quoted in the Bible…??

  30. Mike Briggs said,

    August 5, 2012 @ 1:42 pm

    This whole thing seems to me to be a storm in a teacup. Why not reserve quotation marks for absolutely verbatim direct speech (oratio recta) – and treat all the paraphrases and euphemisms as reported speech (oratio obliqua). Thus: He said to me "You're a lousy fuckn reporter," or: He told me that my journalism was highly inaccurate.

    Perhaps I'm naive, but I do have the expectation that the language inside quotation marks is something that someone, somewhere, actually uttered.

    [(myl) That's exactly what I feel should be true in journalism and other non-fiction writing, unless the reader is explicitly told that the quotes are fictionalized. (I'm fine with editing out disfluencies as long as they're not relevant to the sense of the passage, and similarly I'm not generally in favor of "eye dialect".)

    Obviously the quotes in fiction are fictional.

    There's a region in between journalism and fiction — story-telling about real events for which there's no record other than the memories of participants — where it often makes sense to use quotes based on the memories involved, which are sure to be fictionalized to some extent. But when the participants involved are identified, I feel that you need to be very careful about this.

    Scrupulous attemption to accurate use of quotation marks is also the official policy of (for example) the New York Times "Guidelines on Integrity":

    Readers should be able to assume that every word between quotation marks is what the speaker or writer said. The Times does not "clean up" quotations. If a subject’s grammar or taste is unsuitable, quotation marks should be removed and the awkward passage paraphrased. Unless the writer has detailed notes or a recording, it is usually wise to paraphrase long comments, since they may turn up worded differently on television or in other publications. "Approximate" quotations can undermine readers’ trust in The Times.

    I haven't been able to find any similar commitment on the part of The New Yorker, but I would be surprised if its editors and publisher would openly and explicitly assert that they operate by a different standard.

    However, as I have explained in this and other posts, it's clear that the routine practice of journalists everywhere — including at excellent publications like The New York Times and The New Yorker — is entirely different from this. And ordinarily, violations of such policies seem to be completely ignored, at least by editors and publishers. That's why I became interested in the factors that made the Jonah Lehrer case different. ]

  31. Rhonda Roland Shearer said,

    August 5, 2012 @ 3:35 pm

    @TPR You have to read the article. The stage is set as Wemp telling a revenge warfare story while driving Diamond to do bird research during 2001-2002. Nothing else.

  32. Rhonda Roland Shearer said,

    August 5, 2012 @ 3:36 pm

    @Mike Briggs , I agree with you.

  33. Rhonda Roland Shearer said,

    August 5, 2012 @ 3:41 pm

    @VL, you wrote: "Far more damning are all the errors that besmirch the reputation of those involved." True. But the discussion–not my "fixation" –is about quotations.

  34. Keith M Ellis said,

    August 5, 2012 @ 3:56 pm

    Mike, I think the answer to your objection/question is that there's a genuine need for paraphrase. That is, we don't always want to repeat something verbatim, but we also don't always want to merely report the superficial literal meaning of something someone said, but rather also include something of the subtext implied by various specific choices made by the speaker/writer.

    Sometimes we know we simply don't recall the exact words, but do recall their flavor. Sometimes a verbatim quote is awkward for stylistic reasons. There's numerous good reasons to prefer something that lies between exact quotation and description.

  35. tpr said,

    August 5, 2012 @ 4:16 pm

    I wonder whether the use of direct quotes really does make a piece more engaging or readable or whatever its virtue is supposed to be. It seems plausible that it does, but it also sounds plausible that it's a myth of the same ilk as the idea that active sentences make better reading than passive ones.

    Mark, can you think of a clever breakfast experiment to test this?

    [(myl) There aren't any big collections out there of pre-baked human responses (whether reading rate or aesthetic preference or eye-tracking or GSR or ERP or whatever) to large volumes of text of different sorts. So I don't see any clever way to address this without running a bunch of subjects. Which is easier now that we can recruit people on the internet; but I'm not quite clear what to test in this case.

    As a matter of common sense, it's pretty clear that story-telling — in text form, anyhow — works better with quoted dialogue than with only indirect-discourse descriptions of who said what when. Here's the opening of Elmore Leonard's Raylan:

    Raylan Givens was holding a federal warrant to serve on a man in the marijuana trade known as Angel Arenas, forty-seven, born in the U.S. but 100 percent of him Hispanic.

    "I met him," Raylan said, "the time I was on court duty in Miami and he was up for selling khat. That Arab plant you chew on and get high."

    "Just medium high," Rachel Brooks said, in the front seat of the SUV, Raylan driving, early morning sun showing behind them. "Khat's just catchin on, grown in California, big in San Diego among real Africans."

    "You buy any, you want to know it was picked that morning," Raylan said. "It gives you a high for the day and that's it."

    "I have some friends," Rachel said, "like to chew it now and then. They never get silly, have fun with it. They just seem to mellow out."

    "Get dreamy," Raylan said.

    Do you think you can rewrite that without any direct quotes, so that it reads nearly as easily and clearly, or characterizes the speakers nearly as well? I don't think that I can.]

  36. The Scienceblogging Weekly (August 4th, 2012) | Stock Market News - Business & Tech News said,

    August 5, 2012 @ 9:39 pm

    […] Lehrer, Bob Dylan, and journalistic unquotations and More unquotations from the New Yorker by Mark […]

  37. Sid Smith said,

    August 6, 2012 @ 1:33 am

    I've been a journalist for 30 years; currently a sub-editor on the (London) Times. I agree with LL's stance.

    I would happily tidy the following quotes:
    'It was really really hard.'
    'It was, you know, really hard.'
    'It was really fucking hard.'
    'It was, I decided, really hard.'
    'It was really – what do you call it – hard.'
    …etc.

    You shouldn't do much more than that and still retain the quote marks. All the above egs are of stuff being removed, not added.

    Transcribing tapes is very time-consuming, so it's common for journos to tape an interview and yet rely on their contemporaneous notes, or even just their memories, to write the piece. This can be very wrong or just a little bit wrong – but it's wrong.

    (Transcription is also inexplicably tiring; bizarrely so. Roll on voice recognition.)

  38. Sid Smith said,

    August 6, 2012 @ 1:39 am

    Correction.

    "Transcribing tapes is very time-consuming, so it's common for journos to tape an interview and yet rely on their contemporaneous notes, or even just their memories, to write the piece"

    I should have said "to write the quotes".

  39. Ann said,

    August 6, 2012 @ 8:21 am

    On The Media ran a story last weekend that rails against the suggestion that journalists ask their interview subjects to confirm their quotes before publication:

    http://www.onthemedia.org/2012/aug/03/when-journalists-let-sources-approve-quotes/

    (The objections here seem to be more about the practice of letting politicians clean up their statements before they are released to the public, but there is no acknowledgement of the fact that those quotes may not be accurate in the first place.)

  40. “And will pardon Paul Claudel, Pardons him for writing well” « Statistical Modeling, Causal Inference, and Social Science said,

    August 6, 2012 @ 9:45 am

    […] of plagiarism and fake quotes, a commenter points to two recent posts by Mark Liberman (here and here) where Liberman links to about a zillion cases of journalists publishing quotes that were never […]

  41. haamu said,

    August 6, 2012 @ 9:59 am

    Re "paraphrase marks": English already has perfectly well-established typographical conventions for indicating paraphrase: the square bracket and the ellipsis. (See the original post and several of the comments for examples.) Why invent something new?

    If the actual utterance was

    "Well, you know, that asshole rocketed out at me and I fuckn levelled his ass with a right chop."

    it could obviously be rendered as

    "[He] rocketed out at me and I … levelled [him] with a right chop."

    The only innovation (if it is an innovation) would be to extend the convention in an obvious way to indicate the entire quote was paraphrased.

    "[He came at me and I hit him with my right hand and he fell to the ground.]"

    I have to believe the intent of this typography would be immediately obvious to the vast majority of readers. I have long done this in my own notetaking, to remind myself that the words I have written down represent an utterance, but that the words are not exact.

  42. Catherine said,

    August 6, 2012 @ 10:57 am

    Agree with Sid Smith above (1:33 & 1:39).

    Also concur that transcribing interview tapes is "inexplicably" and "bizarrely" tiring!

    Have been planning to figure out how to outsource to Mechanical Turk.

  43. Clayton Burns said,

    August 6, 2012 @ 1:45 pm

    I do not recommend recording and transcribing for journalists.

    As the owner of a paper, I would not pay people to do it that way.

    I accept that intelligent and ethical editors would want to keep quotes in order. I do think it appears that some practices at The New Yorker go far too far.

    What surprised me was that:
    1.Editors noted that people do not speak in sentences.
    2.Editors did not draw an obvious conclusion from that.

    Students should be taught to speak in sentences. It is not as if we lack the tools. The official high school grammar in America should be the COBUILD English Grammar.

    The dictionary I like is the Longman Dictionary of Contemporary English. If teachers understood students better, they would realize that it is hard to inculcate good dictionary practice.

    Students are always getting separated from their dictionaries and resorting to Internet shortcuts. I have all my students buy the print LDOCE, and this year I am having all of them buy the mobile application.

    Corpus tools from COBUILD and Longman are among the best practical tools of science. That journalism schools largely ignore them, or fail to grasp how to integrate them, is wrong.

    It often takes two to make a mess. Poor communication skills by the person being interviewed. Weak listening and writing skills by the journalist. There is no reason why we can't have a determined national effort to work on both sides of the discourse.

  44. Yael Grauer said,

    August 6, 2012 @ 11:13 pm

    Making up quotes is, apparently, quite pervasive in women's magazines. Consider http://jezebel.com/5917751/why-did-lucky-print-a-bold+faced-lie-about-one-of-its-readers.

  45. Lane said,

    August 9, 2012 @ 11:08 am

    Does anyone have a good set of rules for what exactly counts as cleaning up disfluencies and avoiding eye-dialect, and what counts as "making up quotes"?

    A friend of mine did a long interview with two African-Americans who spoke engagingly and intelligently in fluent AAVE. He then called me worried: these guys had given him great material for a book proposal, but to write down their quotes as uttered posed him the dilemma of either "making it up" — rendering their quotes into standard English — or be seen to be mocking them by writing their words as uttered. It was a real problem, and I didn't know how to advise him. I've been a working journalist for 12 years, and no one has ever given me a handbook on this issue.

    I told him to do his best with cleanup, but not to rephrase whole grammatical structures. If Bill Withers, missing his girlfriend, says to you "Ain't no sunshine when she's gone," you shouldn't quote him as "There's no sunshine when she's gone." But there's no doubt that the "ain't" route would expose the journalist to criticism of mocking his subject.

  46. Experts-shmexperts. Where’s the data? « House of Stones said,

    August 16, 2012 @ 8:52 am

    […] aren't ones who know how to do anything with it. Which is why I take Mark Liberman's conclusions about quotations in the New Yorker to work pretty well for most analyses about the war in […]

  47. Around the Web Digest | Savage Minds said,

    September 7, 2012 @ 6:20 pm

    […] More unquotations from the New Yorker /KF […]

  48. Unquotations and How to Avoid Them | Book How-To said,

    September 10, 2012 @ 5:27 pm

    […] explores the topic further in follow-up articles here and here. Commenters offer a few possible explanations for inaccuracies—transcribing recorded […]

  49. Links « Nation of Beancounters said,

    February 18, 2013 @ 10:21 am

    […] 4. The quotations you read in newspapers and magazines aren't actually quotations. […]

RSS feed for comments on this post