Yeah, I'm not sure if this is true, but I have the impression the error rates have increased in the last decades, despite spell-check becoming more effective at finding them at low cost.
There are some publishers I won't even buy books from anymore, because I've paid too much money for books that are so rife with typos and nonsense that it was clear that it never even touched the inbox of an editor or a proofreader.
I recently read all of Philip K Dick's novels. Spotted typos in every book. I log typos in my goodreads reviews
In the past it wasn't even the author's fault necessarily. The worst case of this was an anthology of Ursula Leguin, "Five Complete Novels" where I quit recording the typos as there would be multiple per page
If the typos are in a legitimate eBook copy, it's because sometimes the people who make the eBook don't have the source file(s), and instead have to OCR a printed book.
They are working hard to lose the newspaper of record status.
They omitted some critical information last week in the Kavanaugh story and blamed it on editorial process.
Reporters Robin Pogrebin and Kate Kelly said in an interview on MSNBC that they wrote in the draft of their Sunday Review piece that a woman who Kavanaugh was said to have exposed himself to while a student at Yale had told others she had no recollection of the alleged incident.
Their editors, they say, removed the reference. “It was just sort of. . . in the haste of the editing process,” said Pogrebin.
Couple more of these and they will be in Buzzfeed level of trustworthiness.
You're going to get downvoted for sharing wrongthink and not screaming "hate!" at Goldstein... I mean Kavanaugh, but you're ultimately right.
The thing is, though, there is no more "newspaper of record". The newspaper industry has been in such a downward spiral for such a long time that, like any other industry bleeding money, quality has severely suffered. Try to find a single notable paper that adheres to the objectivity guidelines of the AP guidebook. Hell, even the AP doesn't do that anymore.
> Couple more of these and they will be in Buzzfeed level of trustworthiness.
This is unnecessarily harsh. The New York Times is a massive company with hundreds of journalists and editors.
If every once in a while a journalistic organization makes a mistake, I think it's reasonable to point it out, have them publicly correct the record, and then let it go.
Harping on every honest (if sloppy) mistake as if it is evidence that an organization is totally incompetent or dishonest is something autocrats do to de-legitimize journalism itself. So long as 99% of the time, they don't make mistakes, and so long as they promptly acknowledge and correct the mistakes when they are called out, I think journalism done in good faith (like the work done at The Times) deserves our benefit of the doubt in an era when it is under unprecedented attack by leaders worldwide.
As a former NYTimes employee, I can assure you that 99% of the work is not done in good faith. It's all activism now, top to bottom. There's a few good eggs granted, but the newspaper has an agenda and uses their pseudo-objectivity/newspaper of record as cover and concealment.
Newspapers have always had an agenda. Op-eds particularly have always been a cesspool of political nonsense.
The difference now is the agenda is obvious because more viewpoints are available online, and in the past the agenda would - sometimes - be supported with hard journalism from real sources, including field reporters.
Now there's a lot more newswire and PR copy pasta and Google searching.
I'm fine with this, so long as they don't insist on their objectivity. Otherwise everything they say must be filtered through an undefined lens that readers either a) don't have time for and are thus victimized by the 'untruth' or b) believe and desire the confirmation bias and are thus victimized by misinformation.
However, this mistake was a piece of information that even the Washington Post said was the reason they did not report on it [1]. To confound it, this was a piece seemingly to drum up interest for the books that the writers of the story also wrote.
Those facts don't make this seem to be an honest mistake. Those facts makes it look to be two authors promoting a book and hoping a bombshell article will drum up sales for their books. It's awfully convenient that the editors took out a couple of details that make it a non-story.
But journalists have not been making mistakes "every once in a while". They have been making mistakes constantly. Nor do they "promptly acknowledge and correct the mistakes when they are called out"; they almost never correct them, and often they double down instead. So by your own criteria, journalism no longer deserves the benefit of the doubt.
At least Buzzfeed never tried to convince the American public that Iraq has weapons of mass destruction and therefor should be invaded and occupied. NY Times is already below Buzzfeed.
The TL;DR is always that the US ginned up the evidence on WMD as a pretense to go back to war.
My take on all of that was the Saddam Hussein was running a con. He wanted his regional adversaries to have the thought that he had WMD but wanted the UN inspectors to never find a "smoking gun". The analogy would be someone making a gun gesture in their coat pocket and then finding themselves confronting someone who really does have a gun...
On top of that he clearly didn't adhere to the restrictions imposed on him after the first Gulf War leading to the second Gulf War to eliminate any possibility of him having WMD and to enforce the restrictions.
Other explanations for our motivation to go to war never made sense to me such as "it was for the oil". Much cheaper (in blood and $$) to just purchase the oil if that was our goal, but the oil economy is basically private so it doesn't even make sense as a government goal. But I digress...
I'd always assumed publishers _were_ doing this already. If not, whence the whining and chagrin over low rent "bloggers" somehow not being comparable to deal authors and journalists?
I realize this isn't as widely held a view as it was 10-20 years ago, but bloggers used to be looked at with disdain, untrustworthy because of the lack of reputability that came with published materials.
This makes me further question whether there was much gatekeeping at all even before this fake news, post truth era.
>This makes me further question whether there was much gatekeeping at all even before this fake news, post truth era.
It was much harder to comment on things and have an international audience for your outrage. Basically to criticize a piece of work was reserved to a few professional critics writing in newspapers. Now everyone who spots a mistake can write about it on Twitter, Amazon review or plenty of other channels.
For a long time, there were multiple newspapers in each city. They would love to catch each other in mistakes. If one paper wouldn't publish criticism, the others would.
These weren't just yellow-journalism tabloids, but two or even three genuine competing reporting organizations. They fought to be the first to publish, but couldn't afford to simply publish falsehoods, or fail to admit mistakes.
Not that they were paragons of honor all the time. They're humans. But it did set up a pattern of keeping each other honest, which they maintained to a large degree after they could no longer sustain multiple papers. That has diminished substantially as they're competing against Internet-style "move fast, break things, nobody cares what you wrote yesterday" publishing (and before that, 24 hour TV that wants to present a constant crisis so that you must never turn it off).
So there was less "suppressing public criticism" than you seem to be implying. It happened, to be sure, but it was very different from the current self-swallowing echo chambers you get today -- at least among the major newspapers. Modern journalism, even the comparatively mainstream stuff, has a lot more in common with those yellow journalism tabloids.
That's a good point I didn't even think of, however it may have been subject to censorship and editing by the newspaper staff and would have limited reach for most newspapers, not a global audience.
Also I don't think I ever wrote a response to any newspaper in pre internet times, that's a lot of work just to be heard.
I did a couple of 'letter to the editor' that got published in those dark and dusty times. I don't recall any editing and there was no censoring, but they did get to determine which letters were published.
It was fun, and I certainly _felt_ like I was being heard - first when I saw my words in print, and again when selected replies would get published a week later or so.
Ultimately, I suppose not so different than posting to $platform_of_choice and waiting for replies/retweets/etc. Just slower.
Fact checking sounds great in theory but maybe in practice maybe it is an exercise in confirmation bias.
For example, in a case cited by the article an author did employ at least one fact checker[0] who was an ostensible expert and who describes a reasonable effort but who still got it wrong (maybe).
In another example, the article alludes to "mistakes" which was actually plagiarism.[1]
If the question really is a matter of trusting books or not, then maybe a reasonable heuristic is to identify the agenda of the author and weight more heavily the writings of authors with divergent priors, similar to Slate Star Codex' idea of adversarial collaboration.[2] An example is Kahneman and Klein's Conditions for intuitive expertise: a failure to disagree.[3]
Nope, it's generally a case of looking up things asserted as fact or otherwise independently verifying them. Quotations may also be verified with the interviewee. If I had known that such a job existed when I was in college, my career path would have been very different.
With the 24 hour news cycle, I've started regarding large popular news-o-tainment companies in the same vein. CNN/FOX are just as bad as assuming some egg on twitter is reporting unaltered truths.
This was inevitable and is the culmination of two things:
- Too much hypothesis-driven science that results in statistically significant, but untrustworthy results. It's easy to find statistically significant data to backup whatever crazy thing your brain comes up with.
- Too much pseudo-scientifically backed popular press. Popular science is useful for disemminating science that has been solidified through decades of falsiability tests, not last month's papers.
Some years ago, when I was working as production editor in New York, I had to manage a manuscript for a book of new-age woo that was trying to pass itself off as pseudo-science (one of these 'quantum magic' type books). Not only was the premise crap, but the author clearly did not understand even basic historical facts about science. I brought this up with the editor, who seemed surprised to be told that the author did not know what she was talking about. I'd guess the editor had not really read the manuscript very closely. I worked with the copyeditor to try to straighten some of it out, but it was pretty much futile.
The NY publishing system is set up so that editors spend most of their day doing tasks unrelated to actually editing their books, of which they have too many. It gets worse as you go up the food chain.
That's a huge problem, but not the one this article is about. The examples given are things like books about 19th century court cases or insider stories about the trump admin, not science.
Awkward surely, but in this day and age I can at least give her credit for listening and not trying to bluster past it like so many public figures when confronted with inconvenient facts.
Add to this Richard Dawkins new book, "Outgrowing God" which is rife with inaccuracies, baseless conjecture about the motivations behind some of the ancients civilizations he talks about (an assyriologist posted a pretty pointed takedown of some of the inaccuracies on twitter https://twitter.com/GHeathWhyte/status/1175081067943997440), and theology so bad you would question whether or not he himself has ever actually opened up a Bible.
As the twitter thread notes, he seems to have gotten some of his historical information about ancient Mesopotamia from a website called historywiz (https://www.historywiz.com/flood.htm), which I suppose is good if you are writing a 8th or 9th grade paper on the topic, but not a book.
However your previous remark about Richard Dawkins opening up a Bible is a bit worrying.
It is obvious that the missing historical information is not in the Bible, and that remark can be interpreted a bit like the point of view of Jehovah witnesses, that all you need to read is the Bible.
I think their statement about Dawkins’ apparently having never opened a Bible was in regard to his theological understanding, not historical.
Having read Dawkins, my primary critique of him overall is that he regularly demonstrates a willful ignorance as to what religious people actually believe. He’s content to attack straw men and not the propositional truth claims at the heart of religious belief.
I'm intrigued by the claim of "dropped footnotes" (and its passive construction). How exactly were these footnotes dropped? There seems to be an implied blame of the word processing software and I find that rather dubious.
I have a problem with "news" sources that don't allow commentary on their articles. It seems that many eliminated the comment sections when the articles were too often found to be incorrect and/or pushed a blatant narrative, rather than just reporting the facts.
Also when they'd get filled with comments from the "other side" of the political spectrum.
Re: just reporting the facts. Look in a journal or in data tables for facts. News is propaganda to sell newspapers. Everything else is secondary to that for them.
That's my impression too, specifically with the Guardian. If I'm not mistaken, you used to be able to comment under most articles, then they started to disable comments under articles on "inflammatory" subjects (read: about immigrants etc.) and now you can hardly ever comment. Which makes sense, because o lot of their articles would be just torn to pieces if commenting was allowed.
I wish there was some mention of major newspapers in this. It seems like a glaring omission. Through all the "fake news" stories in the last years, I have yet to see someone like NYT openly talk about e.g. the gell-mann amnesia effect that the rest of us so openly agree is a thing.
> But a factual controversy can also be costly, driving down sales and, in extreme cases, forcing the publisher to recall and destroy finished copies.
Are there examples of this happening? Michael Wolff's book was a bestseller, for example, despite many of the factual disputes coming up prior to its release.
In theory, lack of fact-checking would seriously hurt a traditional media outlet, as it could reduce traffic or subscriptions. For authors, however, it's not exactly the same calculus. Once the book is sold, it's sold (I realize it's a bit more complex than this with advances and returns, my point is that convincing someone to keep paying them monthly isn't a concern). Once they become a best-seller, they'll be able to put "best-seller" in their "About the Author" section. While there's a future credibility argument (which can be partially mitigated by the value of being a "best-selling author"), I'd argue that from a pure "maximize returns from this publication" perspective, there's really only an incentive to fact check things that are so egregious that it could lead to a recall or stop the book from being published in the first place.
This is not a good situation to be in, but the job a publisher is hired to do is to take a manuscript and turn it into something that can be sold in bookstores. I'm not really sure that it should be their role to police this, unless it's published by a media outlet explicitly as journalism.
Even in science there is too little due diligence.
I'm nearly done a PhD in mechanical engineering. Recently I was writing the review section of my dissertation and I decided to check one of the most used results in my field, the most common empirical equation for the boundary of the "atomization" regime in the breakup of liquid jets into droplets. This dates back to the late 1970s. Turns out that it's off by a factor of about 4 due to a math error. Out of the hundreds of people to examine this result, I think I'm the only one to notice the error. I spent a fair amount of time checking the literature to make sure I'm not crazy but seems no one noticed the problem.
A result being highly cited is no guarantee that it's correct.
Is that "off by a factor of 4" result still "useful"? Can you still create products/devices/things/science with that incorrect result? Even epicycles still allowed reduced-accuracy prediction of the movement of heavenly bodies
The result was a factor of 4 too high (edit: 3.3, to be more precise), which is conservative for certain applications. But for those applications I'd argue that you are typically confident that your device is in the atomization regime a priori and don't need to check.
But it's extremely inaccurate for my application, where you don't want your device to be in the atomization regime. In fact, I only started examining this in more detail because my own experiments clearly contradicted the textbooks, review articles, etc. If someone blindly applied this result then they'd potentially make a bad device.
56 comments
[ 3.4 ms ] story [ 118 ms ] threadIn the past it wasn't even the author's fault necessarily. The worst case of this was an anthology of Ursula Leguin, "Five Complete Novels" where I quit recording the typos as there would be multiple per page
If the typos are in the printed book, that's bad.
At the very least, end it on a high note that you can be proud of -- don't let it degrade slowly until people are openly mocking it.
They omitted some critical information last week in the Kavanaugh story and blamed it on editorial process.
Reporters Robin Pogrebin and Kate Kelly said in an interview on MSNBC that they wrote in the draft of their Sunday Review piece that a woman who Kavanaugh was said to have exposed himself to while a student at Yale had told others she had no recollection of the alleged incident. Their editors, they say, removed the reference. “It was just sort of. . . in the haste of the editing process,” said Pogrebin.
Couple more of these and they will be in Buzzfeed level of trustworthiness.
The thing is, though, there is no more "newspaper of record". The newspaper industry has been in such a downward spiral for such a long time that, like any other industry bleeding money, quality has severely suffered. Try to find a single notable paper that adheres to the objectivity guidelines of the AP guidebook. Hell, even the AP doesn't do that anymore.
This is unnecessarily harsh. The New York Times is a massive company with hundreds of journalists and editors.
If every once in a while a journalistic organization makes a mistake, I think it's reasonable to point it out, have them publicly correct the record, and then let it go.
Harping on every honest (if sloppy) mistake as if it is evidence that an organization is totally incompetent or dishonest is something autocrats do to de-legitimize journalism itself. So long as 99% of the time, they don't make mistakes, and so long as they promptly acknowledge and correct the mistakes when they are called out, I think journalism done in good faith (like the work done at The Times) deserves our benefit of the doubt in an era when it is under unprecedented attack by leaders worldwide.
The difference now is the agenda is obvious because more viewpoints are available online, and in the past the agenda would - sometimes - be supported with hard journalism from real sources, including field reporters.
Now there's a lot more newswire and PR copy pasta and Google searching.
I'm fine with this, so long as they don't insist on their objectivity. Otherwise everything they say must be filtered through an undefined lens that readers either a) don't have time for and are thus victimized by the 'untruth' or b) believe and desire the confirmation bias and are thus victimized by misinformation.
It seems bias is becoming encouraged, quite disappointing.
However, this mistake was a piece of information that even the Washington Post said was the reason they did not report on it [1]. To confound it, this was a piece seemingly to drum up interest for the books that the writers of the story also wrote.
Those facts don't make this seem to be an honest mistake. Those facts makes it look to be two authors promoting a book and hoping a bombshell article will drum up sales for their books. It's awfully convenient that the editors took out a couple of details that make it a non-story.
[1] https://www.washingtonpost.com/lifestyle/style/a-book-on-kav...
My take on all of that was the Saddam Hussein was running a con. He wanted his regional adversaries to have the thought that he had WMD but wanted the UN inspectors to never find a "smoking gun". The analogy would be someone making a gun gesture in their coat pocket and then finding themselves confronting someone who really does have a gun...
On top of that he clearly didn't adhere to the restrictions imposed on him after the first Gulf War leading to the second Gulf War to eliminate any possibility of him having WMD and to enforce the restrictions.
Other explanations for our motivation to go to war never made sense to me such as "it was for the oil". Much cheaper (in blood and $$) to just purchase the oil if that was our goal, but the oil economy is basically private so it doesn't even make sense as a government goal. But I digress...
I realize this isn't as widely held a view as it was 10-20 years ago, but bloggers used to be looked at with disdain, untrustworthy because of the lack of reputability that came with published materials.
This makes me further question whether there was much gatekeeping at all even before this fake news, post truth era.
It was much harder to comment on things and have an international audience for your outrage. Basically to criticize a piece of work was reserved to a few professional critics writing in newspapers. Now everyone who spots a mistake can write about it on Twitter, Amazon review or plenty of other channels.
It was common to write replies in long form and mail then into newspapers, which would then publish them.
Only if the newspapers themselves allowed it. Newspapers could suppress public criticism that came their way through this channel if they so chose.
These weren't just yellow-journalism tabloids, but two or even three genuine competing reporting organizations. They fought to be the first to publish, but couldn't afford to simply publish falsehoods, or fail to admit mistakes.
Not that they were paragons of honor all the time. They're humans. But it did set up a pattern of keeping each other honest, which they maintained to a large degree after they could no longer sustain multiple papers. That has diminished substantially as they're competing against Internet-style "move fast, break things, nobody cares what you wrote yesterday" publishing (and before that, 24 hour TV that wants to present a constant crisis so that you must never turn it off).
So there was less "suppressing public criticism" than you seem to be implying. It happened, to be sure, but it was very different from the current self-swallowing echo chambers you get today -- at least among the major newspapers. Modern journalism, even the comparatively mainstream stuff, has a lot more in common with those yellow journalism tabloids.
Also I don't think I ever wrote a response to any newspaper in pre internet times, that's a lot of work just to be heard.
It was fun, and I certainly _felt_ like I was being heard - first when I saw my words in print, and again when selected replies would get published a week later or so.
Ultimately, I suppose not so different than posting to $platform_of_choice and waiting for replies/retweets/etc. Just slower.
For example, in a case cited by the article an author did employ at least one fact checker[0] who was an ostensible expert and who describes a reasonable effort but who still got it wrong (maybe).
In another example, the article alludes to "mistakes" which was actually plagiarism.[1]
If the question really is a matter of trusting books or not, then maybe a reasonable heuristic is to identify the agenda of the author and weight more heavily the writings of authors with divergent priors, similar to Slate Star Codex' idea of adversarial collaboration.[2] An example is Kahneman and Klein's Conditions for intuitive expertise: a failure to disagree.[3]
0. https://www.independent.co.uk/voices/naomi-wolf-outrages-gay...
1. https://twitter.com/mcmoynihan/status/1093290016632115202
2. https://slatestarcodex.com/tag/adversarial-collaboration/
3. https://www.ncbi.nlm.nih.gov/pubmed/19739881
After the debacle with Lehrer, I'd be amazed if the New Yorker wasn't fact checking everything the print, down to their ad copy.
- Too much hypothesis-driven science that results in statistically significant, but untrustworthy results. It's easy to find statistically significant data to backup whatever crazy thing your brain comes up with.
- Too much pseudo-scientifically backed popular press. Popular science is useful for disemminating science that has been solidified through decades of falsiability tests, not last month's papers.
Wow, never seen this worded so clearly before. Thanks for the great summation buboard!
The NY publishing system is set up so that editors spend most of their day doing tasks unrelated to actually editing their books, of which they have too many. It gets worse as you go up the food chain.
https://mobile.twitter.com/thymetikon/status/113170257787850...
However your previous remark about Richard Dawkins opening up a Bible is a bit worrying.
It is obvious that the missing historical information is not in the Bible, and that remark can be interpreted a bit like the point of view of Jehovah witnesses, that all you need to read is the Bible.
Having read Dawkins, my primary critique of him overall is that he regularly demonstrates a willful ignorance as to what religious people actually believe. He’s content to attack straw men and not the propositional truth claims at the heart of religious belief.
Re: just reporting the facts. Look in a journal or in data tables for facts. News is propaganda to sell newspapers. Everything else is secondary to that for them.
Are there examples of this happening? Michael Wolff's book was a bestseller, for example, despite many of the factual disputes coming up prior to its release.
In theory, lack of fact-checking would seriously hurt a traditional media outlet, as it could reduce traffic or subscriptions. For authors, however, it's not exactly the same calculus. Once the book is sold, it's sold (I realize it's a bit more complex than this with advances and returns, my point is that convincing someone to keep paying them monthly isn't a concern). Once they become a best-seller, they'll be able to put "best-seller" in their "About the Author" section. While there's a future credibility argument (which can be partially mitigated by the value of being a "best-selling author"), I'd argue that from a pure "maximize returns from this publication" perspective, there's really only an incentive to fact check things that are so egregious that it could lead to a recall or stop the book from being published in the first place.
This is not a good situation to be in, but the job a publisher is hired to do is to take a manuscript and turn it into something that can be sold in bookstores. I'm not really sure that it should be their role to police this, unless it's published by a media outlet explicitly as journalism.
I'm nearly done a PhD in mechanical engineering. Recently I was writing the review section of my dissertation and I decided to check one of the most used results in my field, the most common empirical equation for the boundary of the "atomization" regime in the breakup of liquid jets into droplets. This dates back to the late 1970s. Turns out that it's off by a factor of about 4 due to a math error. Out of the hundreds of people to examine this result, I think I'm the only one to notice the error. I spent a fair amount of time checking the literature to make sure I'm not crazy but seems no one noticed the problem.
A result being highly cited is no guarantee that it's correct.
The result was a factor of 4 too high (edit: 3.3, to be more precise), which is conservative for certain applications. But for those applications I'd argue that you are typically confident that your device is in the atomization regime a priori and don't need to check.
But it's extremely inaccurate for my application, where you don't want your device to be in the atomization regime. In fact, I only started examining this in more detail because my own experiments clearly contradicted the textbooks, review articles, etc. If someone blindly applied this result then they'd potentially make a bad device.