Sniffing out edits
We鈥檝e been looking recently at a site called . Its is "to monitor corporate news organisations to uncover bias" and it does this by tracking changes to stories on this website and . It also looks for the "censoring" of comments to our .
On news stories, it automatically detects and shows where they have been amended or updated, then visually highlights the relevant lines or passages.
Having looked at various stories treated in this way, what it mostly reveals is the minute-to-minute editorial processes of 24-hour online news, where stories are written, published, then updated and added to for as long as details continue to emerge. It also shows some of the workings of the writing and sub-editing process in which stories are subbed for length as new quotes are added in, paragraphs are rephrased to accommodate new material, and pictures, links and background are added.
It also, of course, shows up corrections. Our policy is to correct anything that鈥檚 wrong - spelling mistake, factual error or anything else - as soon as we become aware of it. News Sniffer highlights even the smallest of these changes in a way we don鈥檛. Should we do something similar?
When we make a major change or revision to a story we republish it with a new timestamp, indicating it鈥檚 a new version of the story. If there鈥檚 been a change to a key point in the story we will often point this out in the later version (saying something like "earlier reports had said...").
But lesser changes - including minor factual errors, corrected spellings and reworded paragraphs - go through with no new timestamp because in substance the story has not actually progressed any further. This has led to accusations we are "" - a sinister-sounding term that implies we are actively trying to hide what we are doing. We鈥檙e not. It鈥檚 just that continually updating the timestamp risks making it meaningless, and pages of notes about when and where minor revisions are made do not make for a riveting read - as News Sniffer, I would argue, tends to prove.
We are concentrating on providing the fullest, most accurate and most timely account we can and there鈥檚 a risk that adding a lot of detail about the process will get in the way of telling the story - affecting clarity for the reader and the speed of the journalists.
But if sites like this can help show more of the journalistic process and make it more transparent that is no bad thing.
I haven鈥檛 said anything about because, at the moment, we think their tracking is not working properly and is highlighting comments as 鈥渃ensored鈥 which are, in fact, published and live on the Have your Say pages. We are in touch with the architect of News Sniffer to see whether and how this can be fixed.
UPDATE, Friday lunchtime: I've responded to some of the comments raised below here and here.
Comments
Please can the 大象传媒 start editing its presenters so they stop using the horrible expression "bored of" doing something and use the correct form, "bored with". Thank you.
There's one other reason for NOT changing the time stamp with every minor update. I tend to use the 大象传媒's RSS feeds. If you were to change the date stamp every time you corrected a spelling, that would falsely mark it as a new article in RSS newsreaders.
I counted the number of comments shown on the Flag-burning 'Have your say' thing (22 pages, with 15 comments on each page equals 330 comments), and compared this with the number that the 'Have Your Say' website believes you have (378). I wonder if a programming glitch is causing some comments to be not shown?
鈥 "stealth editing" - a sinister-sounding term that implies we are actively trying to hide what we are doing. We鈥檙e not. It鈥檚 just that continually updating the timestamp risks making it meaningless鈥
What a strange statement. The time stamp is there so people can see when changes are made. Not updating after a change, however small, is by definition a stealth edit and renders the time stamp 鈥榤eaningless鈥. These stealth edits are not always minor corrections/changes. B大象传媒, LGF and USS Neverdock have numerous examples where significant changes have been made but where the time stamp remained unchanged. It鈥檚 why the term was invented.
You can maintain a change log separate from the main page of the article, so anyone who's interested can see how the story has evolved.
See the Wikipedia page on 大象传媒 for an example. The history tab shows the update history, the discussion tab contains comments.
This is a common technique, and I'm sure other contributors can point to better examples.
An edit history, like Wikipedia has, might be a good idea to shut up the conspiracy minded dissenters. I'm not sure what people are expecting you to do, though. Create a new, almost identical, article every time you correct a spelling mistake?
Rather than doing a similar service to News Sniffer yourselves why not ensure that such data is easily available to any organisation that wants to monitor or research. In doing so you could set an example or even better establish a standard to be used else where.
Democracy would be underpinned and journalism strengthened.
(Perhaps you already do some of this through your news feeds 鈥 in which case just keep letting us all know about it.)
Hi, I'm the architect of News Sniffer. Regarding the 'Have Your Say' monitoring: it is now fixed.
It was highlighted to me by a 大象传媒 techie on Thursday and I fixed it over the weekend. (as mentioned in the News Sniffer blog).
Some fun examples of "riveting" news article changes can be found here:
My favourite one concerns Cuba:
My guess is there are many more interesting changes to be found in amongst all the spelling changes and breaking stories. You just need to look carefully. Recommend any you find interesting.
I note the registrant of the News Sniffer site - John Leach - doesn't feel the need to display all of his address details on the Nominet site, yet complains about the 大象传媒 not displaying timestamps for minor editorial changes?
A case of do what I say, not do what I do?
News Sniffer is clever - but reading it is irritating and the site does not actually do what it claims.
It's irritating as it refers to every change, edit or addition as a censorship, which is obviously incorrect. More importantly, it does not uncover bias as it claims, it just logs changes.
Bias is uncovered by understanding the syntax of the document and analyzing the nouns, verbs and adjectives used. If the site did that, I would be very impressed. Mind you, I can do it myself a lot faster, which is why I read 大象传媒 news not Fox.
As it is, News Sniffer strikes me as just another paranoid anti-establishment site.
Oh Good Grief!!!
This is one site that will NOT be going on my favourites. There is enough of value out there on the net, and too little time to keep abreast of it all, without obsessing over such piffling details.
On the subject of editing though. Could you please start bleeping out people (Like Alan Milliband) who use the phrasing "the India's and China's of this world". It's not so much that I am "bored with it", it's just so fatuous.
I think the existence of News Sniffer is an excellent development - now you need to maintain your own wiki-style edit history pages to ensure that they in their turn are kept honest.
It might not be a wholly bad thing if a rumour gets about that 'Have Your Say' is censored by the 大象传媒. It will keep the half-dozen or so people who post most of the messages there occupied, and distract them from polluting these blogs, where the real debate goes on.
Slightly more rigorous checking of content before going "live" would greatly cut down on the number of errors and post-publishing corrections. Though it is never possible to eliminate all errors pre-live, and even ardent proofreaders will argue and disagree over the most minor of points, it should be possible to get the number of contentious or incorrect wordings down.
For example, there is confusion between "gaming" and "gambling". This has happened before on the 大象传媒 news website. Currently there is a prominent feature / story about online gambling that is incorrectly titled with "gaming". There's a big difference.
You *do* stealth edit - I've corrected NewsOnline on many occasions, and you have never reflected the changes in your time stamp.
Oh, and why do you put "censored" in quotes? You block comments that you feel breach your policies - you censor them. What's ambiguous about it? Just because you feel you censor in a good cause doesn't stop it from being censorship.
In fact... you didn't publish my last comment to these blogs...
I hope you'll take what News Sniffer has to say about censorship on your Have Your Say pages seriously. I use the HYS pages frequently, and I have noticed that the software you use for it seems particularly bug-ridden. If their figures don't agree with your own, please don't assume that they are wrong.
Censorship on HYS certainly does happen more often than it should: too many of your HYS pages are fully moderated, and even on the reactively moderated pages I've often seen my comments disappear even though I can't see any way in which they could have been thought to breach the house rules. Whether that's censorship or bugs in your software, however, I wouldn't like to say.
You could have two timestamps, the current one, and one changed for any minor edits, or you could simply relabel the existing timestamp as 'Last significantly updated at:'.
What you're doing at the moment, well intentioned or not, is plainly dishonest, since you're claiming that pages have not been updated when they have.
I just clicked on "Have your Say" (https://news.bbc.co.uk/1/hi/talking_point/default.stm).
How many more clicks until I get to Les Roberts answering reader questions on the Lancet Iraq mortality report?
You opened the "Iraq death estimates: Your Questions" on "Monday, 23 October, 2006, 16:04 GMT 17:04 UK". His answers have a timestamp of "Monday, 30 October 2006, 14:09 GMT". But there seems to be no link from the "Have your Say" page itself.
Finally, when you say, about the Congo, "Some four million people are thought to have died", where is the reporting on critics of such estimates on the basis of "main street bias" or "small sample"?
I think a web-based medium is perfect for providing a full version history of an article - it's not something that can be done on a TV report or a newpaper, but can be achieved fairly simply on the web. How it might work, I'm not sure. As #2 said, you wouldn't want to update the timestamp because that would affect RSS feeds, but each article could perhaps have a version log available for those that are interested in seeing it. Common versioning systems such as CVS or SVN do this automatically and with very little extra storage required.
I like the idea - not for 'finding bias', but when you update articles over time, the 'original' article gets lost, and several times I've wanted to go back to that first report.
I don't think you need to bother logging spelling changes - that's a waste of your time and a waste of ours. But where the flow, content or facts of an article have been updated, perhaps it wouldn't hurt to log it.
Why are you wasting 大象传媒 resources engaging with a website like Newssniffer (and then writing a blog about it!)? Do you really care about a few nerds desperately clamouring around to see if you've changed a couple of words in a story? Newssniffer is a clever little tool, but ultimately a bit pointless.
I think the distinction between major edits and minor corrections is very clearly explained in Steve Herrmann's piece.
The Standards and Guidelines information for bbc.co.uk is well version tracked - the decision making process is very clear. Maybe more specific content issues could be included in these guidelines, if they are not already available. A simple guideline stating when the time stamp is updated, with examples, is probably sufficient for most people (with a link from 大象传媒 News).
I agree that having version tracking information on the same page as each news story could get quite messy, although I support the idea of having more editorial information. The Blog is already doing a good job of highlighting these issues.
As (5) and (6) say, having an edit history behind each timestamp as a link would put articles beyond dissent, as everyone would be able to see how they had changed.
This is just a test message.
Are you banning certain bloggers from posting comments here?
Not surprising that the 大象传媒 isn't trusted anymore. As far as I know no one at the 大象传媒 has answered questions about the "impartiality summit" internet streaming - was the public told beforehand or not? If so, where and when?
Stick to your guns, 大象传媒
The comments on here are so full of viral PR people and armchair lobbyists it's sometimes hard to read. Just look at the Lebanon conflict. There were probably a dozen guys trying totag the 大象传媒 as anti-semite or something equally rediculous.
Each comment on here is a grain of salt and no more.
As for updating the timestamp with each edit, that's a little something we like to call "not seeing the forest for the trees."
Good job for taking it into account, though.
As for the "bored of" guy, that's actually a colloquialism, and there are plenty of english-speaking countries where it is acceptable to say it that way.
-j.o.
The idea that presentation can get messy if you include too much information is a relic from the non-digital age. That's what customisation is for and every reader can choose what information they want to see.
Paul Reynolds, I am told, explained that some of the answers and questions put to Les Roberts were omitted because of "space". That just does not make sense in the internet. It's trivial to add a link saying "for more of the same go to the next page". So what was the real reason?
Pity there isn't a news sniffer for downright ignorance, or the promoting of ignorance, as exemplified by so many 大象传媒 "science/health" stories. In 大象传媒 science a perverted idea of "balance" takes precedence over factual correctness.
I'm glad this site is provoking discussion, as the 大象传媒 needs to be as open as possible.
I'm working on a 大象传媒 WeatherWatch program myself.
Regarding timestamps, there are 2 opposing concerns: on the one hand, you don't want to give the impression that you have revised the article when you merely corrected a spelling or grammatical error; on the other hand, there is a valid expectation that if one cites a 大象传媒 report as at a certain time, the article will conform to the citation. Perhaps you should accumulate trivial edits offline and only post them to the website when you have enough that you are prepared to change the timestamp - otherwise those who cite your reports are vulnerable to incorrect accusations that they misquoted what you had posted.
Regarding accusations of censorship, you should not be ashamed to say you do it - censorship is a positive thing that each one of us does when we communicate, and every public forum must do to avoid indecent or banal comments.
Peter in Post 26, could you -possibly- get -more- off topic?
I get the feeling that the people complaining about minor edits are just looking for something, anything, to complain about ... that so many seem to have noticed minor edits go through and the timestamp remain, they say, unchanged makes me wonder just why they're reading the site to begin with.
Such people usually have already made their minds up that the site is going to be biased, and will jump on every triviality to "prove" their assertion ... whilst conveniently ignoring what it says about their own biases in the process.
News Sniffer by its stated reasons for existing shows a prejudiced bias as well. Anyone who doesn't admit that there's media bias, depending on the outlet, and that they don't automatically "trust" the outlet whose biases happen to agree with their own, is deluded.
This isn't a case of the pot and the kettle, more the pot and a singularity. Why on earth do people keep -coming- back here if they dislike it so much?
MOST of us don't sit there and scribble down the last edited time of every post we read to see if the wording changes and the timestamp doesn't. This says something poignant about how a large number of people don't really have lives, but I think the gentle readers here will have already gotten the point ...
Kudos to News Sniffer for coming up with yet another totally pointless piece of 'Net real estate - about the only thing I can see it'll actually do is serve to give certain posters around here even more excuse to nitpick.
And I thought Star Trek fans were bad!
p.s. I freely admit to having a bias of my own ... I'm biased against biased people.
Paul Herrmann,
Much of what you say makes sense but you miss (or obfuscate) the main point. There is no doubt that the 大象传媒 stealth-edits the content of its articles. And what about editing headlines?
jack oatman maintains that it's "ridiculous" to consider the 大象传媒 anti-Semitic in connection with its coverage of the Israeli-Hezbollah war. I recall a headline that screamed out "Israel kills Lebanese civilians." The 大象传媒 would never dream of applying the same headline to Islamic Jihad, for example, even though deliberately killing Jewish civilians is the main reason for Islamic Jihad's existence.
The headline was quickly stealth-edited to something more neutral, but of course it had already gone around the world and the damage was done. The 大象传媒, of course, would not tell us that it was difficult, if not impossible, to distinguish civilians from Hezbollah - who blended into the population, using civilians as human shields.
What happened there with that headline? An over-eager journalist in the field trying to prove his/her anti-Israel credentials? An editor doing the same thing? We'll never know because 大象传媒 editors will never admit their bias.
In the end more information is always a good thing. I think you'll just have to put up with the fact that some people will take it and spin it into ever more elaborate evidence for bias and censorship.
Of course occasionally they might be right!
John Leach (comment #8),
I read the story you linked to on Cuba.
The implication of the original text was that many US investors are keen to invest in Cuba. While I'm sure this is true, this is not a conclusion that can be supported by any evidence presented in the story, which only mentions an interview with a single American rancher. The rancher was already in Cuba, presumably with the intention of investing, so the sample was biased towards the conclusion.
It looks pretty clear to me that the text has been changed to remove an unsupported claim, rather than through any editorial bias. If this story represents the best example of bias you can find, then you've actually increased my confidence in the impartiality of the 大象传媒.
The other stories you point to provide at best ambiguous evidence for biased editing. However, I'm sure there are hundreds of edits to 大象传媒 articles every day, some of which will support various and conflicting claims of partiality. In order to believe that the edits you select are representative of a systematic bias, I would have to trust you and the visitors to your site to rate articles without imposing any prejudice of their own (for example, by selectively ignoring updates which contradict their world view).
Since there is no way of guaranteeing that your contributors will be even-handed, there is no way for me to distinguish claims of biased 大象传媒 editing from selection bias by your contributors.
I'm glad that you're monitoring the media, and I'm sure that your site will expose some interesting stories. However in my opinion your site is far more vulnerable to claims of bias than the 大象传媒.
There are too many ad hominem comments appearing here and in HYS. Bashing those whose opinions differ to yours does not further the argument.
Stick to the point, and play the ball, not the man!
Sparticus (comment #28):
You have a very good point, but the revision stands as an example either way you look at it. If you see the modification as moving more towards "impartiality", it still shows a bias existed.
Regarding News Sniffer bias, this is something I've touched on in the blog here:
I think Revisionista really needs to be used by others to uncover bias. It won't do this on it's own. It is a tool to be used over the long term.
Thanks John (#34, assuming the post numbers are working again):
Regarding the Cuba article; I take your point, though there's a danger of mistaking a simple error as evidence of bias, or even bias on the part of the reporter (or editor) with institutional bias at the 大象传媒.
I've read your blog entry regarding potential News Sniffer bias (linked below - you linked to the Cuba story), and agree with your take on this.
I look forward to seeing how the News Sniffer site evolves. I'm not sure how far it will go to expose editorial bias, since presumably this is reflected more in the original content of the articles than in subsequent changes, but I'm ready to be surprised by the results.
IMO if it prompts the 大象传媒 to include accurate and detailed change logs then it's a job well done.
Why are you bothering to take some website seriously? Who cares what some tinpot conspiracy theorist posts on the internet?
We trust the 大象传媒. We know it's more reliable than the entire so-called `blogosphere'. Don't pander to the geeks; just get on with the job!
I don't trust the 大象传媒 - I don't trust any media outlet, because I know they're all biased to some degree. That's simply human nature
The 大象传媒 however is a lot less biased than the majority, and usually comes up with the middle ground of the rest.
It seems a lot more likely to me that any editing that goes on is much more likely to be driven by complaints from strident advocates of the lies and distortions that are "political correctness" than anything else.
I don't know if the 大象传媒 deserves kudos for being so responsive to the comments and feedback of its "viewers", or to be castigated for trying to please everyone in that respect.
News Sniffer will do nothing more in the grand scheme of things than give more ammunition to nitpickers - not that they need it, their fevered imaginations give them more than enough to begin with).
If it really wants to be of use it should start running diffs of political statements and speeches. That would, however, require more than an automated system (i.e. require actually working on things), so it's an unlikely evolution.
The 大象传媒 isn't perfect - but it's a damned sight better than the majority of the competition right now. I wonder why the site wasn't set up to monitor any of the other mainstream outlets by default - why pick the 大象传媒 to launch to? Makes you wonder, doesn't it?
Father jailed for US mutilation ???
If we're on the subject of edits, maybe you'd like to correct the above headline from one of today's stories.
Did the Father mutilate the US? Surely it should read "Father jailed for daughter mutilation".
We decided to have a look at the Cuba example cited above and - for those interested - here鈥檚 what happened:
The original intro sought to tie the piece into the recent news about Castro鈥檚 uncertain health and speculation about the future. Our reporter called the newsdesk to say he felt the piece as a whole was not focusing primarily on this angle so the intro wasn鈥檛 quite right. He also wanted to add an explanation of why the farms were repossessed in the first place, for the sake of context.
We agreed with the points he made and so agreed to make the changes.
A few more responses..
Richard,
We think there is a glitch which is causing some comments not to be shown. Our tech team is working to put a fix in for this. That doesn鈥檛 mean we aren鈥檛 removing some comments which break house rules 鈥 we are, and we've summarised why.
J.G.
It鈥檚 a question of when it鈥檚 significant enough to warrant changing the timestamp. If we changed it with every amendment to the story you would get no sense of when the story had been substantially updated with new information. Most people are interested in the latest and best account available and that鈥檚 what we are striving to provide.
Spartacus/Magic Dave/Man in a Shed
This is certainly a possibility. Variations of this have been suggested by people in our team here also. Perhaps it鈥檚 something to consider. I don't immediately know technically how much work would be involved. I wonder whether it would present problems when covering difficult legal stories. Also even assuming it鈥檚 feasible and we put the work in, I wonder how many of our users would find it valuable and look at it.
John Kirriemuir,
We have a subbing process and we are always looking for ways to improve it. The reality is that we are writing lots of stories a day 鈥 and updating many more, often very quickly - and in the process errors do slip through. We correct anything that鈥檚 wrong as soon as we become aware of it.
Ewan MacMahon,
This is an interesting suggestion which we鈥檒l consider.
Themos Tsikas
The link should have been there 鈥 we have reinstated it.
Steve R/Diane Sims
We don鈥檛 want to waste 大象传媒 resources - I thought it was an interesting phenomenon worth commenting on. I also thought I should explain a bit about the editorial processes it highlights 鈥 and yes this blog does give us an opportunity to do just that.
Themos Tsikas,
We could almost always publish more. There鈥檚 an infinite amount of information and opinion out there. We know people鈥檚 time is limited. As a news site we have to prioritise what we do and what we offer to readers and we aim to keep it clear, focused and concise as a general principle.
Spartacus
I鈥檝e added a comment on this above
Billyquiz
That wouldn鈥檛 have fitted our tightly constrained headline space, and we wanted to say it happened in the US. So I won鈥檛 "stealth edit" it.
To Steve Herrmann,
Sorry about calling you Paul in my no. 30 post. I was interested in your response to numerous posters, especially the last to Billyquiz. If you wanted to indicate that the mutilation took place in the US, why not simply state: "US father jailed for mutilation." There seems to be a real problem in the 大象传媒's headline writing department. Apart from the stealth editing I indicated in my post above, you put words in quotation marks unnecessarily and you seem to use headlines as an opportunity to propagandise.
It would also be good to get one - just one - response from the 大象传媒 to my and others' numerous comments about the anti-Israel propaganda you pumped out during the recent war. Your silence on this matter speaks volumes.
The fact that most of the "Have Your Say" discussions are "Fully Moderated" enables any censorship to take place undetected.
Some time ago there was an open "Have Your Say" about censorship of Google in China. When people pointed out that the 大象传媒 also censors things, those comments were removed at break-neck speed. People responded to this by pointing out the new censorship and a little war ensued, and the 大象传媒 gave up. You can still see that the top-rated comments are all about the 大象传媒 censorship.
When I complained to the 大象传媒 about this I was told that the postings about 大象传媒 Censorship were "off topic". I find that to be quite frightening and I am grateful that someone has taken the time to create the News Sniffer site.
I find the lists of censored posts from the 'Have our say' sections most interesting. Particularly those made during the have your say on poppies in retaliation to a post referring to the soldiers who took part in those conflicts as 'old farts' who nobody cares about. I note the original contentious post was left on the 大象传媒 site. Oh what amusement that must have caused among the 大象传媒 moderators.
If that is the sort of game they play with reactively moderated topics we can only guess at what they are doing with the fully moderated ones which are often more slanted than the decks of the titanic.
While I'm here what criteria did the people carrying out the user survey use to select who they would question. I was fully prepared to take a day off work and pay for travel to Manchester to participate 'have my say' on 'have your say' yet never heard from them after the initial online survey. Is it a case of 'he who pays the piper?'