Corrections in the digital editions of Free

As some of you may have seen, VQR rightly spotted that I failed to cite Wikipedia in some passages in Free. This is entirely my own screwup, and will be corrected in the ebook and digital forms before publication (and in the notes, which will be posted online at the same time the hardcover is released), but I did want to explain a bit more how it happened and what we’re doing about it.

First, as readers of my writings know, I’m a supporter of using Wikipedia as a source (not the only one, of course, and checking the original source material whenever possible). I disagree with those who say it should never be used. But the question is how to use it.

In my drafts, I had intended to blockquote Wikipedia passages, footnoting their URL. But my publisher, like many others, was uncomfortable with the changing nature of Wikipedia, and wanted me to timestamp each URL (something like this: http://en.wikipedia.org/wiki/Chris_Anderson page viewed on July 8th, 2008), which struck me as clumsy and archaic. So at the 11th hour we decided to kill the notes and footnotes entirely and I integrated the attributions into the copy.

In doing so, I went through the document and redid all the attributions, in three groups:

• Long passages of direct quotes (indent, with source)
• Intellectual debts, phrases and other credit due (author credited inline, as with Michael Pollan)
• In the case of source material without an individual author to credit (as in the case of Wikipedia), do a write-through.

Obviously in my rush at the end I missed a few of that last category, which is bad. As you’ll note, these are mostly on the margins of the book’s focus, mostly on historical asides, but that’s no excuse. I should have had a better process to make sure the write-through covered all the text that was not directly sourced.

Also note the VQR is not saying that all the highlighted text is plagiarism; much of is actually properly cited and quoted excerpts of old NY Times articles and other historical sources. And as you’ll see, in most cases I did do a writethrough of the non-quoted Wikipedia text, although clearly I didn’t go nearly far enough and too much of the original Wikipedia authors’ language remained (in a few cases I missed it entirely, such as that short Catholic church usury example, which was a total oversight). This was sloppy and inexcusable, but the part I feel worst about is that in our failure to find a good way to cite Wikipedia as the source we ended up not crediting it at all. That is, among other things, an injustice to the authors of the Wikipedia entry who had done such fine research in the first place, and I’d like to extend a special apology to them.

So now we’ve fixed the digital editions before publication, and we’ll publish those notes after all, online as they should have been to begin with. [UPDATE: A draft version is here. The final version will live in the right column of this blog permanently] That way the links are live and we don’t have to wrestle with how to freeze them in time, which is what threw me in the first place.

Here’s the statement that my publisher, Hyperion, released yesterday:

We are completely satisfied with Chris Anderson’s response. It was an unfortunate mistake, and we are working with the author to correct these errors both in the electronic edition before it posts, and in all future editions of the book.

Comments

[Glenn Fleishman]

This whole episode was fascinating, because it's so out of keeping with the usual run around with book authors about anything to do with anything.

VQN found something weird, did the thorough research one would expect if you were going to confront somebody with plagiarism, and then contacted you to obtain a response before publishing (if I get the time sequence right).

Your response was to take immediate responsibility for the omissions, explain why, and take steps along with your publisher to right the situation.

I can't recall the last time that happened.

I'm not familiar with the term write-through. Does that mean taking a source and reworking it to avoid citation? I can understand doing that with Wikipedia which is a secondary source.

[Dennis G. Jerz]

If "write-through" means "paraphrase to avoid having to cite," that's still a problem, since you're still building on the Wikipedia article without citing it. (Paraphrase can still be plagiarism of ideas, as the discussion on VQR clarifies.)

If "write-through" means "go directly to the sources cited by the Wikipedia article and cite them directly," that's a lot of extra work, and will typically result in longer passages (with multiple citations).

Having said all that, I recognize that you're doing the right thing now.

[Mark Frauenfelder]

I'm surprised that the VQN is coming down so hard on you about it. It's obvious you didn't try to pull a fast one. You just made a mistake of carelessness, which is human and forgivable.

I think many non-fiction writers share the same nagging fear that their source notes will accidentally get mixed into the manuscript without proper attribution. Because it's so easy to copy and paste, this kind of thing is going to happen to other writers. I'm working on a book now and I really hope I haven't screwed up! I wonder if there's some kind of company I can hire to check my manuscript before the book gets published?

Your candor and proactiveness in this matter is commendable. You are doing the right thing.

[Dennis G. Jerz]

@Mark, "some kind of company I can hire to check my manuscript"

Turnitin.com?

[David Gerard]

Right ... so why couldn't you just cite the specific revision ID, as clicking the "Cite this page" link at the left of every page on Wikipedia allows you to? We make this as easy as we possibly can - what bit wasn't easy, so we can make it easier?

[David Gerard]

@Mark - the VQN comments read like they're trying to get their revenge on everyone who ever used the web or Wikipedia.

[Barry Ritholtz]

I wonder if this is typical for Publishers.

My disagreement with my original (now former) publisher about a similar sourcing issue -- I wanted to add numerous pages beyond end/footnotes called "Additional Resources."

I know I am a sponge and tend to absorb what I read, and I wanted to make sure if anything was subconsciously "copied and pasted," I was covered.

They balked at the extra pages . . .

[Chris Anderson]

David,

That's exactly what you'll see in the online note, but I thought the "Retrieved 19:39, June 24, 2009" was clumsy and confusing when enshrined in paper that would be coming out a year later. So I decided to kill the notes entirely, and then I screwed up the resulting integration process since we didn't have a backup wikipedia citation method. In retrospect, I would have liked to just say "According to the Wikipedia entry on [this subject]...." and then direct readers to the online notes, where the full URL would live. I hope that would have been a reasonable compromise. Next time. Sigh....

[mike]

lifting large spans of copy from wikipedia of all places while retaining inaccuracies... you should just know better than that.

[Galen]

David--
I wondered about citing specific revisions and I see that sure enough Wikipedia actually provides a helpful tool for that. The problem, I suspect, is that the resulting URL is not very human-friendly, whereas the link to the main page is. So for instance
http://en.wikipedia.org/wiki/The_long_tail
is the link to the live version, but
http://en.wikipedia.org/w/index.php?title=The_Long_Tail&oldid=298319774
is the permalink to the current revision. I don't want to have to read that, or have it appear in my document, it's just too ugly.

Is there any way to rejigger the underlying site architecture or create human-friendly forwarding urls that redirect to the php-ified urls? Something like:
http://en.wikipedia.org/wiki/The_long_tail/2009-06-24-1005UTC
?

Something like that would be human-friendly, not too hideous to print, and would clearly indicate the retrieval date within the URL itself.

[Tim]

Wikipedia gives you the ability to link to old versions of the article directly. The old versions won't change and visitors can see what you linked to and then click to see the lastest version of the article.

For instance: http://en.wikipedia.org/w/index.php?title=The_Long_Tail&oldid=272075630

Why not just use that URL?

[Debbie Weil]

Hi Chris, I think your "fix" is just fine. This is a fascinating example of "open publishing." I.e. telling us about Hyperion's wishes, the awkwardness (and datedness) of time-stamping the attributions and your proposal for how to handle it differently. Writing about fluid, Web-based information in a print book is clumsy. Sigh, as you put it.

P.S. I put in a request for a free copy of FREE; fingers crossed.

[Jackson Landers]

I am continually seeing readers attack Chris for using Wikipedia at all as a source. And I was going to post a comment here stating that the problem is really just the issue of plagiarism and that it is a fact that the Wikipedia has been proven to be no less accurate overall than the Encyclopedia Britannica.

But first I wanted to check that to make sure it was true. So I found this:

http://news.cnet.com/2100-1038_3-5997332.html

Which is the study I recall hearing about at the time. But wait, that study was done by Nature. Huh. Chris used to write for Nature. So I dug a little deeper and found this:

http://www.theregister.co.uk/2006/03/23/britannica_wikipedia_nature_study/

Essentially, in March of 2006 the people from Britannica took great exception to the accusations of inaccuracy and in turn accused Nature of 'cooking the books' in the study in order to make what has been characterized as their pet website look better.

At around the same time (I can't see exactly when the piece was written), Chris Anderson wrote an op-ed for Nature advocating that scientific journals conduct peer review through, essentially, comments by their readers a la Wikipedia.

http://www.nature.com/nature/peerreview/debate/nature04992.html

Somewhere in here, there is a snake devouring its own tail.

It would take more time than I have at the moment in between tasks at work to tease it all out, but its here. I can feel it. I cannot help but feel that the perfect man to explore this would be Errol Morris, which I'm sure will be understood by anyone who has read his excellent pieces in the NY Times over the past year. Appropriately enough, or maybe it is a conflict of interest, Morris usually gets to the solution to his seemingly-simple-but-actually-sprawling mysteries through the feedback of his comments sections after each installment has been written.

[Ricky]

Mark Frauenfelder: Nice fawning comment from a guy who worked for many years at Wired magazine. Oops, you forget to mention that.

Then again, you did appear in ads for Apple Computer, another behavior which many people felt was questionable.

[Robert Creutz]

Chris - Classy response. Check out www.ithenticate.com when you get a chance. I'd be happy to allow you a sample account for review. Perhaps you'll even consider a feature article in Wired. In our business, we understand that whatever people feel about the legitimacy of wikipedia, it is a widely used research resource. As such, we have a dedicated crawl of this content to ensure that our content verification software can draw attention to wikipedia matches so authors can mitigate circumstances like yours. On the other hand, your book is going to get great publicity despite. I recall grabbing a copy of Kosinski's "The Painted Bird" from my local library as soon as I learned of controversy regarding the true authorship of the book. Again, kudos for a classy response.

Bob Creutz, GM iThenticate

[Aheli Shooes]

A time stamp for the quote would have been the most logical.

Also very, very, very few really care if you cite wikipedia or any other random website from the Internet.

[Stephen]

One problem with systematically removing all references to Wikipedia is simply that you are removing a valuable piece of information about your process.

I agree that their cite links are "ugly", but that's what footnotes and endnote are for.

[Mark Frauenfelder]

Hi Ricky: You are right. I should have mentioned that I was an editor at Wired from 1993-1998. I left several years after Chris took over, but disclosing my past employment with Wired adds context to my comments. Thanks!

[Air Jordans]

I don't want to have to read that, or have it appear in my document, it's just too ugly .

[spinchange]

Interested in learning more about the online edition, anyhow. Look forward to more details.

[Ricky]

Mark, you seem particularly sympathetic to Chris. Are you supportive of Chris because of some of the criticism you received for doing the commercial? Do you feel journalists are held to too high of a standard?

Finally, do you think journalists should endorse products?

[Mark Frauenfelder]

Hi Ricky -- I like your line of thinking here, but you are stretching too much to try to find a link between my Apple commercial and my defense of Chris. It doesn't really make sense. I have plenty of other character flaws that would be more appropriate for you to use here. Let's take this off line and I can help you develop a stronger argument. Email me at mark@boingboing.net.

[Ricky]

I'm sorry you're not willing to answer my questions in this public forum.

I'll email you.

[pyrate]

If the future of everything is free, why do you expect me to pay for your book? Especially if big chunks of it are lifted from Wikipedo?

[Cville guy]

Jackson Landers, it might be nice if you mentioned you were Waldo Jaquith's twin brother. (they have different last names, long story)

Waldo is the editor at the VQR who discovered this to begin with.