The Authors Guild Is Still Wrong About Google's Book Scanning | Hacker News Reader

The Authors Guild Is Still Wrong About Google's Book Scanning

fortune.com

55 pointswalterbell47 comments

I'm bewildered that such enlightened people as Atwood and Ursula K. Le Guin don't grok the value of this resource to writers.

Just an anecdote: the other day I heard a story about an FBI agent tasked with eavesdropping on the telephone calls of Malcom X. Due to the nature of the assignment, his job was to listen very carefully to everything he had to say. The agent wound up coming around to his perspective.

This story came with a quote but no reference to the book, but someone on Twitter thought to search in Google Books and quickly found it for me. This is an amazing power, unimaginable in the days when research meant Dewey Decimals and microfiche.

I have another similar story! This one kind of blows my mind.

In 1978, Corpus Christi College in Oxford, England was being renovated, which involved breaking down a wall that had been standing for centuries. When they broke open this wall, they found that there was a leaf from a manuscript dating to the 16th century. It had been stuffed in there as a sort of building material.

On this leaf was music, which was quickly identified as being a fragment of a piece by Thomas Tallis (c.1505 - 1585) called Gaude Gloriosa. The piece is known because it survives in other sources. It's pretty popular among music from the period and several recordings of it have been made in the last 50 years.

But what was unexpected is that the text on this leaf was not the same text as we have in the other sources. This text was in English instead of Latin. Clearly the composer had reworked his piece to fit a different text. But what was this other text? Nobody knew, and the pages they found in this wall were highly incomplete, so they only had a small fragment of this alternate text.

The mystery wasn't solved until just last year (2015). A British scholar named David Skinner typed the fragmentary text into Google, and out popped an excerpt from a book. It turns out the author of the text is Henry VIII's sixth and last wife, Katherine Parr. 37 years after the discovery of the leaf, and almost 500 years after the actual event, Google allowed scholars to connect the dots on this mystery.

(I'm not 100% sure whether the search that revealed the mystery came from Google books or the regular Google search, since the scholar who did the search didn't specify. But since the complete works of Katherine Parr are on Google Books (https://books.google.com/books/about/Katherine_Parr.html?id=...), it seems likely that it was a Google Books search.)

By the way, this whole story is being published in an upcoming paper called: ‘Deliuer me from my deceytful ennemies': A Tallis Contrafactum in Time of War. The abstract is available here: http://www.tudorpartbooks.ac.uk/media/sites/researchwebsites... (scroll to the bottom)

This reminds me a lot of the entire Digital Humanities revolution with computers. There is a lot of inconsistency with museum records. Especially for art that is collected for cultural value from certain time periods, the authors are misrepresented.

John Resig does a lot of writing/research on the misattribution of woodblock printings and their sources. Through digitizing the works and being able to use simple image recognition tools, museums across the world are able to iron out inconsistencies.

>In the “friends of the court” brief, it says the law was never intended to “permit a wealthy for-profit entity to digitize millions of works and to cut off authors’ licensing of their reproduction, distribution, and public display rights.”

This "cutoff" language is part of what makes the Author's Guild argument fall down. The authors still get all the normal licensing rights as before, so the argument appears wrong on its very face.

What the guild is actually trying to say, I think, is that this new use is something Google should have to pay for. Which is a stronger argument, because it's not trivially refuted by the current status. By overstating their case I think that their point is lost.

> What the guild is actually trying to say, I think, is that this new use is something Google should have to pay for.

That doesn't make sense to me. Books are no different than any other material, are they? If Google indexes this page, should they have to first get permission and pay somebody (us or maybe HN).

Fair use allows for this kind of thing, but is fuzzy: https://en.m.wikipedia.org/wiki/Fair_use

The web is a little different. One factor is that robots.txt is effectively an opt-out. But no, in general indexing a site doesn't require permission.

Right, and I'm saying that indexing a book shouldn't require permission either. Just as you can use robots.txt to opt out of indexing, so can you contact Google and tell them to remove your work from their index.

Whoops, misread your comment. Agreed.

But they aren't just indexing the book. They are copying it completely. Then using that copy to provide a commercial service to people that reinforces the Google brand.

In a similar way, I could _borrow_ books from the library, take them home and digitise them, then return the library book while retaining the digital _copy_. I'm pretty sure this is not allowed under copyright law.

However, if I go one step further, commercialising my collection of digitised books by providing them online to everyone in the world in a searchable format - but removing every few pages - that somehow makes the original copyright infringement legal?...

Does Google Books provide a useful service? I think Yes.

Do they infringe copyright to do it? I think Yes.

How could they not infringe copyright?

I suggest if they bought every book that they digitise, they would then have the right to index those books so that they could offer their fair use service to people. Optionally, publishing houses and authors could opt-in to the system by providing them with a free book or a waiver.

> In a similar way, I could _borrow_ books from the library, take them home and digitise them, then return the library book while retaining the digital _copy_. I'm pretty sure this is not allowed under copyright law.

Its, on first glance, a violation of the copyright owner's exclusive rights to create copies, sure.

Whether its actually infringement or not depends on whether it falls in one of the exceptions to the exclusive rights, such as fair use. Copying a library book and just to keep it for yourself does not seem likely to do that.

> However, if I go one step further, commercialising my collection of digitised books by providing them online to everyone in the world in a searchable format - but removing every few pages - that somehow makes the original copyright infringement legal?...

Quite possibly. Fair use is funny that way, because the boundaries are not crisply defined but anything that falls within it is an exception to the exclusive rights of copyright, transforming what would otherwise be infringement into non-infringement.

> I suggest if they bought every book that they digitise, they would then have the right to index those books so that they could offer their fair use service to people.

If the service is fair use, then whether they own physical copies or not is immaterial, since the users ownership of physical copies of the work is not one of the factors called to be considered in fair use analysis.

You don't acquire a right to fair use of a copyrighted work by purchasing a copy of the work, everyone has an inherent right to fair use (which was a judicially identified limitation of Constitutional character before being enshrined in statute) inherently.

I'm not so much disagreeing with what you have said, as obviously the various courts this has already been through hold a similar position on the matter, however, I would like to emphasise two points:

1) Google continue to possess _copies_ of books that they do not own. Hence my suggesting they purchase the books (i.e. I don't think that partnerships with libraries is enough). To offer the service of Google Books they must retain these copies, i.e. to allow you to read individual pages of the books.

2) Fair use usually entails the production of another copyrighted work that includes as a very small portion the existing copyrighted material. For example, a quote in a book. However, there are some instances such as for criticism or satire that allow a far greater amount of the original material to be used. But my main point is that a new work is created that is now distributable and is complete in and of itself, i.e. it does not rely on the maintenance of any source materials for its continued existence, e.g., a paper, a book, a video, a movie, a poster, a painting, a program, etc.

The copyrighted work that Google have created is entirely their own, i.e. the source code related to the Google Books service. The copyrighted material they are using is the data used by their system. I don't think it is right to argue that each time a person accesses Google Books that the transformation from data in the database to webpage output constitutes a fair use as the content output is largely just a subset of whichever book it is. There is no extra criticism, satire, or transformation added, as you can then read the output as you would a normal book, and the entire purpose is to read the book as the author originally wrote it (albeit with some pages missing).

Actually, while the amount of the source work used is an express.fair use consideration, more.of the source work can way in favor of fair use or against fair use depending on the circumstances, and many things that involve whole works (e.g., time shifting) have expressly been held to be fair use. It is simply false to say that fair use is usually a very small portion.

  > If the service is fair use, then whether they own physical
  > copies or not is immaterial, since the users ownership of 
  > physical copies of the work is not one of the factors 
  > called to be considered in fair use analysis.

I've wondered this also. Can you point to case law for this? I've seen some seemingly authoritative sites[1] suggest that owning a legal copy is a mitigating factor under "Market Effect".

Separately, I've seen some reference to opinions that say that the "Four Factors" are just a starting point, and that other factors (presumably including legal ownership) can also be considered. For example, here[2] Posner says "the four factors are a checklist of things to be considered rather than a formula for decision" and "the factors and purposes are not exhaustive".

[1] https://www.lib.purdue.edu/uco/CopyrightBasics/fair_use.html

[2] http://www.techlawjournal.com/courts2003/ty/20020530.asp

> But they aren't just indexing the book. They are copying it completely.

Is that different than what they do with web pages? Here's their copy of this site:

https://webcache.googleusercontent.com/search?q=cache:DjfN_3...

Publishers can opt out of the book indexing, too (kind of like robots.txt) https://www.google.com/googlebooks/publisher_library.html

Basically, they are saying that if anyone makes a profit off their work, they deserve a cut of it. Makes sense on the surface of it, but let's try applying this to other categories. Take for example a food critic. He makes his living from reviewing the creative output of various chefs. Without the chefs doing what they do, the critic wouldn't have an audience and therefore no newspaper column. So do the chefs deserve a cut of the critic's revenue? Of course a favorable review does tend to drive extra traffic to the restaurant, just as a positive match in Google Books can drive a sale to a given book.

I like your example quite a lot. I feel like if Google where providing a service that only indexed books, but did not provide any content, the authors guild would not have a leg to stand on.

In this altered Google Books service, let's say I search for the following phrase:

    Then they were looking at the familiar environs of the ISS Flight Control Room,
    which was at Johnson Space Center in Houston. The director of mission
    operations was sitting in front of the camera stroking his iPad. He didn't seem
    to be aware that the camera was on.

This "alternate universe Google Books" would present you with only the following information:

    Book: Seveneves
    Author: Neil Stephenson
    Quote is found on page: XXXX
    [ CLICK HERE TO BUY THIS BOOK! ]

The functionality of being a "Book search engine" would remain, without reproducing any copyrighted material. I imagine that this alternate version of Google Books would not incur any ire from the Authors Guild, or if they did take issue, Google's use of their books would be firmly upheld as fair use.

As it stands now, Google Books shows you some of each book you search for, and I suspect that's what the Authors Guild cares about.

The thing is, often times (and I'm assuming it is based on concessions with book publishers) they only show a very small amount of context for the quote -- just enough so you can make a decision if that was really what you were looking for (essentially what falls under common fair use guidelines). And they do have links to purchase the book.

They normally show you multiple pages surrounding the search result. I have often been able to use the result of a Google book search to never need to buy my own real copy.

Same here. I've used it to find the source of sentences for referencing in my academic papers. And read a page either side to make sure the context is appropriate.

> They normally show you multiple pages surrounding the search result.

Yes, when the copyright holder has given permission

Seems like this article is sort of preaching to the choir around here. I'd be interested to hear if anyone on HN agrees with the Author's Guild or can make their argument in an interesting way.

I'm going to make an argument for the Author's Guild even if I disagree with them...

Google is taking books and digitising them. That within its own right isn't illegal. But Google then takes the digitised books and utilises that content for profit (every time you click buy, Google gets a cut of the sale) and what's more the court has withheld the author's ability to decide how they want their books to be advertised/presented.

I could scan in books, present content without context, and misrepresent what a book is trying to say. For example I could create a climate skeptic website that takes pro-climate change books and misquotes them into supporting me. But because of this "transformative change" from physical to digital, suddenly I am allowed to do that.

Every time Google quotes part of a book without context they're changing the meaning of that book. The author has lost their right to stop that from happening. Essentially Google has committed libel against authors by quoting them without context or permission.

Thanks! That's quite good.

> But Google then takes the digitised books and utilises that content for profit (every time you click buy, Google gets a cut of the sale)

Google doesn't use the books scanned for profit. If they sell a book is because they already have it in Google Books, is not one of the scanned

They do definitely profit. They either gain referral fees, or they refer to themselves (Play Store) and thus get effectively free advertising (that would cost them money otherwise).

When you click a link to Amazon for a specific book from a third party website, Amazon pays that website money for that referral. It is no different here except that one Google department sometimes sends it to another Google department.

> Fiction: Google is paid by booksellers like Amazon to include links on Google Books pages.

> Fact: We provide links to booksellers on Google Books pages because we want to make it easier for users to buy books and for publishers to sell them. Booksellers don't pay to have their links included in Google Books, and Google doesn't receive any money if you buy a book from one of these retailers.

http://www.google.es/googlebooks/perspectives/facts.html

Except Google links to their own store in the US (Play Store) which they definitely do profit from directly.

Click the "Buy Book" link here:

https://books.google.com/books/about/Harry_Potter_and_the_So...

Unless Google are claiming they make no cut from Play Store book sales, which is laughable.

And, as I said, they are not profiting of the scanned book because they own the rights to sell the vook in their store.

Unless you're trying to ay that Google can only link to stores that are not their own, which is laughable

> When you click a link to Amazon for a specific book from a third party website, Amazon pays that website money for that referral.

Only if the third-party website has signed up for the Amazon Associates program -- which must be disclosed on the site.

Edit:

Your argument is largely based on the idea of using copyright to prevent mischariacterization of the work, and more broadly for the author to control perception of the contents of the work. My original comment (preserved below) pointed out that such arguments have little legal footing in the first place and wouldn't even work offline.

I presumed you did not know this, but now I wonder if you're arguing to remove existing standards for fair use. But if you're making that case I would have expected to see different arguments. Arguments founded on the utility to society w/r/t progressing the useful arts and sciences, or a new class of utility all together.

end edit

>and what's more the court has withheld the author's ability to decide how they want their books to be advertised/presented.

Controlling the external perception of the contents of the work is not a purpose of copyright. There is even a well established realm of fair use covering criticism and comment. I don't intend to say that reproduction of excerpts for use in advertising (and google's implementation of it) are fair use. They very well may not be. However, a court will not be swayed by legal reasoning based on an author's ability to wield copyright to control public perception of the work.

>I could scan in books, present content without context, and misrepresent what a book is trying to say. For example I could create a climate skeptic website that takes pro-climate change books and misquotes them into supporting me. But because of this "transformative change" from physical to digital, suddenly I am allowed to do that.

No, this is again an established area of fair use that has nothing to do with digitization specifically. Someone could publish that hypothetical website as a book and argue fair use with the standards of "research and scholarship" fair use (and maybe "criticism and comment," depending on how the quotes were used).

>Every time Google quotes part of a book without context they're changing the meaning of that book. The author has lost their right to stop that from happening.

The author NEVER had the right to wield copyright on the basis of preventing mischaracterization of their work, even if is/was de facto ability.

> Google is taking books and digitising them. That within its own right isn't illegal. But Google then takes the digitised books and utilises that content for profit (every time you click buy, Google gets a cut of the sale) and what's more the court has withheld the author's ability to decide how they want their books to be advertised/presented.

How different is this than if I make a website, review their book, and provide an Amazon affiliate link?

So I'm confused about this. Do Google actually own all of the books they've scanned? If I scanned millions of other peoples books and kept the scans, wouldn't I be in breach of copyright?

You don't have to own a copy of the copyrighted materials to have some use fall under Fair Use doctrine. Making a mixtape for your girlfriend falls under Fair Use. Making fun of the mixtape on Youtube also falls under Fair Use. Unfortunately, the big conglomerate "content" industry has pulled the wool over a lot of eyes lately.

Some use sure, but if I have an scan on my harddisk of an entire book that I don't own, are you saying that's potentially fine?

Whether you bought a copy of the book or borrowed it is pretty much irrelevant. Particularly if your computer is on line.

Google scans, stores, and indexes billions of web pages, with obvious and fairly incontestable public benefits. How are books any different? The contents of an article, a blog post, and a NY Times Best Seller all get the exact same copyright protections and considerations.

What you're proposing is no different from saving the contents of a web page to your hard drive.

The web pages are open public access in machine-retrievable format though. Google purposely doesn't index content that supplies a robots file or header requesting it not to be indexed, and certainly doesn't borrow human accounts to allow Googlebot to also index registered-members-only online content.

I'd argue that Google ought to have a similarly strong presumption that work distributed only in a non robot-parseable format and only for a fee isn't intended to be indexed in a publicly accessible database, unless the copyright owner expressly indicates otherwise. Especially if they haven't even paid the copyright holder the regular fee to read the work

So all the copyright holders have to do is put a robots.txt file at the beginning of the book, and all is good? And really, there is no law that I'm aware of requiring Google (or anyone else) to honor robots.txt.

I don't think robots.txt has any legal standing, but I do think the fact that Google is willing to respect requests not to index content from everyone except print publishers of paid content is indicative of bad faith on their part.

Do you have reason to believe that they do not respect requests from print publishers not to scan their books?

https://support.google.com/books/partner/answer/3365282?rd=1

https://www.google.com/googlebooks/publisher_library.html#op...

Regardless of whether they have a legal obligation to do so, they certainly make it sound like they obey such requests.

robots.txt actually has little to nothing to do with copyright, and more to do with "unauthorized computer access" which is protected by the CPAA.

It's a guard against accidental access or publishing of sensitive data - which isn't necessarily copyrighted but has other legal protections - as well as placing an unwanted burden on the server.

> Especially if they haven't even paid the copyright holder the regular fee to read the work

They have though. All the works scanned were either purchased by google in hardcopy form, or scanned in partnership with libraries that had purchased the books.

If I borrowed from a library and kept a complete scan of the book when returning it, copyright law would still apply. Libraries as far as I know do not have the right to authorise third parties to copy entirities of others work.

> What you're proposing is no different from saving the contents of a web page to your hard drive.

Web pages are intended to be copied by the user - it's pretty much impossible to use them otherwise. Books are not.

If you borrow a book from a library or a friend, scan it to a file on your computer and return the book, you've committed a distribution of a copyrighted work without the permission of the copyright holder. If the copyright holder discovered this they could sue you at which point you could attempt to present a Fair Use defense based on the four factors [0]. Assuming it's a whole published book that's available for sale, you'd need a noble use to explain why you didn't just buy a copy to scan for yourself but even if you lose, the damages could be relatively small assuming you didn't distribute the file further.

[0] https://en.wikipedia.org/wiki/Fair_use#U.S._fair_use_factors

Google Books started their scanning with books held in university libraries (and the New York Public Library). While there might have been some variation from school to school the basic deal was Google and the school each got a copy. The university copies are largely, or entirely, held by HathiTrust [1] (who was also sued by the Author's Guild).

[0] https://en.wikipedia.org/wiki/Google_Books_Library_Project

[1] https://en.wikipedia.org/wiki/HathiTrust

Reply on news.ycombinator.com