My Guardian column on censorship versus copyright protection
Viacom and others want hosting companies and online service providers to preemptively evaluate all the material that their users put online, holding it to ensure that it doesn't infringe copyright before they release it.LinkThis notion is impractical in the extreme, for at least two reasons. First, an exhaustive list of copyrighted works would be unimaginably huge, as every single creative work is copyrighted from the instant that it is created and "fixed in a tangible medium".
Second, even if such a list did exist, it would be trivial to defeat, simply by introducing small changes to the infringing copies, as spammers do with the text of their messages in order to evade spam filters.
In fact, the spam wars have some important lessons to teach us here. Like copyrighted works, spams are infinitely varied and more are being created every second. Any company that could identify spam messages -- including permutations and variations on existing spams -- could write its own ticket to untold billions.
the latest
latest episodes
Wouldn't the online filters only check against copyrights that had been submitted, and possibly verified? Who in their right mind would think it would be an exhaustive list of every copyright ever?
Filters would not look for 'exact' copies they would look at 'fingerprints'. Do you really think the FBI compares .bmp's of peoples fingerprints pixel by pixel to what's in their database stopping if its not an exact match?
I'm pretty sure you know better, so please, stop building straw men, or at least don't put them so close to a slippery slope, idiots believe this crap and are so loud no one can focus on the real issues.
Spoon, that's exactly what Viacom is asking for -- for Google to a priori determine wwhat all the copyrighted works it owns are, and to be able to identify them even in fragmentary form.
Google offered to block any videos that Viacom would enumerate (though I don't think they would have been able to do so), and Viacom turned them down.
Indeed, this is also what the Science Fiction Writers of America have asked for of sites like Scribd -- that they determine, a priori, whether any copyrighted text, no matter how fragmentary, be blocked before it is posted. Included in the works that SFWA served in its last takedown were introductions to short story collections, which they subsequently said that Scribd should have identified as copyrighted works and blocked.
If "fingerprinting" worked as well as you say, we'd be fighting spam a lot better than we are today. "Fingerprinting" isn't a technical term. Are you talking about SHA-1 hashing (that *won't* work, because a single bit-switch will make the hashes nonfunctional). Latent semantic analysis? Support vector analysis? Bayesian math? I'm pretty steeped in this art -- I once founded a company based on this technology, and I'm on the Metabrainz Board, where we make audio fingerprinting technology -- and I don't think there's an adequate "fingerprinting" technology that can overcome adversarial attempts to obscure the identity of arbitrary files.
So, no, this isn't a straw man. Copyright holders are asking for exhaustive interdiction, and fingerprinting as you've described it doesn't exist.
I think he means fingerprinting with actual FINGERS.
[sockpuppet deleted]
I agree that it's impractical to check online content against all copyrighted material, and requiring prior clearance is a draconian measure. The alternative, which is widely criticized on BB, is to "spot enforce" with a heavy hand. i.e. when a copyright violation is found, prosecute the hell out of the violator, so that the other 99% of unprosecuted violators are so afraid of the consequences of detection that they desist. As far as I can tell, the BB party line is that none of these options is acceptable, ergo copyrights are obsolete and should be discarded. This seems like saying that you can't possibly pull over every car that's breaking the speed limit, therefore there should be no speed limits, and everyone will just magically cooperate and drive at optimal speeds.
No, Phasor, I criticize these efforts when they attack fans who are doing things that, had they taken place in the real world, would have been called "culture," not "infringement."
Also, I criticize these efforts when they target non-infringing material, material not rightfully represented by the complainant (as when the Science Fiction Writers of America sent takedown notices on behalf of Damon Knight, Bruce Sterling, and me, none of whom have authorized them to serve as our representatives), and so on.
Finally, I criticize fines and penalties that are disproportionate to the offense -- the $150,000/infringement fines provided for by the No Electronic Theft act, for example.
Of course, the objection you raise isn't germane to this particular post: we're talking about rightsholders (like Viacom and the Science Fiction Writers of America) who have proposed that hosting services should programatically or manually determine which material infringes and take it down automatically.
I'm currently doing research into the origin and development of copyright for two of my law and graduate classes. Read L. Ray Patterson's Copyright in Historical Perspective. If you've got access--damn the monopolists!--check out his article, The DMCA: A Modern Version of the Licensing Act of 1662 in The Journal of Intellectual Property Law (10 J. Intell. Prop. L. 33).
The trends Cory discusses in his article (which I'm going to cite in my paper), in my view, are an unavoidable return to the purposes for which copyright was originally conceived: an anti-competition trade regulation. For two hundred years, copyright existed to keep the English printing cartel in business, protected from both "unlicensed" (read "non-member") presses and foreign imports. The Stationers' Company invented copyright as a publisher's right, not an author's right, to protect themselves. Parliament eventually got sick of the Stationers' monopoly--they were preventing the printing of works of Shakespeare and Milton--and tried to shift copyright to authors, but it never really worked. Now, after another two hundred years of jurisprudence which tried to conceive of copyright as something designed to foster creativity, we're finding out that that isn't really what it was for in the first place. It's always been about monopoly. It's only a matter of time before this point once again becomes too obvious for the copyright cartels to deny.
Wouldn't it be a great day if corporate entities were varred from copyright ownership? I would love to see what would happen if only originating artists were eligible for copyrights!
@ Mod: Hi, I know we've been over this before and the last thing I want to do is take up more of your time. I've tried to go through posts from the sockpuppet to find any offending comment, but I can't find any smoking gun. I just think it's weird the censorship that's going on. Yes, even though you explained it, I still have to trust you, and I would like to, but I don't have any evidence that the sockpuppetmaster was behaving badly
Cory:
I criticize these efforts when they attack fans who are doing things that, had they taken place in the real world, would have been called "culture," not "infringement."
Well, to take one example, sampling of old records by hip-hop artists has been going on since well before the Internet was in use by the general public, and after an initial period of non-enforcement, copyright owners cracked down and you can't use samples without authorization now. And people have been getting sued for stealing copyrighted melodies or plagiarizing written material since well before TCP/IP existed. On the other hand, it's always been an accepted part of the culture that you can be influenced by another artist -- you just can't directly incorporate easily recognizable chunks of their creations into your own work. Redefining this as "culture" is like saying that shoplifting shouldn't be prosecuted because it's so commonplace, stores still make lots of money, and it gets people into the store and makes it more likely that they'll shop there.
There's actually still another couple of problems lurking behind the ones you mention. Let's assume that we can adequately automatically "fingerprint" works to recognize similarity and have a massive database of everything that we're worried about. Let's further assume that there is some form of global cryptographic authentication which people can use to prove their identities and that the database is sophisticated enough to keep track of who is allowed to do what with which works. There's still three more things that we can't easily do:
1) Differentiate what is fair use from what is not.
There's not a legal bright line in the law. Fair use is murky and decided by the courts. The inspectors you mention in the editorial would not only have to be knowledgeable about language, they would have to be lawyers, and even then it's iffy that they could identify what is and isn't fair use in a consistent and reliable manner.
2) Authenticate ownership of works.
Let's say that we have a huge database where all copyrighted works can be registered. Someone adds a work to the database of copyrighted works. How do we know that it's really their work? It could be someone else's work which they just uploaded. It could be a public domain work. It could be a work which isn't eligible for copyright. Only a thorough investigation of each submitted work could yield this information.
The most troubling implication of this are its potential for censorship. If Michael Moore were working on a new film about the evils of the sprocket industry and an industry spy got his hands on a rough cut, he could upload it to the database and claim it as his own. As a result, when Moore finishes his film and tries to distribute it via the internet, the database would recognize it as being a work which was already submitted (or at least as infringing one). Michael Moore would be effectively enjoined from distributing his film via the internet until he successfully appeals the decision.
3) Identify ownership of portions of a copyrighted work.
Owning the copyright on a film, for example, does not imply owning the copyright on all clips used in that film, on pictures which appear, or on songs which appear in the film. Likewise, owning the copyright on a rap song does not imply owning the copyright of any samples it uses.
Consider the following example which illustrates this case:
A talented comedian videotapes a short sketch and submits it to a sketch comedy program. The sketch comedy program likes it so much that they want to air it as is. So, they sign a contract which allows them rights to broadcast it, include it in DVDs, etc., but isn't exclusive. The creator is still allowed to broadcast it or distribute it himself. The sketch comedy show then broadcasts the sketch as part of an episode.
The network registers the comedy program as being their copyrighted work (as it indeed is).
Someone attempts to upload the sketch to YouTube. Should YouTube allow it?
Our magical automated engine would recognize it as being a part of a copyrighted work, and presumedly turn it down. But the correct response depends on what the creator has done in the mean time. For one, he could be the one trying to upload it, in which case it should definitely be allowed. He also could have release it under a CC license or elsewise given the uploader the right to upload it.
In truth, the example above could perhaps be solved through holding more complicated data in the database, specifying what parts of a show are copyrighted by different people. In that example, you could have a time code breakdown of the copyright ownership of different things. But things aren't always going to be that straightforward. For instance, in a scene in a video documentary, there could be a picture beside a collage accompanied by background music with narration over it. Keeping track of what belongs to who is not just difficult, its nigh impossible.
Keith
I have to laugh at Cory's comparison to the Soviet Union. In this article he says, "It would be a great Sovietisation of the world's digital printing presses". In an earlier article, he says, "DRM is Lysenkoism for the digital age." Uh huh. This from the guy who thinks intellectual property should be free - which to me sounds an awful lot like "from each according to his abilities" (Karl Marx). It's pretty ironic that Cory raises the specter of the Soviet Union when he advocates something so similar. Does he really think we'll all run like scared little children because he invokes the ghost of the USSR to scare us? Further, Cory hopes to hoodwink everyone into thinking that copyright enforcement = Soviet Russia. (Just don't upset his fantasy by telling him that the first US copyrights were established in 1790.)
"Viacom sent over 100,000 takedown notices to YouTube last February, but seconds after it was all removed, new users uploaded it again." No, it wasn't. This is an obvious exaggeration.
I'd be a lot more impressed with Cory's arguments if it wasn't so much exaggeration, fear-mongering, non-sequiturs, and strawman arguments. But, he'll always be popular because he's saying what millions of people want to hear: they should have free access to all the music, software, and digital media they want for free. It's like telling children that they can have whatever they want whenever they want it from Toys 'R Us - for free.
Phasor3000:
"prosecute the hell out of the violator"
I disagree with this method. It's a bit like pulling over a few drivers for speeding and prosecuting them for attempted murder. Yes, you could argue that speeding = reckless endangerment of people's lives, but putting people in prison for speeding isn't the way to solve this. Part of the problem is that it's costly to prosecute people, and that makes it uneconomical to prosecute them for small crimes.
@Cory
I've never researched the idea of 'fingerprinting' media but I know there are a few companies working on watermarks (yes they could be removed) and other ways of determining how close mediasample1 is to mediasample2. You're very right that there is no 100% way of doing this since copyright law doesn't lay out a crystal clear line of what's infringing and what is fair use, and even if it did there would most likely be no efficient way to compare samples quickly and accurately. Also: I never said that these systems worked well, just that big Media wishes big Internet would take this seriously. And if you could block as much copyright infringement as you could block spam, you would probably make much more money then the spam blockers.
Brit, it's pretty weird to argue that an endorsement of a regulatory monopoly (copyright) is a free market position, while arguing that a competitive environment is communist.
Spoon, What you're describing -- a comparison of mediasample1 and mediasample2 -- doesn't just produce poor results, it produces terrible results. Meaningless, filled with both false positives and false negatives. The reason that online service providers don't take these proposals seriously is that they lack technical merit and are undeserving of serious consideration.
Furthermore, blocking infringement is HARDER than blocking spam. Spam is generally text-only and thus more tractable when it comes to comparison and analysis (and also has fewer permutations that are still recognizable as the original). Video and audio can be permuted through multiple resolutions and encodings -- and this is before we get to adversarial techniques, such as stegonography and deliberate scrambling -- that programatically discovering comparable items is in the realm of a fairy-tale.
The companies that purport to do this -- and there are many of them -- have poured millions into the problem without any measurable success.
I have researched this in terms of how it affects the practice of visual art. My findings are not yet published, but photomontage as practiced by the dadaists (and other related groups) is illegal now due to the copyright regime Cory describes.
Appropriation of imagery is an essential part of dadaist photomontage. It is all about taking an image with a well-known meaning and juxtaposing it with other images to create a new meaning.
Unfortunately, few people without a knowledge of art history can see the significance of this. Those who do, know that montage was central to many art forms of the twentieth century and it had a defining influence on film and television.
Cory is right. The copyright regime is not protecting artists, it is freezing us out.
Good point, think of all those Max Ernst collages that are chock full of copyright violations and would result in lawsuits if they were done in 2007. I actually remember wondering about that the first time I saw one of his works. Of course, there's a vast amount of public domain material that can be used by artists for collage, so it's not like it's impossible to create great art just because copyrights are enforced.
Yes phasor30000 we can do photomontage, but not in the dadaist/situationist sense. We are not permitted to visually interact with cutural icons (human or designed) because of copyright law.
In other words, the artist's right to engage in dialogue with our surrounding culture no longer exists. That means we can't visually critique anything in the mass culture we are steeped in. We cannot use the visual language our culture speaks.
For example, I would dearly love to use Nelson (of the Simpsons) in a work I have in mind. But I have to figure out another way to express that "har-har", and no matter what I do it won't have the clear impact that pasting Nelson in there would.
What if I want to use the face of, say, George Bush? Do I knock on the White House door and say I want to take a photo? Do I choose between getting arrested doing that, or getting arrested for using an official photograph? Or pay the price? (I live across the other side of the world, which makes option 1 very expensive too.)
We can take a few photos and layer them together, but that is not the same as creating a meaningful work that challenges some aspect of society.
Of course, there is a fair bit of civil disobedience going on, in the dadaist tradition.
JB Lavin (9): You know why you can't find objectionable material warranting censorship? Because that's not why the comments got removed. I don't see how you can have missed this, because I've explained it repeatedly, but yon sockpuppet is a person who's been suspended, and so is not currently allowed to post.
Since he's thick as two short planks, he's had trouble grasping that you aren't currently allowed to post comments also means you aren't allowed to post comments under a variety of transparent pseudonyms, either.
The stuff about how he's tried to resolve the issue has zero basis in fact.
Have I finally managed to make this clear to you, or are there parts of it I need to explain again?