Take the A Train

Worthwhile speculations sometimes arise with an element of serendipity. For example, on a mailing list that I am on, there was a thread on how to describe complex content packages to facilitate interlinking.

John Mark Ockerbloom, a computer scientist lodged in the University of Pennsylvania Library, noted a conundrum raised by web 2.0 services, and was able to contextualize a difficult issue with great clarity.

In turn, his thinking was informed by the recent article in Slate by Tim Wu (“Tolerated Use: The Copyright Problem“) on how our often draconian copyright laws are often not uniformly enforced — creating a growing Sword of Damocles over the general population.

With his permission and that of the mailing list owner, I have reproduced what John wrote below. John sets the stage by first citing a catalyst provided by another correspondent:

Jon Noring wrote, regarding a mother posting video of her daughter playing “Take the A Train”, with no more metadata than “my daughter plays the piano”.

In the traditional view of content, this video will simply be standing alone online. A few people may find it (in this case probably by random browsing), view it, and that’s pretty much the end of story.

If I may go off on a bit of a tangent, I find this an interesting example of some of the potential pitfalls of the “2.0” world.

In the usual scenario, this is a case where minimal metadata can work as a benefit: the video’s seen by a few people, they enjoy it, and no further effect occurs on markets or people’s lives.

Now, let’s say that someone helpful comes along, realizes that what’s being played is Strayhorn’s “Take the ‘A’ train”, and annotates it accordingly. It’s then found by lots more people, many of whom link to and include it in their virtual music video collections. (Let’s assume for the sake of argument that her daughter’s a particularly talented pianist.) A friend of the family annotates the video with the identity of the pianist.

Then, let’s say along comes an agency who represents Billy Strayhorn’s music publisher, who use the metadata to find the clip, identify the performer, and send them a demand for a large chunk of the family’s life savings for copyright infringement. (The piece was, after all, composed in 1941, and is almost certainly still under copyright in the US; and statutory penalties for infringement can run up to $150,000 under current law, and that’s apart from any showings of actual damages.) What started out as a simple desire to share a daughter’s talents with some friends ends in a case that could ruin the family. (And, to be fair, if the daughter’s performance is popular enough, a free video of it could hurt sales of other, royalty-collecting performances of ‘Take the A Train’, and therefore hurt Strayhorn’s estate.)

Would this actually happen? It’s not very likely, since it would be an expensive and unpopular tack for most publishers. As Tim Wu points out, what’s now happening in the US is widespread low-level violation of copyright laws that’s largely tolerated. But legal action or threats are not unthinkable. You don’t have to look far to find copyright holders acting against their economic self-interest for one reason or another.

Once something is out on any public website, it might as well be on national TV as far as exposure to copyright infringement liability is concerned. And if someone powerful has a reason to go after someone not — powerful (whether to make an example, or to attack someone who is troublesome in other ways) copyright law gives them a potent weapon to use if they choose to.

Laws that have pervasive coverage and severe penalties, but that are only occasionally enforced, can all too easily turn into weapons that the powerful can use to keep down the less-powerful. (“Do something we don’t like, and we can have you ruined in the courts with infringement suits.”) We’ve seen this scenario play out before: the current state of US immigration is a similar legal dynamic in dire need of reform.

In the past, various technological and legal safeguards prevented this from happening. Performances were naturally limited in audience unless special, complicated arrangements were made. Infringement penalties were lower. Various fair use and automatic licensing scenarios were agreed upon for common distribution channels, including the right to cover a song in a club or on a record for a standard, capped fee — I’m not aware of any current analogue for YouTube. And copyrights themselves expired earlier, especially copyrights that no one particularly cared about. (Even renewed, the ‘Take the A Train’ copyright would have expired by now if the maximum copyright length was still what it was when the song was composed.) But changes in both technology and law over the last 15 years have made the gap between public norms and legal codes ever wider and deeper.

If we really are going to move in the “reading 2.0” world, where ordinary citizens and content creators can share their work, and have clear and fair expectations about their risks and compensations, we need serious copyright reform as soon as possible.

John Mark Ockerbloom