RSS and Copyright

Internet

by John Palfrey, Berkman Center at Harvard Laaw School

There’s been a flurry of posts and comments on a topic I’ve long been watching, which is the status of copyright and syndication technologies. It’s arisen this time in the context of a project that I’m involved with outside of my Harvard work, called Top10 Sources (seemy disclosures for more; it’s important to note that what you read below is potentially colored by my obvious interests here, though I believe I’ve been 100% consistent on the merits of this argument since I became involved in the discussion).

The issue, raised by a few respected members of the blogosphere,Mike Rundle, Om Malik and Adam Green among others, is whether Top10 Sources is doing something that violates copyright or, separately, is doing something that is outside of “the bounds of accepted aggregator behavior” (perhaps related to the furor over splogs). My view is that the site is doing neither. I believe also that this issue is a very important one to vet fully, as a community, because this debate is going to recur and recur until we sort it out.

What Top10 Sources does is to introduce readers who ordinarily don’t spend all their time reading blogs into the medium. The idea is to offer a directory of reading lists, available as web pages and as OPML files, as well as a quick synopsis of what each of the chosen sites is saying. Top10 Sources is meant to be helpful to the RSS-offering community by directing readers to great content, to get people to subscribe to your feeds, to get people clicking through to blogs. The Top10 Sources editorial group also ends up learning about communities built around ideas. (Soon, Top10 Sources will enable anyone to create their own, competing Top10 lists and upload them to the site, which will add another dimension to the analysis below.)

On the copyright matter: what Top10 Sources does is instructive to whether it’s lawful.
First, an editor, as part of an editorial team chooses a topic, spends a LOT of time in the community of people writing about this topic, consults some technical metrics for the sources, and chooses 10 online sources (defined simply as offering a feed syndicated using some flavor of RSS) that cover that topic. The editors periodically repeat this process, taking one source off the list when a voice fades or stops covering the same topic, and adding a new voice as it emerges as important and topical. The point is to create a human-edited Reading List by topic, and to contribute those sources into a human-created, limited search engine.

As the editor compiles the site, the editor sends out an e-mail to the person who appears to be responsible for the site, or, sometimes, posts a comment to say that the site has been chosen. The site renders a list of those sites offering the feeds as directlinks to the page. The site also subscribes to those feeds and renders them all together on a single page. It is this latter activity that I take to be the concern.

The issue raised here is whether it is a copyright violation to render these syndicated feeds in this way. As a matter of copyright law, I contend that it is not. The strong form of the pro-copyright argument runs like this: the creator of the RSS feed retains, automatically, all copyrights in the content in the feed and retains all rights in its republication, use as a derivative work, and so forth. Given that those rights have been retained fully by the creator of the site, the argument goes, it is unlawful for someone – presumably in a commercial context — to republish that copyrighted context without license to do so. This is the Web 2.0 variant of the argument that is litigated frequently in the context of web-based content, with plaintiffs like the RIAA and the MPAA (in the p2p context), the publishers (like McGraw-Hill, or Perfect 10) who are suing Google, and the like.

Though I don’t believe this to be the end of the story, to be fully responsive to this argument, Top10 Sources offers to remove any feed chosen as a top source immediately. So far, out of the 1,500+ sources chosen to be included in one of the site’s recommended “Reading Lists,” only two sites have asked to be removed, both owned by the same copyright holder. (Out of deference to them, I won’t list them here, though the publisher is well-known for his stance on this topic, which I respect.) So, as an open invitation, for anyone included in one of those lists who wishes to be taken out, just write to terms of service, in the footer of every page, which includes a section on Copyright.

Why this is not the end of the story is that there are several other factors to consider. One is a defense of fair use, which is a four-factor test that excuses some activity that would otherwise be unlawful. Another is the concept of implied license: why, after all, would someone in fact offer an RSS feed if they did not want to be included in aggregators? As an empirical matter, the fact that far fewer than 1% of those that Top10 Sources has included in aggregators in fact have complained about inclusion suggests a norm around what people are expecting when they decide to syndicate their content. As a broader sample size, consider all of the aggregators, whether public or private in the market, which now number in the hundreds, and the fact that we have not yet had a train wreck around the presentation of content in these web pages. Another is the fact that many people have written in, asking to be added to the aggregators.

This is so because, fundamentally, RSS is ads. As Dave Winer haswritten, “RSS itself is an advertising medium, if you use it correctly.” Or, putanother way by Mitch Ratcliffe, “RSS is not content, it’s a channel.” The point of many public aggregators is a place to run these ads, or a TV Guide to these new channels. Some people also embed ads in their feeds, presumably so that these ads will run other places and be seen or clicked through. Another way toput it: “People come back to places that send them away.” (Recall what happened to the AOL walled-garden model.)

If a publisher of RSS feeds thinks of it differently, that publisher has options. First, the publisher can and should put a license in the feed that says what they want people to do or not to do with their feeds. Creative Commons licenses, as I’ve argued on this blog, are the way to go — to embed them into the RSS feeds when they go out, with clear instructions for your intent. If you want people to run your feed in private aggregators, but not in public aggregators that are for-profit, to re-offer your content just as you’ve offered it, and to attibute authorship to you, why not add to your feed aBY-NC-SA license? Second, the publisher of the source, as some have done, can make clear on their blogs or by writing to those who aggregate or allow others to aggregate their content not to do so, pursuant, for instance, to the DMCA 512 procedures. If an aggregator does not abide your wishes, then the publisher can seek to assert a copyright complaint via the courts or otherwise. But to switch the presumption, somehow, back to a strong form of the copyright argument would do far more harm than good.

I’ve been worried about this issuesince early in 2003. Is history repeating itself? Is the blogosphere arguing itself right into a trainwreck of the sort that has played out over music and movies? Consider the world that A (prominent) VC envisions,here andhere, wherein content is micro-chunked and syndicated. This world cannot emerge if every plausible copyright claim is asserted and litigated. Is it a “permission culture,” as Lawrence Lessig has argued, that we want to head for, where every use of syndicated content must be pre-approved?

OK, so maybe you don’t like the micro-chunked and syndicated version of the future. Even without that version of the future, the rights in syndicated content should be clarified. There’s no doubt that common practice is to share the content that you are syndicating for a wide variety of uses. That’s the default that has emerged. Simple, clear, online licenses should demark those feeds that are not meant to be consumed broadly in such a fashion, before the train-wreck hits.

Back to Top10 Sources, I expect to take up this issue again with the management team. I don’t think there’s anything being done wrong from the perspective of the law. But we should take up for discussion some of the ethical issues that Mike Rundle and Om Malik raise and suggestions that Adam Green makes about how much of a given feed that the site republishes — maybe a truncated version of the feeds is the right thing to render. The point is not to “steal” someone’s content, but rather to direct readers to that person’s content after giving a snippet of it. Perhaps the right answer is to limit how much of a feed’s offering is republished in the aggregator.

The broader issue of RSS and copyright remains. The community is speaking, to large extent, by creating a norm around syndication and aggregation which is very important. It would be a great shame if the terrific changes being wrought by online publication, syndication, and aggregation were to be brought down by an aggressive (and in my view, wrong) reading of the world’s copyright laws. As my friend and colleague Jonathan Zittrain might say, the Internet and its communities have a terrific way of “self-healing.” This topic is a great one for the Internet community to solve on its own before it becomes a (self-)destructive fight.