A copyright reform bill containing the new exception for search engines has passed the Japanese Parliament and will come into force on January 1st, 2010.
I will post an English translation and offer a couple of comments. The English translation will depart from the unreadable style of the Japanese original by relocating all information in brackets into their own sentences.
Article 47-6 Copying for the purpose of searching for uniform resource locators of information made available for download
A business with the purpose of searching uniform resource locators of information made available for download and displaying the search results in answer to requests from the public may, as far as necessary for such search and display, store and adapt information made available for download on storage media, and may serve automatically in answer to requests from the public a copy of such information relating to the uniform resource locators served, including making the information available for download.
An uniform resource locator is a string of characters, numbers and other code used to distinguish information made available for download.
Search engine businesses include those providing only partial services, but are restricted to those that comply with the standard for collecting, organizing and providing information made available for download laid down by government ordinance.
For information made available for download only after requiring input of an access code identifying the user, the exception applies only if the person who restricted the access gives permission.
Update: The above paragraph is incomplete. It must read:
For information made available for download only after requiring input of an access code identifying the user or other means of restricting the access are applied, the exception applies only if the person who restricted the access gives permission. (End of update)
The search engine business may also store secondary works in the course of storing or adapting information.
The search engine may display such secondary works in combination with the information made available for download. This is called “records for displaying search results”.
However, if the search engine business knows that the information made available for download regarding records for displaying search results is in violation of copyright, it may not serve automatically in answer to requests from the public such records for displaying search results, nor may it make the records available for download.
In case of information made available for download in foreign countries, this restriction applies if it would violate copyright in case of serving the information in Japan.
This translation is probably not perfect. Any hints for improvement are welcome. Now for a couple of comments.
I had the chance yesterday for a short discussion of this new law with Masayuki Matsuda, who is a leading copyright lawyer with numerous publications and was involved in the legislative process.
He said that this legislation was not needed, since search engines could operate legally even without it. I disagree with that. In my opinion, search engines should be restricted to opt-in like everyone else in the copyright system.
But if one sees this only as a clarification, there remain some questions. For example under German law, a recent article by former judge of the Federal Court of Justice Joachim v. Ungern-Sternberg (GRUR 2009, 369) argues that authors give implied permissions to search engines the moment they put information on the Internet. Since German law has no search engine exception, without this kind of theory the current operating model of many search engines would become illegal.
Now, again, as an author of web content I object strongly to the idea that I give any permission to a search engine. Especially Google. This kind of theory is, at least in my case, making up a permission out of thin air where there not only is none, but strong objection. If you want to take my content against my will, then write an exception into copyright law, as the Japanese just did. But don’t tell me I agree to something where there is no such consent on my part.
And the theory seems to be fueled by the premise that it is impossible to operate a search engine under opt-in. I don’t believe that.
But even if one makes up permissions where there are none, this new legislation raises a couple of questions. For example, what exactly is the scope of that “implied permission”? Is it restricted to copying content in the database search engine, or does it also extend to serving cache copies to users? The article by v.Ungern-Sternberg for example does not extend the reach of the fictional permission to serving of cache copies (see page 373), which would mean that the new Japanese exception would reach further than his theory.
And even v. Ungern-Sternberg concedes the author’s right to opt out. As most proponents of implied permission he raises that as a reason against respecting copyright. If the author can opt out easily by adding some lines to a robots.txt file, insisting on copyright is an abuse of copyright (Rechtsmissbrauch).
So, under his theory, I can opt out, as long as I do it in a machine-readable way. Not so under the new Japanese law. Any search engine is free to ignore an opt out in robots.txt, as long as the website in question does not require a password for accessing it. Again, the new Japanese exception goes farther than the above theory, and farther than current practice, where most search engines do respect at least opt-out.
Update: The above paragraph is wrong as far as it states that under the new Japanese exception search engines are free to ignore an opt-out declared by editing the robots.txt file. That would be a case where “other means of restricting the access are applied“, see also the update of the translation above.