From Peter Suber's Open Access News
http://www.earlham.edu/~peters/fos/2006_11_05_fosblogarchive.html#116318960519020369

If you remember, Google blocks access to Google-scanned public-domain books outside the US. Finally we have Google's explanation:

Only books in the public domain -- books no longer under copyright -- have the download feature available. For users in the United States, this typically means books published before 1923. For users outside the U.S., we make determinations based on appropriate local laws. Since whether a book is in the public domain can often be a tricky legal question, we err on the side of caution and display at most a few snippets until we have determined that the book has entered the public domain. These books...may be in the public domain, but until we can be sure, we show them as if they are not.

We're working quickly to digitize and index as many books as possible so we can make Google Book Search truly comprehensive and useful. One way to treat digitized books that may be in the public domain would be to exclude them from the index until we were sure. However, our goal is to make the index as useful as possible, and that means including books as soon as we can rather than waiting for a perfect determination of public domain status. So, some books may initially show up in "Snippet View" and then later, be expanded to "Full View."

Comment. In most countries on Earth the duration of copyrights is the same as in the US. So why isn't it easy for Google to provide access to all of those countries as soon as it decides to provide access to the US?

At least Google admits that these books "may be in the public domain" and that it's temporarily treating them "as if they are not". That is, it hasn't wrongly classified them, but only delayed classifying them. Still, in most cases, it's hard to understand why any delay is necessary.

We have in this weblogs the following entries (in English) on this topic:

http://archiv.twoday.net/stories/1073534/ (How Google Print is Blocking Not-US-Citizens, 2005, Oct 19)

http://archiv.twoday.net/stories/2609488/ (Burning Money: Google's Scanning Nonsense 2006, Sept 1)

http://archiv.twoday.net/stories/2643658 (Google and Michigan block access outside U.S., 2006, Sept 8)

Comment:

It is not right that in the most countries the copyright rules are the same as in the US. Unfortunately the pre-1923 rule is US-specific. Most countries have a 70 (or 50) years post mortem auctoris term (the EU has 70 years).

Arguing against Google (and UMich) is speaking with a wall. Until now the best solution for people outside the US is to install a US free proxy (I have choosen a separate browser, Firefox users can use SwitchProxy http://www.erweiterungen.de/detail/SwitchProxy_Tool/ ). Downloaded PD works can be put on free respositories like Wikimedia Commons:
http://commons.wikimedia.org/wiki/Category:De_Wikisource_book

If UMich is unwillingly to change the rights management according the life data of the authors which are given in the UMich OPAC - why should Google do any work in this direction? Google Book Users are apparently accepting the restrictions. There is no broad discussion on this topic (nor a small, Peter Suber is the only I know who cares on it).

The explanation Google gives is not sufficient for the lot of pre-1800 works Google presents only as snippets. No one can believe that these books are still protected in any country in the world (Mexico has a 100 year pma term).

See e.g. for the date 1600-1650:
http://books.google.com/books?q=date%3A1600-1650&btnG=Search+Books&as_brr=0