2006/9/8, Perry Willett wrote

> 1) We base our decisions about access to any item on information in our MARC
> records (and for MARC fans, in the fixed fields). If information is
> incorrect or incomplete, we may block an item that should be freely
> available. Please let us know if you run across an item that you believe to
> be restricted in error, by using the Feedback form at the bottom of our
> Pageturner.

The MIRLYN evidence refutes this words.

First I have to speak about Google.

I have enough experience with Google Print (now: Google Book Search)
since March 2005, see
http://wiki.netbib.de/coma/GooglePrint
to know that Google is blocking thousands of Public Domain books for not-US users. There is no clear pattern and Google is giving no
information about the rules it has. With hundreds of tests I could
find out that for foreign users (although I only have experience with
German IPs and not testes French or UK open proxy server) all
publications after 1864 (i.e. 1865 or later) are not available as full
text. We see only snippets or nothing.

And there is a lot of pre-1865-stuff we are blocked too because of
unclear reasons.

In the US you cannot proof easyly the following assertion:
Link
is the list of matches for full text books published between 1865 and
1900 which Germans can see and the number of matches is 130.

With nodeproxy.com (anonymizer using free US proxies) the number is
299. It is clear that this date search is'nt working correctly but
there is no reason to hesitate that US citizen can see a bulk of stuff
which Germans are unable to see.

Mathematicians of insurance companies may correct me but I think it is
true that from the authors living in 1906 only 50 percent were alive
still in 1936. In Germany and the EU we have a clear rule for
copyright expiration: 70 years after the death of the author.

I do not think that it is important to take into account the Mexico
copyright term (100 years after the death of the author).

If one will not check each entry separately the only reasonable rule
for digitizations project which will avoid conflicts with copyright
terms of foreign countries is to handle those works as PD which are PD in the US (pre-1923) and of which the creator is dead 70 years.

In reality US and Canadian projects are not so cautious: They are
making available freely works which fits the national PD criteria but
are EU protected.

Accepting the reasonable rule it is clear that Google has a copyright
paranoia when preventing German users from viewing works published in 1870 or 1880. No one will sue Google (or UMich) for these items.

And now to Mbooks (UMich).

As everyone outside the US can proof with an US proxy there are
hundreds of books which are not available for public viewing outside
the US. It is therefore a fact that UMich is using an IP-based
blocking mechanism like Google.

It is the same mechanism like Google? No, the results are not in all
cases the same. One can see in MIRLYN works published in the US after 1864 which are blocked by Google for non-US users.

> 2) We take a careful approach to access. We cannot consider and research the
> copyright status of every item. Instead, we have developed fairly broad
> definitions of public domain that we can apply to the entire collection.

It is clear that this isn't true. UMich is using at least a part of
Google's paranoia rules when blocking foreign users.

As I have understand the Mbooks rules all books can be viewed from the
UMich community and the general public. UMich students or staff cannot
see copyrighted books which US citizen cannot see. But "general
public" means: US general public not worldwide general public if
foreign works are concerned.

Some examples:

The "Chroniken der deutschen Städte" are a multivolume edition of
German town chronicles. The first volume was published in 1862. There
is no evidence that the XIXth century volumes are causing copyright
problems - the majority should be definitively in the PD also in
Germany.

German users cannot see a single volume of this edition.

US users can see 3 Luebeck volumes and 1 Dortmund (? I cannot check
this exactly because Google is blocking proxies like nodeproxy. The
page image - I need the title page - cannot be seen but it is possible
to search and to check if the books are marked as full view).

Yesterday I have checked with an US proxy (not an anonymizer like
nodeproxy) that US users can see in MIRLYN some additional volumes
(e.g. Nuremberg volumes) German scholars are blocked by MIRLYN.

Strassburg 2 (published in 1871) is free for US citizen at:
http://hdl.handle.net/2027/mdp.39015026717424 (checked with nodeproxy).

German users only see the message:
" Page images and the full text of this item are not available at this
time due to copyright restrictions. "

The possibility to search in the book is useless because Google's OCR
cannot recognize Fraktur. See e.g. a Geibel poem in UMich's text mode:
http://de.wikisource.org/w/index.php?title=Die_beiden_Engel_%28Geibel%29&oldid=90866

Bericht über die fortschritte der anatomie und physiologie ... 1856-71

All volumes not available for German users.

Vol. 1 checked with proxy: available for US users.

I spent a lot of time checking some other journals and can say that
UMich is restricting the access to journals in the German language in
the same way for foreign users like Google. There is no evidence for
a own "UMich copyright policy". There is no evidence for a reasonable
rule taking into account foreign laws. Mr Willett is wrong when
asserting that this is a single case problem.

UMich is asserting that they are doing diligent copyright clearance
("We take a careful approach to access"). The fact seems to be: they are doing nothing.  I have spent a lot of time to find out that UMich has the same
IP-blocking policy like Google if foreign works are concerned. I have
no explanation for the 1865 date after which foreign users are usually blocked by Google or UMich.

(If you are asking: How can I filter Mbooks - there are also licensed
sources from Gale etc. as "electronic resources" I do not want in the
list of MIRLYN matches? The answer: Try "keyword" as additional
keyword.)

Geibel died in 1884. All three Geibel works in MIRLYN with Mbook
access (published in 1871-1873) are not free for foreign users. Peter
Suber has shown at
http://www.earlham.edu/~peters/fos/2006_09_03_fosblogarchive.html#115766385751334831
that no one outside the US has access to the Geibel test item (Gedichte, 1873).

Friedrich Spielhagen died in 1911 (source: MIRLYN) - not free for us
http://hdl.handle.net/2027/mdp.39015014810520

Sammlung spanischer gedichte. Ausgewählt und mit anmerkungen versehen
von Adolf Kressner. Kressner died in 1907 - the work is not free for
us.

Homer / von Georg Finsler. Finsler died in 1916 - not free for us.

Studien /von Johannes Scherr. Scherr died in 1886 - not free for us.

Aus Amerika: über schule, deutsche schule, amerikanische schule und
deutsch-amerikanische schule. By Rudolf Dulon who died in 1870 - not
free for us.

Aus natur und wissenschaft. By Ludwig Büchner who died in 1899 - not
free for us.

Einführung in die theorie und den bau der neueren wärmekraftmaschinen.
Von Ingenieur Richard Vater. He died in 1919 - not free for us.

Gespräche über Gott und natur und über unsterblichkeit. Von dr. J. C.
Bluntschli. He died in 1881 - not free for us.

Truly a sad litany!

In short: I haven't found any German book published after 1864 and
before 1923 that is available as full text for us although the author
is according MIRLYN longer dead than 70 years.

I have checked also some French titles with the same result (e.g.
"Histoire de la philosophie Cartésienne") - UMich doesn't care for the
author's year of death although it is provided in MIRLYN.

Thus the Geibel case is no exception. There are hundreds or thousands
of other German (and French and Italian and ...) Public Domain books
which are available for US users but not for foreign users. UMich
students or staff have no privileges: if an book is regarded as
US-copyrighted by UMich they can only do a search. If the book is in
Fraktur (and published after 1922, the critical year for US access)
this possibility is worth nothing because Google's OCR cannot handle
Fraktur.

If there is a clear indication in MIRLYN that the author is dead more
than 70 years (like Geibel) and the book was published before 1923
UMich can give wordwide free access to the Google files without risk.
Regarding books before 1900 there is a very low risk that they are
copyrighted in the country of origin with a 70 years pma term. It
would a reasonable rule to give also free access to them.