[d@DCC] Unintended consequences of treating the anonymous part of the Internet as other than BY-NC-ND

[mskala at ansuz.sooke.bc.ca]

On Sun, 22 May 2005, Russell McOrmond wrote:
> Matthew and I seem to agree on the following about files offered to
> anonymous audiences.
>
>   - That permission for royalty-free copying is granted for files
> offered to anonymous audiences under specific conditions:
>      - Attribution must be kept (how is another question)
>      - non-commercial use (you can not be selling the work)
>      - verbatim (you can not change it in any way)

Sorry to be difficult, but that's really not what I think at all.  I think
that posting something on the Web, without a licensing notice, only means
that permission has been given to copy the file in the ways the technology
of the Web demands - including, for instance, viewing it in a Web browser,
caching it in a proxy, and *maybe* (I'm undecided about this - it may fall
under fair dealing for research, or news reporting or something, or it may
be an implicit license that can be revoked by robots.txt) archiving the
document in a facility like archive.org or the Google cache.  Any other
rights a visitor to the site may have come from fair dealing, not from
permission.  The important difference is that because they're fair dealing
rights, they can't be taken away by the copyright holder.  You don't have
permission for fair dealing because you don't *need* permission for fair
dealing; fair dealing is allowed even over the objections of the copyright
holder.

On the other hand, the only way that permission to receive the document
via HTTP could validly fail to be given would be by the HTTP server not
providing the document in response to requests; otherwise the copyright
holder would be in an unfair position to bind site visitors to a contract
they haven't been fairly notified about and given the chance to negotiate.
Note that I don't believe "terms of use" documents on Web sites can bind
visitors just by the act of visiting the site, notwithstanding that they
virtually all claim to.

With fair dealing, there is generally an attribution requirement.  There
is sometimes a non-commerciality requirement, depending on the purpose of
the fair dealing - but some kinds of fair dealing are permitted to be
commercial in nature!  For instance, I'm allowed to print out your Web
site in order to read it myself because I find paper easier to read
(that's private study) even if I'm doing so in order to evaluate your
company as a possible investment (that's a commercial purpose) because the
non-commerciality requirement doesn't apply to private study.  I'm also
allowed to write a critique of your novel, including quotes, and *sell*
copies of my critique without paying you a royalty, because
non-commerciality isn't a condition for "criticism" fair dealing.

I'm annoyed by the recent trend in some quarters to attempt to exclude
publication from fair dealing - e.g. the "how do we allow one copy but not
a million copies?" discussion:
   http://ansuz.sooke.bc.ca/lawpoli/copyright/consultations/2004051201.php

Fair dealing includes personal use, but it's important to remember that it
also sometimes includes the right to publish quotes from the material,
even for profit.

I don't know what "verbatim" means, I don't think you can define that in a
technologically neutral way, and I don't think it's a requirement for fair
dealing.  I'd substitute a requirement for reproduction to be
non-misleading, but I think that might be based in libel law rather than
copyright law.  Problems with "verbatim" include, does "verbatim" mean it
has to be "complete"?  Then do I have to reproduce your entire Web site on
my own server in order to comment on it?  Does "verbatim" mean I can't
convert the file format?  Does "verbatim" mean I can't rescale an image?
The plain meaning of "verbatim" seems to be something like "in the same
words", but how does that apply to something like an image that can't be
expressed in words?  And so on.  There may be an appropriate place for a
"verbatim" requirement in a Creative Commons license, but there the author
has the chance to select the precisely correct license, or not use it, so
that the requirement will be applied to things for which it makes sense.

Incidentally, one reason I haven't jumped on the CC bandwagon a lot more
enthusiastically myself is that I think the Creative Commons licenses seem
to encourage authors to voluntarily give up privileges that I don't think
the authors could ever legitimately claim in the first place.  There are
sound legal reasons for drafting the licenses that way (greater
certainty), but I don't want to grant exceptions that would prove the rule
of authors being able to claim excessive privileges.

Above I'm talking about the current "fair dealing" law rather than a
proposal for a new one, but I think that the current law on "fair dealing"
is generally not broken and generally doesn't need to be fixed.  The
educational exemptions could be clarified, technology-specific references
removed, and protection for parody added (ideally, with a clear indication
that parody still isn't an infringement of copyright even when the
copyright in question is not the target of the parody - see the Penny
Arcade/American Greetings and JibJab episodes); but none of those are of
overwhelming importance.  I don't think documents posted on a Web site are
any different for copyright purposes from any other documents published to
an anonymous audience, for instance, the religious tracts one of my
neighbours has recently taken to leaving in my apartment building's
laundry room.

Why does it matter that the audience be "anonymous"?  I don't think things
would be any different if the audience wasn't anonymous, and if it *is*
relevant, someone could start raising issues about how anonymous Web
browsers really are.

>    - what form does this attribution take?

Depends on the case, and should not be codified in a general way.

>    - does the "location" (URL, filename, etc) form part of the attribution?

You've mentioned this "URL as attribution" concept a couple of times now,
but I don't know how you got that from what I wrote, or why you think I
think it's relevant.  What impression did I give you about my view on
that?  What did I say about URLs forming part of the attribution?

>    - In order to make false claims of authorship (violate attribution),
> does one have to DO something (State in some other page that the author
> is different than it is), or FAIL to do something (IE: fail to indicate
> in every page that references the document who the copyright holder is)

The attribution requirement isn't just that you mustn't claim to be the
author yourself (if I said it was, earlier, I hadn't thought it through
completely); the requirement is that you must inform readers/viewers of
who the author is.  No representation is still a misrepresentation,
because it isn't the representation you're obliged to make.  You have a
positive obligation, even if I've usually phrased it in negative terms.
You must attribute correctly, so you must attribute at all.

In the specific example, which you seem to be avoiding, of copying an
image and embedding it onto a Web page, there's also an implicit
representation, when none other is made, that the author of the page is
the author of all the embedded images, because that's the way Web sites
work.

I also think courts generally take a dim view of the "I didn't do
something bad, I only failed to do something good!" defence, and they
should.  It's often not a meaningful distinction.

>    - what about translations/transformations.  Is permission granted for
> the document modifications made by anonymizer (changes URLs in HTML
> documents) and automatic translation services?  Can these be transmitted
>   to the person requesting under a personal/private use type of
> exception only to the person requesting it, but never redistributed to
> others?

It's because of those kinds of issues that I'm disinclined to endorse a
"verbatim" requirement.

>    - what about printing for personal/private (obviously non-commercial)
> use? What about archiving on other than paper media?
>    - (part of the current policy debate) Is use by educational
> institutions considered commercial use or non-commercial use?

I think those points probably can, and definitely should, be answered by
"fair dealing" rules.  Note that private study doesn't have to be
non-commercial to be allowed by the current "fair dealing" rules.

> and Archive.org are doing.  Google has a banner at the top for HTML
> documents, but not for images.  Does this mean that it is
> document-dependent whether Google is considered infringing? They are
> "failing" to do something to show attribution for documents other than
> HTML documents:
>
> http://images.google.ca/images?q=tbn%3Awww.flora.ca%2Frussell%2Fpic%2Frinaruss-s.gif

I'd say it's more like *you're* providing instructions on how to remove
the attribution Google provides in their image search service.  They
didn't fail to give attribution; you chose to skip past it.  A normal user
of Google's service, accessing it in the way Google intends people to
access the service, would not see the image at the above URL without also
seeing the attribution information provided on the enclosing page.

> In the case of HTML documents they are modifying the HTML file, which
> can be said to be a violation of the "verbatim" requirement?

That's why I don't like a "verbatim" requirement.

> Is the URL really what is being trusted?  What if in the future we move

What's trust got to do with it?  For that matter, what have URLs got to
do with it?

> In this case there are those that would argue that what Anonymizer is
> doing is not a "verbatim" copy as it is modifying all HTML documents.

I could argue it either way.  The documents aren't byte-for-byte
identical.  On the other hand, they have the same words.  I don't like the
"verbatim" requirement.

> Would those in this forum consider this to be an infringement?  Is this
> not an infringement because these pages are not being "republished", but
> being considered part of the reception by the user of anonymizer?

One claim the anonymizer operators could raise would be that they are a
network intermediary.  It would have to be decided in the actual case,
though; I don't think this is one I can evaluate as a hypothetical.
-- 
Matthew Skala
mskala at ansuz.sooke.bc.ca                    Embrace and defend.
http://ansuz.sooke.bc.ca/