Fairhaven, The River

Ownership and copyright

This weekend's work finds yet another oddball use for "own".  In the world of BPEL and BPMN the "owner" of a task is the person who is currently working on it.  Another way to spread confusion.

John asked about an "XML schema" for privacy.   There are two answers to this.

1. I think a consent structure similar to the Creative Commons copyright licenses is feasible.
2. A general XML encoding for privacy, consent, or copyright remains a matter for research and exploration.  The translation of law and decisions into a structured form has been a matter of legal research since the mid 1970's or earlier.  It's still research.  There is also ongoing research into forms of license and contract.  None of these are ready for routine use.

A bit of history

While I think that common consents are feasible, it will take a lot of time.  The history of work that led up to the the Creative Commons licenses shows how much time it can take:

• 1974 - The CONTN commission recommends that copyright law be extended to software.
• 1980 - Copyright is extended to include software
• 1988 - The Emacs General Public License is written.  This is the first effort to formalize a publicly usable license.  It is a reaction to the serious problems resulting from inadequacy of the Gosling license for early emacs work.
• 1991 - GPLv2 issued.  This became a major influence on copyright and public domain thinking.
• 2002 - The Creative Commons is established.  The needs for open culture, open publications, etc. are not met by the GPL and similar variants.
• 2009 - The Creative Commons licenses reach version 3.0 (the current version).

Creative Commons is now to the point where there are about a dozen standard copyright licenses.  For a very large number of people it meets their needs.  You answer a few questions and get a recommended license.  You also get an HTML code snippet that identifies the license, and provides links for further information, attribution, etc.

This has reached the same level of ease of use as the typical publisher's copyright assignment forms, with substantially better commonality.  Every publisher, etc. has their own standard assignment forms,  with their own terms, etc.  They all have lawyers who customize things to maximum advantage of the publisher.  Common Criteria took the perspective of the authors, and after a few years experience with actual author preferences, has a set of common licenses for the open culture authors.

A Creative Commons for Consents

There have been occasional discussions of whether privacy rules could be managed like copyright.  Zittrain has written on this, and there was a recent seminar on the topic at the Berkman Center.  These make clear that the larger issue of privacy remains extremely complex.  But in a much narrower domain like patient privacy consents, there is a better chance for success.

I can see a situation where a group defines a set of patient oriented consents.  These consents would differ from much of what I've seen in current work by being patient perspective rather than provider perspective.  Rules like the HIPAA rule are impenetrable to the typical patient.  They deal with the many issues that are visible to the provider, rather than the issues that are visible to the typical patient.

I expect that getting this right will probably take a decade, given that it took two decades for the copyright licenses to evolve a public oriented set.  We can learn much from that effort, but one of the things that you learn is that real experience was crucial to the evolution of these licenses.  Real experience takes time.

"Own" is the wrong verb

The verb "own" takes us down the wrong path in understanding the issues around medical data. This comment is triggered by Keith's post, but this misuse of "ownership"has a long and sad history in medical records and informatics.  It was less of a problem in the 19th century, when copiers and computers did not exist.  Then, ownership of the physical paper holding the record controlled the possible uses of the medical data.  But "own" was never really the right verb.

All of the semantics and law implied by the word "own" are derived from management of atoms.  You "own" things.  These things are made of atoms.  "Ownership" deals with breaking things, stealing things, selling things, lending things, etc.  All of these make sense for collections of atoms.  If you sell or lend something, the atoms are gone.  You no longer have the atoms.  None of these verbs make sense for medical information like patient records.

Copyright is much closer.  It uses the proper verbs.  Copyright deals primarily with two questions:

1. Who is allowed to make copies?  This may be a restricted right, as in allowing a magazine to print copies of an article in one and only one issue of that magazine.  Electronic copies, CDs, DVDs, audio, etc. are not allowed.  Or it may be an unrestricted right. There is a huge body of details, expectations, defaults, and considerations around making copies.
2. What uses are allowed for those copies? A host of potential use restrictions are normal in copyright.  A copyright license normally specified exactly what uses are allowed, and what uses are prohibited.

These are the issues that arise with patient records.  The entire issue of disclosure and disclosure consents is defining who can make copies and what uses can be made of those copies.  Privacy rules are all around allowed copying and use.

Copyright also deals with more minor concepts properly.  In medicine the actual creator of the record is a machine or doctor.  From the legal and social perspective the ownership world would assign ownership to the creator, subject to sales contracts, mechanic's liens, etc.  In the copyright world this is a "work for hire".  The terms found in "work for hire" accomodate all the uncertainties of the creative process in the arts and medicine.  The contract law around manufacturing of things make little sense in medicine.

Copyright uses the right verbs and structure but the body of rules and expectations is all wrong for medicine.  It distinguishes between owning a book and having the copyright to the contents of a book.  This is the same as the distinction between owning the paper medical records and having the rights to use and make further copies of those records.  It's normal for a doctor to own the paper records, have the rights to use for treatment, and have no right to disclose to the public or use for research.  This falls nicely onto the copyright treatment of owning a copy of a play, having the right to read it to your children, and having no right to put on a public performance or make copies for the actors.  Copyright lacks the proper words for the rights and uses in medicine.

We should be adding words to deal with the kinds of uses that make sense in medicine.  Instead of public performance, derivative work, plaigarism, and "fair use" we need kinds of medical use, purpose for disclosure, etc.  The "work for hire" concept is there, but all the details and contractual rules make sense for employees creating documents, photographers on assignment, and artists painting to a commission.  What are the right words for referrals, treatments, reviews, consultation, billing disputes, etc.?

So forget "own" and the world of atoms.  For medical records in the electronic age we want:

1. Rights to make copies, (who, what, where, why, when, ...)
2. Rights to use copies, (who, how, where, why, when, ...)

For those who want to dig deeper into the world of copyrights, I suggest the Groklaw references or the MIT Open Courseware short course.  Copyright practices are a large subject to cover, and you will find that all of the work and examples has been around the creative arts and more recently software.  The basic concepts and verbs are right for medicine, but the rest needs to be defined.