Tuesday, December 14, 2004

Patents and Research.

Erich Kunhardt has written an op-ed in the NYT today suggesting that "inventing", or the filing of patents, become a part of academic life, along with research and teaching. Prof. Kunhardt unfortunately never makes clear how he distinguishes inventing from research itself.

The primary purpose of a patent (at least originally) was to enable innovation by encouraging the dissemination of ideas while protecting the inventor. The bargain went somewhat like: "you agree to disclose the nature of your invention, and in return no one else gets to make money off of it". Not anything can be patented: the usual test of what is patentable is that it should be novel, have utility, and be easily describable and reproducible. As was drilled into me a couple of times by patent experts, a patent does not give you ownership of an idea: it merely prevents anyone else from making money from it, for a period of time (17 years for "standard" patents).

The world of patents today bears only a small similarity to this original view. Many companies generate patent portfolios for use in cross licensing agreements, in which case the number of patents filed is often a useful statistic (quality be darned). The US PTO has laughably bad standards for accepting patents, leading to absurd situations where a 6 year old patents swinging sideways.

So at least by current mechanisms, asking academics to file patents appears to be a bad idea. So maybe the academic community should come up with its own standards for the value of a patent:
However, unlike research, there is no established peer-review process for evaluating inventions, no way to evaluate the academic significance of a new idea beyond its potential economic value. That must change, and the academic community, perhaps with a push from the professional societies and the financial support from the government, should take the lead in clarifying the principles for doing so.
But Prof. Kunhardt never quite says how a patent should be evaluated "beyond its economic value". In fact, earlier in the article he says:
Patents, along with available investment capital, are an excellent measure of the potential for job creation. America's competitive advantage in the global economy has long rested on our ability to generate intellectual property - patents and other expressions of creativity - and to leverage it by creating companies or increasing market share.
I would argue that the primary purpose of a patent is economic: this is why the utility clause is important when establishing the value of one. The novelty and reproducibility elements are already familiar parts of any research endeavour: to publish we need to be novel, and reproducibility needs no defense.

The line separating the university from the technical school is already blurring. Students already view a university as a place that provides a service, rather than as a place where one gets an education. Start making researchers file patents as a condition for tenure, and the quality of "blue-sky" research will go down tremendously as one focuses on the economic factors influencing one's research rather than the scientific ones.

Update I: emmous points me to this wonderful article by Jeff Ullman (shame on me for not knowing this) on the patent process. It has some interesting anecdotes from his time as an expert witness in patent lawsuits.

Update II: Jason Schultz, a lawyer at EFF, writes this commentary on the current state of affairs in the patent world. (salon.com - paid registration or free one-day pass)


  1. Here is one interesting article about patent by Jeffrey D. Ullman.

  2. But hasn't that already happened to some extent? IIRC, Brin and Page patented PageRank at Stanford and have exclusive rights to a license from the University for a certain number of years.

    As for tenure, academics are already burdened with the publish or perish thing. They seem to have to focus on non-blue-sky kinds of things. Things that they can readily publish soon or else they won't get tenure.

  3. I find the discussion of patents as a measure of productivity understandable, but misguided. In cryptography, the lesson we've learned is that patents more often tend to restrict adoption of a technique rather than encourage it. The few exceptions, such as the RSA algorithm, are for needs that were so compelling that there was no other way they could be met. Therefore, I try not to patent things I come up with whenever possible. It's hard enough to convince someone to change even when you're giving something away for free...

    (On the other hand, there's at least one patent out there with my name on it. Them's the breaks of working in startups and industry research labs - you don't have the luxury of not patenting things.)

    -David Molnar

  4. This comment has been removed by a blog administrator.

  5. This comment has been removed by a blog administrator.

  6. This comment has been removed by a blog administrator.

  7. This comment has been removed by a blog administrator.

  8. This comment has been removed by a blog administrator.

  9. This comment has been removed by a blog administrator.


Disqus for The Geomblog