Wednesday, March 21, 2012

"Property Makes us Stupid," part 1

At the end of yesterday's post, I insinuated that I think the notion of "autonomy" as it floats around many contemporary ethics discussions won't do the work it needs to.  Today, I want to look at another mismatch between digitality and traditional normative concepts, this time "property."  Intellectual property - copyright, patents, and trademark, most centrally - has been at the center of a lot of battles about digital culture.  For example, just yesterday the Supreme Court unanimously ruled that a procedure designed to apply observed correlations between drug dosages and patient outcomes could not be patented, on the grounds that allowing this patent would effectively close off the natural facts of the correlations to anyone but the patent holders.

This sounds trivial enough.  But what immediately struck me (and other commentators - see here) is the possible relevance of this to a much more important topic: patents on genetic sequences.  The current law is established by the appellate decision in AMP v. USPTO, which was decided last summer.  It's a split opinion, but the governing part says that patents on isolated gene sequences are ok, based largely on the argument that genetic sequences in the body are something different from the ones outside the body and subject to patent claims - in other words, moving them does meaningfully transform them (there's at least one good reason this is true: genetic sequences in the body involve a lot of non-coding information which gets excised in the patented sequences).  What the appellate court did not do a very good job on was the preemption argument that the Court used for yesterday's decision.  That argument, recall, says that you're not allowed to patent something if, in so doing, you'd effectively block others from accessing a law of nature.  As this paper argues, that's a very plausible way to construe what gene patents do: they bar others from using the genetic sequence to develop tests, therapies, etc.  The Court's unanimous ruling yesterday suggests to me that there may be an appetite there to take up the gene patents case.  At least four justices have to vote to take up a case; here, the tea leaves following such a decision might indicate that there's four votes for striking down gene patents.  The Court's been saying for a while that it thinks there are too many patents, even explicitly a few years ago in a case about automobile accelerator pedals (not for nothing do we get the phrase "patent geek!").

What does all this have to do with digitality?  It seems to offer confirming evidence of a couple of points.  The first is that the gap between biology and informatics is rapidly shrinking.  That is, biology is increasingly a science of information, statistics, and probabilities, with human bodies being treated not so much as somatic wholes, but as repositories of information about risks, propensities for certain diseases, and so forth.  The gene patents case, for example, concerns Myriad Genetics' patent on the BRCA1 and BRCA2 breast-cancer mutations (btw, Myriad's stock slipped 5% following yesterday's ruling, and is off another 4% today, as of this writing).  Women who carry those mutations have a highly elevated lifetime chance of getting breast and ovarian cancers.  But the tests themselves only establish probabilities; they don't actually diagnose a symptom.  If it's true that 40-60% of women who carry the BRCA genes will get breast cancer during their lives, it's also true that the rest won't.  Those are terrible odds compared to the 12% baseline, but the point is that it's all a matter of processed information, not of medicine in the old-fashioned sense of "Doctor, I have a headache."  This is where a lot of the action is: the body is being mined as a source of information in order to slot people into risk categories for certain diseases.  Those who fit the categories can then make informed decisions about whether to preemptively do things like have their ovaries removed.  We thus see the rise of what Nikolas Rose calls the "presymptomatic patient."  Such an entity is only possible, I submit, when we understand the body as a source of information, as a product of information society.

The other point is that the law just isn't adapted to this.  It's a really, really good question whether genetic sequences should be subject to patents.  The existing vocabulary for patentability might not handle it well, because that vocabulary was developed long before genomics developed very much.  This problem sits at the tip of a good-sized iceberg.  To take just one more example: patentability generally requires naming a single inventor.  Such a person is the primary patent holder.  But this model is ill-suited to the current scientific enterprise, where collaborative activities are spread across large numbers of people.  It also enables companies like RiceTec to attempt to patent basmati rice.  It took a herculean effort to invalidate this patent, precisely because patent law has trouble seeing inventions where one cannot also name the inventor or point to a first publication.

When Michael Hardt and Antonio Negri, then, say that "property makes us stupid," they're partly talking about the misfit between our understandings of property and reality.  "Reality" is partly about our new "posthuman" status.  It's also, for them, the thought that insisting on proprietizing everything causes us to forget that our daily lives function quite well without asserting property rights all the time.  I'll talk some tomorrow about more of what they mean, and the assumptions embedded in our copyright regime.


  1. Gordon, This is an exceptionally keen-minded take on the human body/genome as a data set. I was immediately reminded of Myriad Genetics before I finished your first paragraph. I'm doubtful, though, that there will be any change in human gene patenting. After all, Myriad's original argument, rejected in summary judgement but supported on appeal, was essentially that this train left the station decades ago, and there's too much money at stake to turn back now. (Cf. your inclusion of the respondent fluctuations in Myriad's stock price.)
    It would be great if we as a society took a fresh -- and comprehensive -- look at gene patenting. During the first round of the Myriad litigation the ACLU stated that 18.5 % of human genes are the property of various biotechnology entities. How high does that percentage have to go before alarms start to sound?

    1. Sorry for the slow response - but there's actually news on this front: on Monday, the Supreme Court vacated the Myriad decision and kicked it back to the Federal Circuit to review in light of Mayo. So we'll see what happens... Myriad has at least bought itself some time.

      My reading of the original decision was that the Federal Circuit was restoring order - the "train has left the station" argument you make. On the other hand, this Supreme Court is starting to send a pretty consistent signal on patentable subject matter.

      I find myself of two minds on gene patents - I can see the arguments both ways.

      One final thought: I think that one effect of gene patentability has been the support of start-up biotech firms (I'm getting this from a combination of Kaushik Sunder Rajan's Biocapital and Robert Merges' Justifying Intellectual Property). Big Pharma actually would rather see gene fragments declared unpatentable - that way, their upstream costs are minimized. So a ruling against gene patents might tend to further centralization in the biotech field. I don't know what to do with that - but I do think it illustrates the complexity of what's at stake.

  2. I just go through the post really nice. I think people should go through it. Thank you for sharing.

    Property Management Software