One of the tragedies of the modern age is that we have to prepend the word “open” onto science in order to differentiate Open Science from the regular kind — you know, the kind that’s overrun with IP restrictions and practiced by way too many academic and research organizations. It should never have come to this considering science is about reproducible results under controlled conditions, a process that intrinsically demands transparency and openness. Fortunately, the alarm bells have gone off and many in the research community are taking matters into their own hands.
In my experience IP barriers in science come about in three major ways. First, there is the decreasing ability of the traditional publishing process to support reproducibility. A “publication” without data, source code (the methods), and good documentation is nearly impossible to reproduce and causes many to question what is actually being shared. Some of the computational papers I have seen are so complex that reproducing the science behind it may take years (assuming you can acquire the data), something that few do due to the pressure to create novel ideas (career recognition tends to fall to those that create something new versus those that confirm a result).

Second, there is a decided resistance to the open research process. Part of the research community is uncomfortable with the inherent messiness of innovation and do not want others to see the process unfold. Others want to protect their “ideas” due to the competitive environment that is science today, created largely by the way funding science and tenure works.

Lastly, formal IP organizations have been established at universities and other research organizations to control and license IP. I have seen so much innovation killed by these offices that it turns my stomach. While there is the occasional Gatorade success, statistics indicate that there are significantly more failures than successes and that these come at the cost of squelching innovation, the scientific process, and even business. For more information on University IP systems, see Melba Kurman’s blog post “Fear, Uncertaintly and Doubt and University IP Strategy,” which claims that 99% of university patents never earn money, or Kimberly Moore’s article “Worthless Patents,” which shows a 53.71% abandonment rate of patents that takes place after spending tens of thousands of dollars to acquire the patent in the first place.

With the emergence of networked science and the increasing reliance on computing, novel methods are emerging to address these IP barriers. These methods typically boil down to providing Open Access (publication), Open Source (software), and Open Data to support the Open Science (community). Many people are aware of the fantastic Open Access journals such as PLoS and BioMed Central; even more innovative (at least from a software point of view) are journals that accept data and software submissions and then automatically test and score the software before handing the submission off to the human reviewers (see the Insight Journal here).

Progressive publishers are also experimenting with novel processes; for example, the Optical Society of America provides their Interactive Science Publishing system by enabling linking of active documents to data that can be downloaded and interactively viewed. Indeed, one possible business model for publishers is to get back to their roots and provide services to the community in the spirit of networked science, which could include hosting open communities, data, and software (can you imagine publishers becoming software developers?). There are serious needs for such services; just consider the need to host and archive (large) computational data, a huge and increasingly important problem as data rains down on us.

As computing becomes ever more pervasive, it’s critical that we maintain the innovation highways that will enable us to drive happily into the future. Open Science is central to this; yet despite the fears of many, practicing it does not preclude commercialization. Gated communities representing well-crafted solutions and user experiences will always sell, and technology integration services that move ideas into practice offer significant business opportunities. Without Open Science, we run the risk of creating feudal societies that lock innovation behind intellectual walls, with closed data repositories that we don’t share and software that we have to needlessly reinvent. In the future, success should be measured not only by impact factors, but by how well research output is shared and reused – to see further, we must produce results that enable others to stand on our shoulders.

About the author: Dr. Will Schroeder is president, CEO and co-founder of Kitware, Inc., a company focusing on the creation and support of open-source, leading-edge scientific computing software. He contributes to several open-source communities including the Visualization Toolkit where he is a developer and first author on the companion VTK textbook.