Last night I went to a talk at Temple University given by John Barrie, CEO of Turnitin. It took me back to the days when I learned about natural language processing. Turnitin seems to have—quite deservedly—become an indispensable service for schools, universities and academic publishers. I asked Dr. Barrie about the IP protections on their code, but my question was probably not clear enough as he talked about one or two other things I mentioned in my question. The question I wanted to ask was: Given that this system is now indispensable for teachers and researchers, shouldn't the code be open? What happens if the service becomes unavailable, e.g. if Turnitin gets bought out by Google, which then loses interest in the business? If the code is proprietary then scholarship and education have a big problem. Also, would it not be safer if everyone could see and maintain the code producing the fingerprints and searching the databases? Otherwise, how do we know that some bias is not being introduced in the checking process? I am sure Turnitin's answer would be to judge them by the results they produce—if their results are strange then they will lose business. But how do we know if their results are strange if we can't see how they produce the results?