TurnItIn.com – Copyright and "fingerprints"

With all of the debates about the legality of TurnItIn.com, I sometimes forget the basic facts. My intellectual property is submitted to TurnItIn.com. They store it. They use it to provide a service and profit from it. Their service could not exist without my intellectual property and the intellectual property of thousands of other students.

That said, let's take a look at their defense. It hinges on the claim that they do not make copies or true derivatives of my work. Here, I'll show you why that's false.

I read this in TurnItIn.com's Canadian Legal Document:

The TURNITIN system makes a "fingerprint" of the work. This is merely a digital code that relays the fact that certain pre-defined content is present in the work. The ‘fingerprint' does not include any of the work's actual contents, and is therefore neither a copy nor a true derivative of the original text.

First, it's not clear to me what they mean here. They don't define "the work", or what they mean by "fingerprint." "The work" could be the papers stored in the database, or a paper currently being tested.

In technical terms, a fingerprint is something that "characterizes an object or substance." A common example is the MD5 algorithm, which can return a ‘fingerprint' of a block of text. It has two important properties: 1. Given a fingerprint, you can not determine what text it represents, 2. The same text, and only the same text, will always yield the same fingerprint. However, it is possible that they are using something completely different.

Regardless, the argument is that they are not making a copy or derivative of the original text papers. However, in my experiments, it seems that not only do they make, keep, and continually use these copies, but they also take the sharing and use and – worst of all – the distribution of MY copyrighted text completely out of my control. Let me explain.

In December of 2005, I signed up as an instructor, created a class, and got my "students" to submit their "assignments". I was testing how effective the detection was (the verdict: not very). The class was scheduled to end on January 5th, 2006.

Three months later, I logged in as this instructor, and was able to download a fully formatted copy of my paper. Clearly, TurnItIn.com has made a copy of my copyrighted paper, is retaining it for an undocumented period of time, and is providing it to my professor (and possibly others?) long after the class is over. But it gets worse.

I created a new instructor, a new class, and a new assignment. I submitted an "assignment" that contained text copied from the first assignment. The "originality report" correctly identified it as being plagiarized from a paper submitted to Dalhousie University. I clicked the name of the paper, and was given this message:

Because submitted papers remain the intellectual property of their authors, instructors, and respective institutions, we are unable to show you the content of this paper at this time.

If you would still like to view this paper, please use the button below to submit a permission request to the author's instructor. We will send the instructor an email detailing your request and include any information the instructor will need to respond if your request is accepted.

Dear Instructor,

Turnitin is forwarding this request on behalf of Test Instructor, an instructor at Dalhousie University. This instructor requests your permission to view the paper, "A Test of TurnItIn.com," submitted to your CSCI 0000 class at Dalhousie University on 12/03/05. This instructor has found a __% match to this paper in a paper submitted to his or her Test Class class. To grant the instructor permission to view the paper, please reply to this e-mail and the paper (included below) will be forwarded to the instructor.

Sincerely,

Turnitin

--------------Begin Paper--------------
<< Entire Text of Paper Followed >>

The reply-to on the email was the address of the instructor – so replying to the message would send the entire paper to this other instructor. At no time was the student involved in this decision, or even informed of what was happening. Control of the student's copyrighted material was left up to TurnItIn.com and the professors.

What's my point? The Canadian Legal Document, and the American one even more so, relies on this principle that only a "fingerprint" is used, and the fingerprint is not a real copy or derivative. But as I have just shown, full-text, exact copies are available on the system, and these copies are used and displayed long after the student is finished a given course. The legal document either lies about this issue or avoids it completely, depending on how you interpret it.

TurnItIn.com takes students' intellectual property, copies it, retains it, uses it, and exercises control over it, all to make money. That is not what I signed up for when I enrolled at Dalhousie University.

This paper is part of Mike Smit's continuing discussion of TurnItIn.com, as found at http://MikeSmit.com/turnitin