As A Grad Student, Who Owns My Code?

Question

[Chris Miller]

If I write some code while a graduate student at a university, who owns that code? If they're scripts used for data analysis, but clearly aren't substantial enough to warrant publication or patents, can I safely open-source it and post it online? Will this somehow run me afoul of my school's IP office?

For the record, most of the scripts would be of the data-munging variety. There's no novelty or lost revenue if I just post a 4-line script to convert file format A to file format B.

So, is this one of those "better to ask forgiveness than permission" type deals, where I'm unlikely to have any problems? Or is it something I should seriously be concerned about?

Answer 1

[Egon Willighagen]

There are plenty of scientists (like me) who argue that it's best for science is the source code is Open. As such, I agree with Andrew that if your advisor agrees, you are safe; I do not think the university will tell it's researchers what good science is.

But that's about licensing, not about who owns the code. There too, make an agreement with your supervisor, and most scientists I worked with keep the copyright with the authors.

However, even if the copyright is written to be with the university, you can still make the source code Open Source, as you would argue this is needed for peer review. And again, who in the university will argue that peer review is not important for science.

(Yes, I know the argument that closed source can be peer reviewed too, but there are not organizations like publisher who get that done for you, so Open Source is just the only viable available option to scientists.)

Bottom line: talk to you supervisor today, and agree on making it Open Source with you as copyright holder.

[Phis]

As regards open source: certainly some journals demand that any scripts/software used in a publication be made publically available/open-sourced. As to who actually owns the code, I suspect the situation may be different from university to university, actually. (At least, looking at the contracts/regulatations I've seen in the past would suggest that).

[Malachi Griffith]

Even if a journal does not require code be made open source, you as a peer reviewer can strongly recommend it if you are an open source advocate...

Answer 2

[Andrew Su]

I think you should be most concerned about what your adviser thinks. If s/he is on board with open sourcing it, then they will likely go to bat for you in case the lawyers end up having an issue. OTOH, if s/he isn't on board and you do it anyway, then I think there are many more possible negative outcomes, mostly related to reference letters and the like... IANAL, but my two cents...

Answer 3

[Will]

I'm not sure how it is at all universities but the two that I have been involved work like this:

It is assumed that all code, inventions, IP, etc. are owned by the university if they were created using university resources (employees, lab-space, etc.). However, they will transfer the IP to the scientists involved following the completion of a seeming blizzard of paperwork.

This allows the University to monetize the IP of anything in its portfolio even if the scientists are too "lazy" to sell it. It also allows the University to get "good press" whenever a scientist requests the IP and creates a thriving business.

I agree with the previous answers that what really matters is what your adviser thinks. Mine actually made me fill out IP forms for every single piece of software ... which was a nightmare.

[Mrawlins]

The copyright technically belongs to the university. Most of the things you write will be of no use to them, they won't care, you do what you like and they'll never know. If you run afoul of this, though, you can get in some murky waters. If you're worried about it, you probably ought to talk to someone in the legal or tech transfer office at your institution. Your adviser knows what's useful, but legalities don't always line up with the science, and this is a legal question.

Answer 4

[Mary]

I can remember the day my class cohorts and I sat in the department office and got to the form where we had to sign that our IP belonged to the University. This was stuff we had to fill out on the first day. Without any legal guidance.

I don't know if they all do that, but I remember being pretty miffed by that. One of my classmates had the same reservations. But we thought we were stuck. It didn't end up mattering. But it was disturbing.

Be careful about what you signed on intake. You may have already signed it away.

Answer 5

[David Quigley]

I think Egon is basically correct. For an example of how amazingly complicated this issue is, browse this email exchange, part of an informative (though U of Washington CS department-specific) site (http://www.cs.washington.edu/commercialization/). Your supervisor may or may not have the correct answer, since this can be a very counter-intuitive area. Work with someone from your university technology transfer office to get the official answer for any non-trivial piece of code. Google "research software ownership" with your university if you don't know where to get started, as many schools have standard information online.

Answer 6

[Derek Fairley]

As others have noted, it depends mainly (i) what country you're in , and (ii) what you've signed. I've learned something about this process here in the UK through filing patents arising from university / NHS research - although not software related. In the UK & europe I believe that "programs for computers" are not patentable in their own right, whereas in the US software can be patented.

For stuff which can be patented, in the UK your employer owns your IP in most circumstances (see section 39 of the Patents Act). There is a catch-all term in the Act ("circumstances ... such that an invention might reasonably be expected to result from the carrying out of his duties") which effectively means that IP for anything you "invent" which is related to your day job is owned by your employer - even if you have the idea at home in the bath.

As a student you're not technically an employee, hence you may be required to sign your IP over to the institution. You're still the "inventor", and you have a right to a share of any resulting income. UK universities will generally give 1/3 to the inventor/s.

Health warning: I am not a lawyer or attorney etc. etc. ;-)

Answer 7

[Nicojo]

The answer to your question is highly dependent on the place you work at:

As several people have pointed out, if you have signed a paper that states your work is owned by the University, then that's pretty much it.

However, some countries have laws that protect people. In Sweden for instance, the person doing the work owns the IP and there's nothing the University (or the principal investigator) can do about that, besides burning your lab-book and any copy thereof you may have made.

Finally, discussing this with your supervisor is a good thing. But don't only rely on one source. Many people (including the supervisors) are misinformed. Always look for the source of whatever anyone tells you when it comes to IP.

Answer 8

[Dave Lunt]

It may depend on where you work, but I would also add that you shouldn't expect to get the right answer from your university. They will almost certainly tell you that they own it. But do they? Your supervisor may say that as long as you arrange it before hand with him/her you can licence it as cc etc. But is that right? This is a legal issue and universities and supervisors tend to say the thing they want to be true!

There have been several protracted legal cases about who owns your lectures as an academic. Universities pretty much always claim they do and some have even taken legal action against staff, they have always lost I believe, lectures are NOT owned by institutions.

Unless you have specifically signed something to the contrary do not accept that somebody else owns your work just because they say they do.