Over at Adventures in Ethics and Science, Janet
Stemwedel, our resident ethicist, has been writing about academic
dishonesty and how professional researchers should respond to it.
I've been on the receiving end of dishonesty on three occasions -
ranging from a trivial case (arguably not dishonest at all) to the profound.
I'll describe my three experiences, along with how I did respond to them, and how I could have responded to them. Unfortunately, my experience isn't very
encouraging, and most of my advice comes down to: always, always keep a paper trail: it can't hurt, but don't count on it being useful.
I don't want to be too discouraging here. I don't think that there
are many dishonest researchers out there. The overwhelming majority of professional
researchers are scrupulously honest people who give credit where it's due, and who would never do anything to
take credit for anyone else's work, who would never steal an idea, and who would
never do anything even remotely questionable when it comes to
intellectual honesty. The problem is, it doesn't take much to poison
the well - one person out of a hundred is easily enough to create a
huge problem. And the nature of power and politics in research makes it
possible for that dishonest one to get themselves into a position where
people are scared to come forward about it.
My first and least serious experience involved a conference
paper. In computer science, most peer reviewed publications are
actually in conferences, not in journals. Work is published first in
conferences, and conference papers are actually more prestigious than
journal papers. (Acceptance rates for journal papers are quite high,
well above 50%. Acceptance rates for good conferences are typically
between 15 and 20 percent.) In general, you don't write a journal
paper until you have multiple conference papers that you can cite in
it. The conferences are really where it's at.
So I submitted a paper to a top conference - acceptance rates in
the 15% range. The reviews were the best I've ever gotten. The reviews
summarized their opinions for different categories using a 1 to 10
scale. My paper averaged between 8 and 9. Two of the three reviewers
recommended it for the "best paper in conference" award. But it got
rejected. The program chair was offended at the way that I described
the program chairs system in my related work. (I wasn't critical of it; he
just thought that I didn't discuss it in enough depth.) So he spiked it.
In my opinion, there's something wrong with the PC overriding the
opinions of the reviewers in that way, because of what he percieved as
a personal slight. But technically, he was within his rights to do it
- it was allowed by the reviewing rules of the conference. So I was
upset, but I moved on. There really wasn't anything I could do about
it. Looking back at it now, I still don't see anything that I could
have done. It was a case where all that I could do was accept it, and
submit the paper elsewhere. I ended up publishing it in a much less prestigious conference. But that's the way the cookie crumbles.
My second experience was much more serious, and much more
upsetting. I was at one of the major programming language conferences,
presenting a poster. One of my friends introduced me to a student from
UIUC, who was doing work that was related to mine. We spent a couple
of hours at the conference banquet talking about our work. About 9
months later, he published a paper proposing a change to his system
that was clearly based on some of the ideas I discussed with him. I
wrote him a polite note saying, roughly, "I'm not sure if you
remember, but back at XXX conference, we talked about our work, and I
told you that I was trying YYY idea. In your new paper, you use YYY,
but you didn't cite me. If it's not too much trouble, I'd really
appreciate it if you could add a citation to any other papers where
you discuss that."
His response was shocking. He basically said "You didn't publish it, so
you've got no proof that I got the idea from you, and who's going to believe a student from a dinky little school like UD over a student a UIUC?"
Looking back, what could I have done differently? Nothing
good. What I learned from that, and subsequent experiences is
don't talk about unpublished work in progress to people you don't
know. I hate the idea of telling people to do that. But I've seen
far too many examples of things like that happening - people hearing
about unpublished work-in-progress, and scooping it. I do still talk
to people about my work, but I keep it very vague, and don't talk
about new ideas unless I know and trust the person that I'm talking
to. Nothing get discussed until it's published in some form, so that
I have evidence that I did it first.
The third experience was the most serious by far, the most upsetting,
and the most intractable. After getting out of school and getting a job, I was discussing work with a coworker. The coworker took the ideas, and published them
under his own name.
In this case, I had the paper trail. I had the documents that had
been presented to the management at my lab more than a year before the
scum had heard about them. I took advantage of what was known as the
"open door" policy - we were allowed to go to any manager or executive
when something like that happened to raise a complaint. I went to one
of the upper-level managers who I had presented the work to, and
described what had happened, along with dates of meetings where I had
presented to him, to other people, and when I had described the work
to the scum. I expected some kind of serious action to be taken - here
was proof, hard evidence of outright intellectual theft. What
happened? Nothing. The scum was apparently well-known for
doing that (but no one had told me), and because he had excellent
political connections, no one dared to do anything to reprimand him. So the
whole matter was dropped.
I don't know what I could have done differently. I think I did the
right thing. It would have been wrong to refuse to discuss work in
progress with a coworker. I had a strong paper trail showing what I
had done when - I couldn't have kept more or better evidence that it was
my work. I raised the issue in the right way, with the right person. I
was calm, professional, and thorough. But I ultimately relied on the idea
that honesty could trump politics. That wasn't true. No amount of evidence
could have made a difference. The thief got away with it - as he had done on
numerous occasions in the past, and as I assume he continues to do today.
To provide one (sort of) encouraging case: I know people who used to work at another, now defunct, research lab. One of their lab directors had, for years, been
using the labs "publication permission" system to deny permission to publish to
his underlings, taking their papers, putting his name on them, and submitting them. He even won a fellowship an a major professional organization on the basis of his stolen work. He was, eventually, caught with the help of a corporate ombudsman, fired, and stripped of his fellowship. Alas, he'd done this for close to 20 years, and the lab only survived for about six months after he was fired.