General Question about Competition in Science?

Feb 19, 2013 23:22

I'm a young student in the early stages of graduate school and I have this idealized view that information should be freely exchanged and that science is more enjoyable and productive when this happens. But a week ago I was asked by a collaborator not to discuss our project with a friend I was meeting that day (another scientist working in the same ( Read more... )

Leave a comment

Comments 6

jaipur February 20 2013, 12:37:45 UTC
There are different attitudes toward that--I see fellow scientists as my friends, but some of my fellow scientists see other scientists only as "competitors", and don't talk freely about their latest ideas with anyone who has anything to do with their field. I think it definitely pays to pay attention to whom you are speaking--my graduate advisor once told me to keep an idea under my hat for while, and when I looked at him stunned he pointed out that Investigator X was visiting the lab that week, and Investigator X has a tendency to "forget where he heard things". I.e., my advisor had had experience with this researcher talking freely about ideas, and having the other guy run home and publish without giving any credit or additional discussion; so that ended that free exchange of information with that investigator. They were still friends, but my advisor just didn't give him the hot off the press ideas any more ( ... )

Reply


kailen February 20 2013, 20:50:12 UTC
I think this largely depends on your field and the people in it. The area I conducted research in was pretty small and most of the people in it had an advisor/student relationship at some point, so there was a relatively free exchange of ideas, and bouncing things off each other. However, we also had a specific 2 month field season and there was always a backlog of projects due to only being able to collect so many specimens in a field season, so it's not like someone could just run home from a conference and try out an idea right away. I don't know if this was true for some of the other people in my program (ecology, evolution, and behavior), but when you're dealing with animals and plants and have to collect or grow your specimens and have more experiment ideas than you have time or resources to conduct, this seems like less of an issue than in a chem program, for example. That said, I would still be judicious with who I tell and what I tell them. It's one thing to share skills and help each other out, but another to share ( ... )

Reply


mokele February 21 2013, 04:01:24 UTC
I'm lucky enough to be in a very open, scoop-free field with lots of sharing and collaboration. I think it comes from the fact that we're a) small, b) academically incestuous, c) there's still lots of "room" to stake out an area without competitors, d) many projects require uncommon skills/equipment or rare animals, and e) each experiment takes months or even years to do and analyze, so by the time it's "conference ready", anyone trying to duplicate it would be 3 years behind you.

Reply

kailen February 21 2013, 05:26:48 UTC
Yeah, the academic incestuousness really cuts down on the probability of scooping, IMO.

Reply


nervous_neuron February 21 2013, 12:15:44 UTC
Ah collaboration vs competition.

I suppose it does depend on the area. As cypherangel, industry may thrive on this competition, but what about academic, basic science? With the current system, it can be competitive as a good idea can net you a publication in a prestigious journal and a potential to attract more funding for your lab. So competition may increase the quality of the science. Sometimes idea sharing won't influence the competition - for example getting published in Nature, where there are many different fields being published, so a scoop would be difficult and fairly obvious. The problem is within similar fields. What's stupid is when two 'competitors' are working on similar projects and its a race to publish first when the competitors can become collaborators and combine their data into one paper. I've seen a paper where two different labs did the same experiment, and it added to the validity of the paper, because two independent labs produced the same results and it was all there on the one paper. Of course, there is the problem of ( ... )

Reply


british_spy February 22 2013, 06:31:32 UTC
Thanks so much everyone for your comments.
From reading all your responses, I feel like it is field dependent and something I'm going to have to experience myself through my academic career. My current environment is super collaborative and I guess reality isn't often like that.

Reply


Leave a comment

Up