A colleague of mine likes to share a saying passed down from a very famous neuroscientist who used to give this advice to all his trainees. “In science there are producers and there are consumers, don’t be a consumer!” What exactly does this mean? Well, consumers are those that pour over the literature endlessly, reading voraciously and going into excruciating detail on all papers they read (obviously a caricature, but you get the point). Producers are those who know the literature but prefer to experiment, trusting their own work rather than relying too heavily on what has been shown before.
Why do I bring this up? Three major reasons, creativity and interpretations of negative data (which are linked) and time management. But before we get to those, let’s begin with a bit of an ode to consumers.
In my humble opinion, there is a time to be a consumer. This is when you get into grad school and are trying to get a grasp of your field. At this point you need to get into the lit in detail. Learn about the relevant methods and what they are good for. Look carefully for the limitations in interpretative power that come with these methods. This is also the time to learn about the history of your soon to be area. What were the major milestones and advances and how did they occur. Finally, an excellent grasp of the literature can reveal what the major current problems are. Finding those gaps in knowledge are key to understanding how to plan out your Ph.D. studies. They are also key to grasping where and what to do for a postdoc, especially if you want to stay in the same area.
But, with all of this consuming comes a grave danger: talking yourself out of creativity. All too often this is linked to negative data in published manuscripts (at least that is my opinion). You know how it goes, so and so shows nicely that X is involved in some effect and a single experiment shows that Y “does not play a role”. I, for one, cannot stand that crap (especially when it creeps into the title of a manuscript — a pervasive problem in my area). Now, don’t get me wrong, there are papers out there which rigorously rule out a given hypothesis; however, there are all too many that do so with an overly cavalier attitude toward their single experiment. We all know how difficult it can be to demonstrate that a given effect occurs, how much weight do you really think should be given to an experiment which does not show an effect? Absence of evidence is not evidence of absence!
Another potential trap in being a consumer is getting bogged down in the minutiae of the literature. All too often this can lead to the illusion that small problems are actually big problems. Groups A and B cannot figure out why something occurs, we should go after the solution. If you are thinking on grand scales, this can be an appropriate approach. However, be careful. First off, this can place you into a bad situation in terms of funding. Two groups are already doing this, the problem will need to be a big one in order to convince a funding agency to throw more money at the issue. Second, this approach can distract you from the creativity of your own work. In my opinion it is always dangerous to think about trying to solve the problems of another research group. The appropriate response is what does their problem tell you about your work and how might a solution that you dream up lead you in a new direction for your interests?
Creativity is a funny thing and it can come from a variety of sources. While I have little question that it can come from a careful reading of the literature I think that experimentation is a much better source. There is no replacement for doing an experiment. All too often (at least in my experience) trainees allocate an inordinate amount of time to reading manuscripts and it subtracts from their experiment time. This is not a GoodThing. Planning meaningful experiments requires a good deal of time and there is no substitute for experiments that have interpretative power. Performing those experiments is also time consuming and those items combined can frequently add up to a full day of work. My advice, learn to fit your literature perusals into the downtime but don’t dedicate time to reading the lit that subtracts from experiment time. Produce, produce and produce.
So all of this makes it seem like I must pay next to no attention to the literature. Not true. I think it is the case that most producers that progress through scientific training eventually gain the ability to read papers by looking a figures. I generally don’t read an entire paper (unless I am reviewing). As a general rule I read abstracts and then glance through the figures. A rather funny thing has happened to me in this regard. In general, I find myself writing the paper in my mind based on what I see in the figures. From time to time I will read a discussion to see if my interpretation matches up with what the authors write. Usually it doesn’t, but I have a different perspective based on the biases of my interests. You see how this can work to your advantage, I have avoided the bias of the authors in preference of how I would interpret the data. I find that this keeps my creative juices flowing quite nicely.
Enough for now, time to get back to producing an R01 application.
Welcome all you folks from Genome-Technology.com! Apparently, from the comments, some are interpreting this post to mean that I have some sort of disdain for reading the lit or that I want hands and not brains in the lab. Neither of these are true.
1) As I stated above, at the start of training (or the start of a new project) one’s focus must be tilted toward consuming the lit. You’ve got to know what happened before and where the gaps in knowledge are to make an impact through your work. My concern is more with open mindedness and creativity rather than some sort of enforced ignorance of the literature. Perhaps I did not outline that clearly enough.
2) I have no interest in having sets of hands in the lab with no minds attached. I want to see creativity and originality in the thinking of my trainees and I want to help cultivate those properties into successful careers. If I wanted someone to just do what I say all the time I would build a robot.