I’ve mentioned before that running a lab is like running a ship. For the most part it’s like running a sinking ship which you are frantically trying to keep afloat, but the analogy extends further. And maybe here’s where the analogy can diverge depending on how you run your lab. In many large ships you have a crew of specialists. One can navigate, another steers, others fixes sails, others keep the cannons in shape, a bunch of lackeys will swab the decks and coil the ropes, someone cooks, that sort of thing. The captain’s job is to set the course and make sure everyone gets along and does their job, the goal being to get to where they need to get, and deal with any trouble along the way.
This can be an efficient way to run a lab. You can have a bunch of specialists all working on different aspects of a core project. The molecular biologists are running the DNA chips and generating the transgenic animals, the imagers are working the confocal microscopes, the electrophysiologists are cranking along in their rigs, the engineers are devising new devices, the spies are infiltrating competitor’s labs, the programmers are modeling your data and the nuclear physicist is plotting the global takeover. This division of labor ensures a steady throughput of papers and allows the lab to efficiently take new directions and tackle new projects. So from an efficiency perspective, this is the way to go.
What is less clear is what the contributions of the first authors on these papers is. If everyone is a specialist, what makes someone first author? Also, to me the training potential of these labs seems questionable. How can you be expected to run your own lab if you spent your postdoc doing the same thing over and over? Probably someone who is smart will pick up things along the way and by observation learns how shit gets done. But many people will be headed to nowhere, despite their first author paper on some fancy glamor journal.
So there’s an alternate model, a model where people are assigned projects and they are then responsible for everything in the paper. In this model, some people are better at some things than others, but everyone then gets a chance to learn and do multiple techniques. So if your project requires imaging, genetics and behavioral testing, then you learn to do all these things by talking to the folks in the lab that are experts at these techniques. And if you don’t want to learn a new technique, then you are responsible for arranging some sort of internal collaboration. In this way everyone mentors each other and learns about each other’s projects, as well as learns how to put the big picture together. While this may cause papers to take longer to finish, I think that over the period of 3 years or so the total output of the lab is comparable to the previous model. It also ensures that say, the expert in time travel leaves your lab, then you are not left high and dry until someone else figures out how to run the Neutrino Hyperaccelerator so you can go back in time and submit before your competitor does.
I’m not sure which model is better overall. I did my postdoc in a lab that ran using the second model and that is mostly how I run my lab. I try to get everyone to learn and use multiple techniques and to pilot their own projects. But I see the merit of both approaches. Any thoughts?