Wednesday, August 21, 2013

Submarine science, or doing projects on the sly

"Quiet! Management thinks we're doing that stupid Project A!"
(Credit: metroactive)
I was both surprised and not surprised to read this comment by Validated Target in a recent In The Pipeline thread on innovation:
..."loose cannons and mavericks" does not equal non-conformist innovators. The history of Drug Disco is filled with examples of the latter. 
1. Cimetidine started out as a two-person project (chemist and biologist) who were inspired by Black's work on H2R antagonists. As it was told to me by those who were there, they were repeatedly told to stop work on the (back burner) project. As other SKF projects withered and died, pre-cimetidine results FINALLY merited further examination, the project was green lighted. Cimetidine / Tagamet saved SKF. The two guys who took the heat were non-conforming innovators. 
2. I've heard that Prozac was a 2-man (chem + bio) project for a while at Lilly until it picked up speed. I think it was the biologist who, from the literature or a meeting, first wanted to go after an SRI. It was not some corporate committee that set a mandate or a glorified thought leader. It was a scientist with an idea and it worked and everybody took a pill and chilled out (and cashed in!). 
3. I can't think of others cases at the moment. (IBM threatened to fire Bednorz and Mueller if they didn't stop work on the unofficial yet Nobel worthy superconductor project.) 
Later in the thread, London Chemist has another one:
the Germans working on gabapentin were repeatedly told to stop working on it by W-L's* US management.

I think it is terribly interesting how often scientists will say "Oh, yes, sir, I've been working all day on Project A" and in a hidden hood somewhere, they're actually running Project B experiments, too. I've done it before, and I don't doubt that I'll do it again sometime in the future. In the classic New York Times Magazine article "Lethal Chemistry At Harvard", the chemistry students of Harvard actually had a name for it:
They spoke of ''submarine science,'' in which graduate students feel obliged to spend their days doing experiments suggested by their adviser -- even if they are certain they won't work -- and wait until after hours to try what they wanted in the first place.
I wonder if advisers/managers ever realize that, chances are, their students/employees are working on something on the side? I wonder if they ever try to make sure to quash that behavior ("Johnson, I want 100% effort on Project A! 6 reactions a day!") or if they just don't care, just as long as Project A is meeting its milestones.

(I'll tell ya one thing -- if I'm ever the boss (that'll be the day), I just won't care.) 


  1. I have heard of a PI who told his students to do these types of experiments on Friday afternoons. Can't remember who, though.

  2. Other submarine projects that paid off handsomely include Zoloft (Pfizer) and Topiramate (McNeil, now J&J) . More on Topiramate here:

  3. Depends on the manager: at an oil company I worked at, it was okay as long as official project milestones were met. Got a few patents that way.

  4. I thought Heck was doing his work on Friday afternoons at his industrial job. I remember something about the company allowing 10-20% effort on your personal projects (?) at the time.

  5. My graduate advisor (in organic chemistry) actually encouraged us to do submarine chemistry!

  6. If there's something I've really wanted to try, I've always found a way to try it. Many companies actively encourage "bootlegging."

  7. Open source software is quite useful for supporting submarine projects and have been the norm everywhere I've worked (but maybe that's just me).

  8. The AIM-9 ('Sidewinder') heat-seeking air-to-air missile was developed this way at the China Lake Naval Weapons Centre. What is less often appreciated is the sheer *efficiency* of submarine projects. The following quotation is from Ron Westrum's book *Sidewinder: Creative Missile Development at China Lake (Annapolis, MD: Naval Institute Press, 1999), pp. 12--13 (notes in original).

    'While they were learning how to run an R&D operation, Rabinow and McLean were evolving a style that later became identified with the Lockheed "Skunk Works," but actually such informal operation was typical of OSRD laboratories in World War II [24]. Rabinow, who has a gift for turning a phrase, over the years evolved a set of laws:

    '[Law] #13 I think, says that everything you do illegally, you do efficiently. This, of course, is perfectly obvious. For one thing, you do not write at all because writing on an illegal project is suicide. For another thing, you work with whatever equipment you have on hand, and of course, you do everything on your lunch hour, which starts at 8:00 in the morning and finishes at 5:00 in the evening. Another thing, when it doesn't work well and it is illegal, you drop it very quickly and kill the project. When it is legal, you carry it on to doomsday, hoping that someone else will carry it on, so that when it finally fails you won't be blamed. If an illegal project does succeed, you will be a hero, but if it fails you would like no one to know about it, so you bury it quickly. Illegal projects are very, very efficient from many points of view. We were allowed to do much of this.[25]' "