It was an awkward conference.

The DNA computing conference in Holland in 2001 sat at the crossroads of two barely related disciplines: computer science and molecular biology. The computer scientists sought to apply principles of molecular biology to new computing platforms, and the molecular biologists wanted to make computers that use DNA for information storage.

In other words, everyone's work relied on a field in which they themselves had no expertise.

While it was encouraging to witness the collaborations that sprung up between disciplines, the gulf in understanding made for some uncomfortable moments. In one of those moments, a Romanian computer scientist described his recently invented computational model with a structure that somewhat mimicked that of biological cells. Somewhat. As he described his new "membrane computing" idea, and terms like "Boolean satisfiability" flew fast and furious, the biologists in the room lost interest—until he expressed a desire to replicate his model using actual cells.

The problem was, his computational model was just a model, and the "membranes" were mere theoretical barriers that divide objects from other objects. A biologist pointed out that some of his required processes—such as selectively dissolving just one of several nested membranes inside a cell—were impossible to do. In reply, the computer scientist frowned and said dismissively, "I am not a biologist."

As scientists, nothing pleases us more than collaborating with researchers in other disciplines. It makes us feel like part of a community that is cooperating to solve the world's problems. It reinforces our sense of self-importance; after all, we've managed to convince a smart person that ours is a star worthy of wagon-hitching. It gets our names into wacko journals we've never heard of. It lowers our Erdös number.

At the same time, we risk treating it as an outsourcing of responsibility. There's no denying that another purpose of collaboration—a part of its value—is in delegating portions of our research to someone else, someone whose science, to us, is indistinguishable from magic. Part of our project then becomes something for the other folks, those mystical collaborators in their incomprehensible discipline, to figure out.


CREDIT: Hal Mayforth
Click the image to enlarge

In grad school, I used to work on designing new malaria drugs. (Now I work on a malaria vaccine.) Sometimes people would ask me how easily my new drugs could be synthesized, and I always said that was something for the organic chemists to figure out. Or they'd ask how much the drug would cost (if it worked, which it didn't), and I'd say it was something for marketers to figure out. Or they'd ask when I'm graduating, and I'd say it was something for my principal investigator to figure out, but that was a different problem.

How did we synthesize the drugs? We'd e-mail our collaborators in a synthetic chemistry lab in Japan. A few months later, we'd receive a package from Japan. In my mind, it was a two-step synthesis.

* * *

It is a truth universally acknowledged, that a scientist in possession of a grant must be in want of a collaborator.

But such collaborations often come with special challenges—beyond, of course, figuring out which of the grad students in your lab will mate with the grad students in your collaborator's lab, which most granting agencies require. So if you choose to work with a scientist in a different discipline, please bear in mind the following guidelines:

  1. The unfamiliarity of your collaborators' work is what makes their science intriguing—but it can also make you regard them as mere service providers. In other words, though your collaborators are valued colleagues, the idiots just don't seem to be able to understand that you needed those results on Thursday. What are they doing over there in that other department, screwing around all day? Chill out and remember that, in their minds, you're the weird collaborator with the odd priorities and uninteresting interest.
  2. When collaborators come to work in your lab, they may be as confused and reckless as undergrads. When I was in college, I collaborated with a physics lab to build a device on which I could perform biological experiments. A postdoc in that lab mentioned that I'd need certain custom-made brass pieces—and I mistakenly assumed that he meant he intended to make them for me. After all, I'm a molecular biologist: I can breed bacteria, I can purify proteins, and I can clone one hell of a fruit fly, but I wouldn't know the first thing about machining brass. When the brass pieces failed to materialize, I sent him a friendly e-mail asking when I could expect them, and he replied angrily that he was not there to do for me "things that a 3-year-old can do."

    This is how miscommunications between collaborators arise: One field's boilerplate is another's alien gibberish. (I still doubt, though, that physicists let 3-year-olds near their metalworking equipment. My daughter will be 3 in a few months, and just yesterday she asked if I'm married to Grandpa Jeff.)
  3. Sometimes your fresh perspective can help your collaborators solve a problem in which a solution had been practically staring them in the face. More often, though, a fresh perspective will lead you to mistakenly think that you've solved a problem staring your collaborators in the face:

COLLABORATOR:  We just can't get the chimps to speed skate.
YOU:  Have you tried making them wear chimp-sized skates?
COLLABORATOR:  Of course we have. In fact, we've tried hundreds of types over the past year—
YOU:  Yup, chimp-sized skates. That's the ticket.
COLLABORATOR:  See, it's more complicated than that. The curvature of a chimp's plantar fascia—
YOU:  It's a skate, but smaller. Like for a chimp, you know?
COLLABORATOR:  My Ph.D. is actually in chimp skates, so I'm well aware of the challenges—
YOU:  I'm going to patent that idea. Chimp skates! Now I'm off to solve world hunger. I think the solution is … wait for it … food.

  1. Be aware that differences arise not just between scientists in different fields but also between scientists at different types of institutions. When an academic scientist collaborates with an industry scientist, for example, each should know that the former faces pressures of grant deadlines, while the latter faces pressures of whether to gild or platinum-coat the valet key for the Lexus.
  2. Don't forget to dazzle your collaborator by making your work sound complex. "I hyper-configured my trans-gronkulator," you can say, "but you know how that is." This comes in especially handy when deciding who gets how much grant money. Those trans-gronkulators sure are expensive, and goodness knows they're not going to hyper-configure themselves!
  3. Most importantly, try to understand your collaborator's discipline as well as possible. One of the great advantages to collaboration is the opportunity to learn something new, and you'll sound less oblivious when someone asks you a question at a seminar and you don't need to say, "Ah geez, I don't know. You'll have to ask my collaborator because I'm a big tool who cares more about appearances than about personal enrichment and discovery."

* * *

A few weeks ago, at a cryobiology conference, I witnessed firsthand how a lack of interdisciplinary communication can sink science. Cryobiologists are always looking for good ways to freeze things and keep them viable. At this conference, a physicist presented his theories on how to do that—but his theories were all couched in the language of physics, so no one listened.

Even worse, he knew no one was listening. It turned out he had submitted the exact same abstract for the exact same conference 24 years ago, and no one listened then, either. His present resubmission of the 1989 abstract, he told me—and his second identical talk (albeit with more advanced PowerPoint)—were meant as passive-aggressive reminders to the community that he had solved their cryobiology problems during the first Bush Administration, but because he was a different kind of scientist, no one had listened.

Did the physics equations he described actually solve our cryobiological problems?  I don't know. I wasn't listening.

That's the biggest problem with interdisciplinary science: It can make us complacent. If we just work on our part, we think, others will solve the problems we find difficult, so we look the other way when the equations feel foreign.

That rarely works, though, because there's no such thing as a "biology problem" or a "physics problem." The word "interdisciplinary" may appear to let us off the hook for solving whole problems, but that appearance is deceptive because science itself is innately interdisciplinary. The spectrophotometer in my lab—used primarily by biologists—was invented by an optician, improved by chemists, manufactured by mechanical and electrical engineers, and distributed by a sales rep who gave us free pens. In a sense, we're all collaborating, even if we don't know it.

Which is wonderful—until the time comes to determine authorship.

Adam Ruben, Ph.D., is a practicing scientist and the author of Surviving Your Stupid, Stupid Decision to Go to Grad School.

10.1126/science.caredit.a1300179