My word

In defence of passionAndrew Murray

Graduate students and post-docsoften ask “What do I need to do tosucceed?” One popular answer is toboost your career by networkingfeverishly, finding powerful mentors,learning to pinpoint hot areas, andpublishing in sexy journals. Thisapproach has two problems: it makessucceeding in biology sound likelearning to sell used cars, and it reliesheavily on prognostication, a skill thateven politicians have yet to master.

I want to discuss a more old-fashioned solution: do work that youfeel excited and passionate about.The first step is to find a problemthat is hard to solve, located in anunpopular area, and interestingenough to push you out of bed everymorning. If your problem is hard,you’ll have to endure more failure inyour quest. Because failure hurts,most of us learn from it faster than wedo from success. In addition,succeeding after a long struggle givesyou confidence that you can tackleother difficult problems. Working inan unpopular area gives you the timeto fail and learn without worryingabout an enormous factory labovertaking you in the night. Finally,your problem must generate theexcitement and passion that will keepyou going through thick and thin.

How do you find the perfectproblem? Start by consulting lessfashionable sources. Try leafingthrough solid but unflashy journalsfor papers that identify new puzzles,and talking to older scientists aboutinteresting problems that have beenwaiting for fresh young blood.Information from more than acentury of interesting experimentson an enormous diversity of creaturesis lodged in the cerebral crinkles of

people whose careers began beforethe dominance of molecular biologyand model organisms.

If you’re interested in cellbiology, consider reading E.B.Wilson’s classic book The Cell inDevelopment and Heredity. Anyquestion that Wilson asked in 1925and is still unanswered is likely toidentify an interesting problem. Forexample, what is a centrosome madeof, how does it duplicate, and howdoes it position itself within a cell?

A good problem should be hardto solve and located in anunpopular area

Another way of finding goodproblems is to consult yourself. Listthe questions that you think we haveto answer in order to reach anintegrated understanding of howcells, organisms, and populationswork. I bet half the questions onyour list are unanswered in principle,let alone in detail. For example,although we’ve made great strides inunderstanding how transitions in thecell cycle are regulated, we knownext to nothing about the factors thatdetermine how cells control the rateat which their mass increases (thereal meaning of cell growth).

Although your problem should behard, it has to be soluble. There arethree ways that insoluble problemsbecome soluble. The first is by thesort of mental leap that occurs whensomeone rephrases a question in away that allows existing techniquesand resources to answer it. The moreyou know about your field andbiology in general, the better yourchance of asking your favoritequestion in a new way. The second isthrough the use of new techniques.The more techniques you can useand understand in detail, the morelikely you are to invent new ones.The third way in which intractablepuzzles often seem to be solved is

through a chance observation,although most of these discoveries areless lucky than they seem. Near theend of my graduate work, I failed topursue a strange finding that wreckeda measurement that I was trying tomake, choosing simply to get the dataI badly needed by taking a differentapproach. Later, Dan Gottschling’sbetter-prepared and more flexiblemind took a similar observation andmade it into the beautiful story ofhow telomeres silence nearby genes.

Once you’ve found a problem youlove, you need to find a boss whoshares your curiosity or is benignenough to welcome a well-meaningeccentric. You’ll hone yourcommunication skills by convincing amentor that it would be fascinating toknow why echinoderms have five-foldradial symmetry. This exercise is alsovaluable preparation for giving talksto similarly unenlightened audiencesafter you’ve solved the puzzle.

Finally, you can embark on yourproject. Of course, you need to dowell-designed, meticulousexperiments, be determined in theface of failure, and generous to yourpredecessors when you succeed. Buteven more importantly, you need towork openly and get as muchcriticism and advice from otherscientists as you can. The beauty ofworking on an unpopular problem isthat you won’t need to worry aboutsome dark prince or princessskewering you with your ownunpublished results, and becauseyou’ve picked a hard problem you’llbe forced to seek advice andconsolation on a regular basis.

I spent my graduate careerunsuccessfully trying to discover whychromosomes care how long they are,published a paper last year that haddata from my thesis in it, and am stilltrying to persuade my poor studentsand anyone else who is interested tohelp figure out the answer.

Address: Departments of Physiology andBiophysics, University of California, SanFrancisco, California, USA.

R267

Magazine