Perhaps the most vexing challenge facing early-career scientists is how to demonstrate independence while working for someone else. And doing so is crucial. These days, an increasing number of institutions require compelling evidence that you're ready to function as an independent scientist before they're willing to hire you even for an entry-level faculty position. So how do you prepare for that transition? How do you demonstrate independence while still a postdoc?
"You need to get yourself to the point where you can write that proposal and ... seriously demonstrate that if it got funded, you'd be able to complete it," says Chris Armbruster, a social scientist in Berlin, who runs a network fostering postdoctoral independence, the Research Network 1989 . There are three ways to accomplish this: negotiate autonomy (and access to the requisite resources and space) with your postdoctoral research adviser, obtain a fellowship designed to facilitate your independence, or take advantage of one of the "junior investigator" programs intended for independent postdocs, which are growing in number, especially in Europe. "The earlier you can get experience [and] get some confidence, the more competitive you are going to be," says Sibby Anderson-Thompkins, director of the Office of Postdoctoral Affairs  at the University of North Carolina, Chapel Hill .
There is great variation in the degree of autonomy postdocs enjoy. According to the 2005 Sigma Xi Postdoc Survey , which looked at the experiences of 7600 postdoctoral scientists at 46 American research institutions, 38% of postdocs reported that they shared responsibility for planning new research projects with their advisers, 35% reported that their adviser controlled their projects mostly or completely, and another quarter said that they were mostly independent.
So how do you get yourself into the more autonomous fraction? Start when choosing your postdoc adviser and working environment. "One of the biggest challenges is postdocs finding themselves in labs or settings that are not the best fit," Anderson-Thompkins says. "It's critical that, as early as possible, they're asking questions about how decisions are made, about the expectations that the PI has [and] about the ability to negotiate." It's also important to make your expectations clear, she adds.
Some PIs encourage you to be autonomous, and others are rarely around to tell you what to do. But in most cases, "you're not going to walk into the lab and say, 'Can I have this space? Can I use this to set up my own experiment?' " Anderson-Thompkins says.
First, you have to build trust. So volunteer to look after a student, run a piece of equipment, or take charge of part of a research project. "Demonstrate to the lab leader that you are willing to carry part of the workload, take on responsibilities, and make sure that the results are good and delivered on time," Armbruster says. Then ask if you can do just the same thing but with more independence, he adds. "The [lab] head gets used to you becoming more independent. ... So if then, some time later, you go back and ask for even more independence, you're likely to get it."
Also, "it is very important that [postdocs] have good communication with their PI and really use ... the day-to-day contributions that they're making to revisit expectations, to revisit goals, to revisit space and time and resources in the lab," Anderson-Thompkins says. Progress on an experiment, a grant, or a paper--or getting a new project off the ground--can give you leverage. After each success, "come to the PI and say, 'Now that we have won this grant, would it be possible for me to now build in some time so that I can start to build my own research agenda?' "
"The next step would be to suggest some sort of alteration ... or different direction to the lines of research that are being pursued," Armbruster adds. Once you've demonstrated you can do all this, leverage your skills and reputation for reliability to "develop your own research plans."
Start writing grant proposals as soon as possible. See whether you can help your PI with a grant renewal or new application. As a postdoc, you may also be eligible for small research or travel grants from your institution, private foundations, or a government agency. Take advantage of opportunities to visit other institutes, as this could help you expand your network of independent scientists. This could eventually open the door for joint grant proposals on which you appear as co-PI, advises Peter Peters, principal investigator and dean of postdoctoral affairs at the Netherlands Cancer Institute  (NKI) in Amsterdam and cofounder of the Postdoc Career Development Initiative .
Aim for independence in your publications, too, appearing as sole, lead, or corresponding author as often as you can, Armbruster suggests. If you're mentoring students as a senior postdoc and write a paper together, you may even be able to claim the honored last-author spot; that is possible at some institutions, such as Peters's NKI.
In science, as in life, one of the keys to true independence is financial autonomy. Public bodies, private foundations, and research institutions offer funding opportunities for postdocs (see Independent Postdocs: Resources ).
Astronomers are typically granted more early independence than are scholars in, for example, the biomedical realm, and NASA  administers a cluster of fellowship programs that are rare, if not unique, in the amount of independence they grant fellows. The Hubble , Einstein , and Sagan Fellowships  are not mentored; the research plan is written by the fellow and merely "endorsed" by the host institution. The fellowships provide generous stipends--$60,500 for the first year, plus benefits--and a $16,000 allowance for travel and research expenses.
Individual awards like the U.S. National Institutes of Health's  (NIH's) Transition Career Development Award  (K22) offer postdocs a salary and research-related funding to help the transition to independence but still have a strong mentored component. Postdocs must develop both a research program and a career development plan with their sponsoring institution. "You're still building a research agenda within your faculty member's broader research, but by bringing in their own funding, they are able to carve out a niche and function more as colleagues," Anderson-Thompkins says.
NIH's new Pathway to Independence  (K99/R00) program takes a different approach. These awards offer early-career scientists a salary with research funds to continue in the labs of their PIs for a couple of years, followed by funding for their own research program during the first 3 years of their assistant professorship.
Also this week: "Reshuffling Graduate Training" 
Nobelist Roald Hoffmann believes that taking graduate students off grants and giving them fellowships would be good for U.S. science. But others say such a radical change isn't in the cards.
On the European side, one program that has been giving postdocs a fellowship to spend several years at the foreign institute of their choice is the European Commission's Marie Curie Actions . "Your proposal must always be in the interest of the place where you want to go," Armbruster says. But "the pure act of moving ... makes them more independent."
The European Research Council  recently launched its Starting Grants  to help postdocs of any nationality establish their own groups in a European country, with 5 years' funding. Many European governments and institutions have created equivalent national or ad hoc programs. The Research Institute of Molecular Pathology  in Vienna and NKI, for example, offer junior group leader positions that include institutional support, lab space, and funding sufficient to start a small, independent research group. Altogether, "this adds up to quite a significant number" of opportunities, Armbruster says.
"Of course, it can go wrong, and if it goes wrong, you're the only person that will be blamed," Armbruster adds. "But if you do it reasonably well, you will, in due course, receive probably more job offers that you can handle, because you're a rising star."
Despite the growing emphasis on early independence, it's important not to rush things: You need to spend your first postdoctoral years building your research skills and publication record. "If you start too early talking about independence, you lose momentum because you need your PI first to work together with you on a paper [on] which you are first author," Peters says.
Another risk is losing the support of your adviser as you claim more independence. Most PIs want you to do well, but they may perceive your increased independence as the loss of a valued worker--or, worse, as gaining a new competitor. "This is this double-bind situation which you need to navigate with the PI or laboratory head whereby you move on and ... move out, and they nevertheless must not feel threatened," Armbruster says. The solution: Talk and be explicit about what research projects you're free to pursue independence on, Anderson-Thompkins says.
These mostly new programs offer postdocs opportunities to establish their independence early, but they have also upped the ante in the competition for scarce faculty positions, especially in Europe. "A relatively clear hierarchy among postdocs" has emerged, Armbruster says, with ERC-type grants and junior PI positions at the top, soft money positions at the bottom, and fellowships and standard employment contracts in between.
Such increased competition means that many postdocs will have their hopes dashed early, Armbruster says. "But on the other hand, ... it is quite good that you're being forced to consider your options relatively early" so that you can get started sooner on a different career trajectory. "There is nothing sadder than the prevalent waste of talent in Europe, which is all these people who are allowed to go on with their academic careers until they are 40 only to find out that they'll never make it."