Well Served by ServiceJanuary 11, 2013
Careers in the Math Sciences
Service is always listed third in the academic career-development triad: research, teaching, and service. Though typically the last area emphasized and cultivated, whether in academic, government, or industrial careers, service is richly varied and rewarding. Service can create a sense of community that is far more diverse than the community you might experience as a specialized researcher. Too strong and too early an emphasis on service can rob you of the concentration and dedication required to advance your career. But many types of service are broadening and perspective-building. Too weak or too late an involvement in service can lead to narrowing and even burnout. No general article can prescribe the type, schedule, or level of service that is appropriate for a particular reader, but we hope this one might encourage junior colleagues to consider new pos-
sibilities--perhaps as a volunteer with SIAM.
Peer Review: Privilege and Responsibility
I often emerge from review panels marveling at the unanimously converged on rank of the proposals left on the board---and its modest correlation with the rank on the spreadsheet I had studiously prepared beforehand. What transpired in between was an almost magical process called peer review---a practice of obvious significance to the proposers, but also of great stretching for the reviewers. The dinner following a well-run panel is a high moment: Friends who were strangers a day before celebrate a group achievement requiring wisdom, candor, vulnerability, and prolonged concentration from each that has identified new talent, offered constructive feedback, and highlighted technical paths forward. Few events past a thesis defense occasion the free-wheeling, wide-ranging discussion of fresh ideas among experts who are committed to the importance of the occasion. Few better illustrate the superiority of the whole to the sum of the parts.
The contract between society and its research scientists and technologists has evolved during the post-World War II era to depend upon a strong engagement of scientists with society. This was not always so, but never before in history have so many---as much as 1.7% of the population---been scientists or engineers! Earlier, science was typically the realm of the well-to-do or the monastic---very rare members of the species who were free from concerns of making ends meet. Today, economically competitive societies need to grow researchers and developers more systematically, and we tax ourselves to accomplish this, spending as much as 2.5–3% of our GDP on R&D. Societies dispense much of this support through competitive processes.
In the U.S. and most G–20 countries, applied and computational mathematicians are expected to earn support for their research from granting agencies or from centralized discretionary corporate or laboratory programs. While most development and some research support is dispensed top-down according to short-term goals, the majority of basic research support is awarded by peer review. A high level of expertise and careful separation of interests are required for a peer review system to work well, and the human effort required by the system has to come from researchers themselves, whether as volunteer fractional-time reviewers of proposals or as paid full-time (at least for some term) managers of the process.
Members of the SIAM community are vital across the globe to agencies that support research in the mathematical sciences, staffing panels as described above or working solo to prioritize investments. Volunteers do the majority of the technical work accomplished by these organizations, facilitated by generally lean executive staffs. Voluntary participation is essential to the intellectual health of the profession and to the various specialties within it. Participation of many volunteers diversifies the perspectives and keeps the work relatively light (in the long view) for each one.
Agency service can take many forms, including reviews, panels, and site visits; resource-allocation boards (for computer time, beamline time, etc.); and advisory boards, report-writing panels, and workshops, wherein grantors set their strategies. What agencies are we talking about? In the U.S., many government agencies support some aspect of research and training in the mathematical sciences: DHS, DoD, DOE, DoEd, DOT, NASA, NIH, NIST, NOAA, NSA, NSF, and USDA, to name a dozen. Many states in the U.S. also conduct peer review of technology incubator proposals. Of course, some peer review takes place across national boundaries; this is especially important when too many investigators within one country are too interconnected in a particular campaign to provide impartial review.
If proposal review is "upstream," peer review surfaces again "downstream" of the research endeavor, in the publication of results. Service to journals includes editing and reviewing. SIAM journals are among the top ranked globally in applied and computational mathematics, and maintaining their standards and scope would be impossible without the approximately 10,000 reviews of the approximately 4000 papers submitted by the community each year.
Peer review is also required in running conferences. SIAM conferences are intentionally rather light on selectivity compared to most, favoring networking and exchange over conferring prestige. Still, a lot of expert judgment is required in setting the themes and populating the podiums for each of the dozen or so conferences SIAM holds in a year.
Reap What You Sow
Peer review provides major indirect rewards. Your field prospers and polices itself. Strong participation as a volunteer engenders a favorable reputation for you within budget offices and editorial offices, and with other stakeholders. Your personal reputation as a contributor grows with your participation in individual and panel reviews. Influential colleagues come to know you and your judgment and can subsequently review you well in turn. Your technical knowledge deepens and broadens, and your "street smarts" in championing, proposing, and executing future research grow. You also become more valuable in the eyes of enlightened employers.
Lobbying for the Health of the Discipline
Another form of service to agencies and professional societies is lobbying---where the societies meet the agencies. SIAM has an extremely well-run Committee on Science Policy, made up of about two-dozen SIAM researchers. Twice yearly, committee members converge on Washington for two days, hosted by our lobbying organization, Lewis-Burke.
On the first day, the CSP hears from invited guests, usually heads of agency mathematics divisions. That afternoon, or overnight, CSP members prepare brief position papers on priorities for support and advances from previous applied mathematics research. On the second day, committee members pair up and visit legislative staff---often those of their own Representatives in Congress, or of key legislators for science funding. They follow up with letters to agency representatives and legislators, and with news articles for SIAM's membership.
Legislative staffers are often recent college grads---ambitious and eager, but generally having evaded math en route to their current jobs. Those who stay don't learn more mathematics, except from people like us. My congressman's senior staffer once greeted Linda Petzold and me by putting his feet up on his desk and informing us that earlier in the week lobbyists for medical research on various organs (heart, kidney, eye . . .) had stopped by to plead their cases; he challenged us to match them in importance. We replied by describing mathematical modeling and simulation research that was helping to enhance understanding of some of the same organs. This was not the kind of mathematics he had been shown in school. When it comes to community service, surprising amounts of fun and low-hanging fruit await!
Members of the SIAM Council decide on the sunrise and sunset of activity groups, conferences, journals, contests, prizes, and all kinds of initiatives, collaborations, and statements. SIAM members also join their counterparts from other professional societies in populating advisory committees to granting agencies. I remember the first time it dawned on me that I was writing text that might go into a future agency solicitation to which I (and hundreds of others) might apply. What a sobering fact to sleep on! Our systems for support of science have many flaws, but how wonderful is this feedback loop amidst all of the other forces and how important it is to maintain!
Are agencies and professional societies not yet asking your advice because you are a little ahead of their domain? It is natural that new paradigms arise bottom-up, rather than top-down. Internet blogs, wikis, and open-source sites for distribution of codes, data sets, and technical tools are previously nonexistent forms of nonrefereed or lightly refereed service to the community that can potentially rapidly accelerate the transformation of a field. Only later are such contributions codified into supported programs and standards.
Following frontiersman Davy Crockett, "Make sure you're right---then go ahead."
Control your Expectations
Contributing technical expertise allows you to stay current in your own and nearby fields, to swell your Rolodex, to impress your colleagues and grant monitors, and to truly influence the direction of a field. However, some career cautions should be borne in mind vis-à-vis agency and society service. Don't expect immediate gratitude---a little here, a little there, from different stakeholders, is hard to notice at first. Those in charge come and go. But you're in the profession for the long haul, and thanks come over time. Don't throw yourself into too much volunteer work too soon. It can be addictive! Research and teaching are more important for immediate advancement. There is time for agency and society interaction when your career is solidly established. But when asked, everyone who benefits from the system is expected to flex to participate.
What is a reasonable commitment for community service? If you have a grant from an agency, expect to be asked to serve on a panel or two per year, or to combine a panel and a batch of mail reviews. If your work is published in a journal, expect to be asked to review up to a few papers per year for that journal. Such service is not quid pro quo, however; follow your instincts, accepting invitations from those you respect, in areas to which you can contribute the most specialization. If you have not yet been asked to serve, do not panic. It is fine to make your willingness to serve known, but don't target too finely---you could appear to have a private agenda. Ultimately, invitations will cascade. Don't be too picky at first, but you will eventually have to prioritize.
There are many categories of service to consider, from the broadest forms of service to the community described above, to your employer, and, finally, to your closest circle.
Serving your Employer
Service to an employer can take a dizzying variety of forms in academia, perhaps fewer in industrial and national laboratories, where the diversity of the employers' activities can be somewhat less. To begin, universities, corporations, and laboratories may have their own internal granting programs, for which the advice is similar to that given above.
Academic service is usefully categorized as university-wide or program-based, though the two have some activities in common. Both include committees for student recruiting, graduate admissions, faculty recruiting, faculty promotion and tenure, library or facility acquisition, curriculum review, qualifying examinations, thesis evaluation, student chapter advising, conflicts of interest, intellectual property, university development, student prizes, faculty prizes, alumni relations, publicity, international programs. Many of the same considerations regarding career cost, career benefit, and priority described for community service apply here as well, except that the community is much smaller and typically has a much longer memory.
Serving your Circle
The smallest circle of service is within a department or group of students and postdocs. Here, the dominant service is mentoring and career counseling. The interactions are more intense and typically lifelong. The needs and the unique roles required to fulfill them are more obvious than for community and institutional service. The most important items on the career checklist are balance and priority. Your students and junior colleagues are more uniquely dependent on your time than, say, SIAM or NSF, but rewards from service come from all levels, from all sectors, and at diverse timescales. Practice on activities with low stakes, by volunteering to judge undergraduate poster contests or science fairs, or by organizing single-day campus-wide or regional technical symposia.
Diversify for experience. Specialize for effectiveness. But don't miss out!
David Keyes has contributed to more than a dozen U.S. agency reports on computational mathematics or computational science and engineering and has served on two NSF advisory committees. He is the inaugural dean of the Division of Computer, Electrical and Mathematical Sciences and Engineering at the King Abdullah University of Science and Technology. Keyes is the 2011 recipient of SIAM's Prize for Distinguished Service to the Profession.
Sue Minkoff (email@example.com), of the University of Texas at Dallas, is the editor of the Careers in the Math Sciences column.