Since I started doing research 10 years ago or so, this is a question that I asked myself or talked about around me many times. It also happens to be a common question for student at the end of their PhD, faced with career choices. This recent piece by Philip Ball was an excellent reminder of the unique role of the instrument in the scientific process. There are two kind of researchers out there: some who touch upon many techniques (the majority), driven by a problem to solve, while others (the minority) are specialized in one technique only.

When doing a PhD, there is a good chance that one technique in particular will be particularly useful or relevant for the topic you are investigating. In my case, I was investigating the low-temperature surface degradation of zirconia-containing ceramic materials during my PhD. Around the end of my first year, an atomic force microscope was bought and setup in the lab.

Following the advice of my supervisor, I started playing with it, since nobody else was using it, and it turned out to be a killer technique for the phenomenon we were looking at. I spent the next two years doing almost exclusively AFM, and since we were the first ones to use it to look at the surface changes induced by the degradation, we learned a lot on the phenomenon and I ended up with a dozen peer-reviewed papers by the end of my PhD. At this point, I started looking for a postdoc position and considered applying for a CNRS or university positions, and was given what I believed was a solid advice: do not specialize in one technique. Long story short: by a series of coincidences, I get interested in freezing and eventually secured a CNRS position a few years later. I have not touched or even approached an AFM since the end of my PhD.

The blog post by Philip Ball on scientific instruments brought the original question back to my mind: it is a good idea to specialize in one technique ? Does it make it easier to get grants, papers, and a stable position ? Is it rewarding ?

Since the majority of researchers are not specialized in one technique, it is easier to see the cons, first. The first and obvious constraint is that if you want to specialize in whatever you chose, you have to be the absolute best one with it. Maybe not worldwide, but at least at the local scale or preferentially larger. Once you mastered the technique and became a reference with it (congratulations), which should take you a couple of years, depending on the technique, the bad news is that you will have to upgrade your instrument, and do so sooner rather than later. Techniques are changing rapidly (think confocal microscopy, for example). Your continuous success with it usually means that you must be up-to-date. And doing so can be very costly, in particular if you are into transmission electron microscopy, for instance. Congratulations: you just entered a never-ending race. Having an out-of-date instrument can be a killer for everything you did before and you are at the risk of becoming obsolete rapidly.

The other major con I see (and I guess this is the main view) is the difficulty of developing a research line on a medium or long-term. Looking at a particular phenomenon almost always requires a range of techniques to give you a complete picture. Each technique provides a biased look, with its one strength and limitations. It’s just one part of the story you’re trying to make sense of.

There are nevertheless many advantages of being an expert with a particular instrument. Bearing in mind what I mentioned before (about the upgrades), if you have the best instrument around, you have a unique opportunity to push the limits of the knowledge rapidly, and can do so in a number of fields or topics. Microscopes with higher magnification and/or better resolution (space or time) will let you see features never seen before. There is also an intense satisfaction of pushing the limits of the instrument you have in your hand, something academics seems to really enjoy.

Once you established yourself as an expert, you won’t have to look for samples or topics anymore. People will naturally come to you will a range of problems to solve. You are certainly saving some time here, although you are at the risk of getting lost in a myriad of samples and topics. This also means that you don’t need to look for fundings anymore (except to upgrade your instrument, see previous point. It can be even worse, actually, as getting grants for equipment only can be very tough). It becomes really easy to jump in proposals, if you provide a unique and essential contribution, without being in charge.

The main advantage I see is that you get to work on many topics, which can be both intellectually satisfying and rewarding. Since you have a dispassionate look at the problem underlying the sample you’re probing, you also get a different perspective on this particular problem.

Pushing the instrument to its limits will also make you a very valuable customer to the people who sold you the piece of equipment, since you will provide a very unique and thorough feedback. If you enter into a trust relationship with them, you can also be the first one to try new upgrades (detector, column, etc.) before they come to the market. Which will again bring you in the position of being the first to unveil cool new features/phenomenon. If you’re building your own piece of equipment, which is something physicists seem to be passionate about, you are learning a lot in the process. I recently discussed this with Jim Smay, who’s been building and selling robot-casting (3D printing) machines for many years now. He’s not getting bored of it, and told me he’s always learning something new every time he’s building a new one.

At the end of the day, I guess it comes down to personal preferences. Most people will choose a problem and not a technique, but becoming an expert in one can be a very rewarding choice and the beginning of a fruitful career. I don’t have any evidences of which one is best on the long-term to secure a permanent positions and grants and keep doing cool science. I like to touch upon many topics, and freezing is certainly greatly satisfying from this point of view. I prefer the exploration of new topics and ideas to the tedious, deep investigations on well-identified problems. We need both kind of people of course, but this will probably be the topic of a different blog post.

A few days ago, I asked on twitter why there are so few french academics on twitter. I started tweeting about a year ago or so, for reasons I will discuss below, and a year later, my feeling is that the relative number of french academics on twitter is clearly less than in the US or UK. Correct me if I am wrong. Whenever I discuss twitter with colleagues, I get this social-media-is-for-teenagers kind of look and comments. I am truly amazed how underestimated twitter is as a tool and a mainstream communication platform for science today.

Twitter adoption is just the tip of the iceberg, I believe. When it comes to communicating science, french academics have long had a peculiar attitude. I have been fortunate to spend two years in one of the best place to do science in this world, in the bay area around San Francisco, and learned a lot about communicating science while I was there. US academics take communication very seriously. Ok, a bit too much, sometime. Overselling your results is not a good communication strategy. At LBL, there was an entire service dedicated to communicating the science of the lab (ok, it is a 3 000 persons lab, but still). They could design your posters and talks if you wanted to. A professional photograph was working full-time for the lab. Coaches were decorated with giant photographs of science leaders of the lab.

In France, people have the opposite attitude and (mostly) do not seem to see the point. « If the science is good, people will notice my work », seemed to be a common belief for a long time (and still is for many, unfortunately). The state of many lab websites, for instance, is a good example of the current interest (or rather, its absence) of french scientists for communication. Things are changing, for many reasons. Because taxpayers are challenging the use of their tax money and want to see the outcome of fundamental research. Because people are curious. Because funders require it. Because a new generation is coming, too. A generation that is born with internet, its tools, and its social codes.

Having shared my time between the two countries, I am balanced between both attitudes. The absolute first requirement if you want to communicate is that the science has to be good, of course. Communicating on bad science will be disastrous. I have seen too many talks at american conferences where the speaker was obviously an expert in communication skills but the science presented was absolutely ugly or completely pointless. Worse, and much more common (at least in France): some stunning science badly served by catastrophic communication skills. Slides with bullet points and 200 words, crappy figures with terrible color choice, you get the idea. I, for one, have become a bit obsessed when it comes to figures, posters, and slides design, largely inspired by my former mentors and Edward Tufte principles. My latest student can testify about it (and somehow became even more extremist than me) but he is becoming good with Illustrator. Good communication skills are difficult to acquire, it is a long journey, but the rewards can be awesome. It is definitely an interesting exercise to explain your work in accessible terms (like when you have to do it for you family during Xmas or Thanksgiving dinner) and also happens to be a good training for your next grant proposal. My ERC interview started by a 5 minutes pitch, and you can bet I practiced it before (in the train to Brussels, actually).

Twitter is in a different category. Using twitter for self-promotion might not be the best idea, at least if that is the only use you are planning with it. Many reasons brought me to twitter: I am working in a small lab (10 to 15 people, depending on the year), in a small town (no university around). Even though I travel a lot (almost every week), it does not make up for this modest size, below the critical mass where you get random interactions with others. I therefore use it more and more for the informal exchanges you can have with fellow scientists, but from all over the world in this case. Which can happen either in french or english — the language is irrelevant here. The second interesting use is to keep the generic public informed about our work and the way we do it, on a daily basis. #overlyhonestmethods was an interesting one, for instance, and brought the scientists out of their ivory tower for many people. I occasionally use it for self-promotion, when a new paper is coming out, of course, but I am not sure yet if it is the best way to reach fellows and people. And shameless self-promotion is badly perceived or at best blatantly ignored on twitter. Twitter has also become one of my main information channels for generic science news of different domains for which I have a passing interest (dinosaurs !) (I still keep track of new papers through RSS feeds). You can also get interesting behind-the-scenes peeks from the journals as many editors are on twitter (Andrea Taroni (Nature Materials), Pep Pamies (Nature Materials), Stuart Cantrill (Nature Chemistry), Robert Garisto (PRL), Joerg Heber (Nature Comunications), providing useful advices. My feeling is that for most, in France, all these aspects are not perceived yet. Twitter still appears as a distraction for youngsters and is not taken seriously.

Unexpected things can happen if you start to tweet. After exchanging with @BenjAbecassis for a few months, he came to the lab to chat and give a seminar (hint: in real life, he has a beard, do not trust his profile pic). Following the Science paper on open access journal accepting a fake article (referred to as the « Bannon sting »), there were many papers in the daily press about it, and most of them (unfortunately) missing the point. I shouted to one of them (Rue89) over twitter, and they invited me to answer to their paper, which I did here. I kept track of the comments for a few days, both on the journal website and over twitter, and had interesting exchanges with regular (non-scientist) people, who were truly interested in understanding the issues at stake. I recently discovered that @fxcoudert, with whom I exchange a lot on twitter, is going to supervise a PhD student that is shared between his lab and mine. You get the idea. Try, and see what happens next.

Things are improving faster at an institutional level. Not everything goes through press release anymore. The CNRS for instance, makes great efforts to communicate to the public, through both its journal (available both in french and english. I am a big fan of it), its redesigned website, and yes, twitter (Facebook also I think, but I am not using FB). Most (if not all ?) institutes (chemistry,  physics, nuclear physics and particles, etc.) are now tweeting news about the institutes and their activities and stories. In the « News from the lab » page (here for chemistry), papers of interest are briefly discussed in fairly generic terms. I grew to really appreciate it over the years, as it gives a different perspective than the one you get when you read directly a paper in your field. I would be really curious to know the audience of the site (150k daily visitors according to Wolfram), as well as the profile of the visitors. Public events recently organized, such as Les Fondamentales, seemed to have met a large audience, too. The public (i.e, non scientists) is clearly happy by such endeavors and very keen on getting more. Twitter is just another channels of the communication stream. And a versatile one.

France is good at picking american habits with a 10 years delay. So maybe all we have to do is wait until we reach a critical mass of younger people in the labs. In the meantime, my next move is to go to ScienceOnline. We will see what happens, I will tweet from there if I go. Oh, and you can trust my twitter profile. I have an outstanding coffee machine at home. @ajsteven130 can testify.