Science Is Not Solitary

There was another round of the "who counts as a scientist?" debate recently, on Twitter and then on the Physics Focus blog. In between those, probably coincidentally (he doesn't mention anything prompting it), Sean Carroll offered a three-step definition of science:

  1. Think of every possible way the world could be. Label each way an "hypothesis."
  2. Look at how the world actually is. Call what you see "data" (or "evidence").
  3. Where possible, choose the hypothesis that provides the best fit to the data.

This isn't quite the way I would put it-- I'm an experimentalist where Sean's a theorist, so I tend to give more weight to observation while he gives more weight to theorizing-- but it's not bad. It is, however, missing a critical element, in my opinion. There's a fourth step to the process, without which it can't be considered complete, namely:

4) Share your results with everybody else.

This may not seem like it belongs on the same footing as the more philosophical and abstract steps Sean gives, but I'm not much for philosophy. And on a practical level, you can't really have science as we know it without sharing results. It's not an accident that the rapid explosion of scientific progress in Europe roughly coincides with the rise of the professional scientific societies and all the other trappings of modern institutional science-- regular meetings, journals reporting the latest discoveries, etc.. Things really get moving, in terms of scientific knowledge, between the early 1600's when Galileo and his contemporaries were encoding their astronomical discoveries in Latin cryptograms and the early 1800's when Michael Faraday provided his own three-step guide to scientific success: "Work. Finish. Publish." That last step is a big part of the reason.

My knowledge of the history of science generally is not terribly systematic-- I've read bits and pieces about specific eras I find interesting for one reason or another, but not that wide a range. I have the impression, though, that rapid progress is always associated with groups rather than individuals. The fragmentary knowledge we have of ancient science generally involves groups of philosophers in dialogue with one another, and even where you have dramatic progress attributed to single individuals, on closer inspection, they're usually tied into a network of other like-minded folks. Modern myth has Einstein devising relativity in isolation, but even when he was a patent clerk, he exchanged ideas with friends and colleagues, chiefly Michele Besso. And Marcel Grossmann was essential to the development of general relativity, which stretched out over nearly a decade, during which Einstein became extremely well integrated with the network of physics in Europe.

If you want a real example of a lone solitary genius in science, it's not Einstein you should talk about, but Ernst Stueckelberg. If you just said "Who?," that's exactly my point. Stueckelberg was a Swiss physicist who almost certainly worked out something very close to the modern understanding of quantum electro-dynamics (QED) years before Schwinger, Feynman, and Tomonaga. He's not well known, though, and his discoveries are basically ignored (relegated to regretful postscripts in most of the histories of 20th century physics) because he didn't share them very well. His only real connection to the international physics network was Wolfgang Pauli, who apparently talked him out of publishing some results that later made others famous. When he did publish results, he mostly sent them to obscure journals, and used idiosyncratic notation that other physicists found difficult to follow. Consequently, he had very little impact.

(Interestingly, he was interviewed for a few of the histories I've read recently, and surprisingly, didn't appear to be bitter about any of this.)

While pop culture might suggest otherwise, science is fundamentally not a solitary pursuit. Even when you're dealing with pencil-and-paper theory, where a single scientist can in principle make a key discovery (as opposed to needing a thousand-person LHC collaboration), science doesn't make progress in isolation. A discovery isn't a breakthrough until it's shared with others, who can check the idea for themselves, and help build on it.

This is also why stories like Wiles proving Fermat's theorem, or Grisha Perelman solving the Poincaré conjecture, or this story about a Shinichi Mochizuki maybe solving another huge problem in math have a lot of pull among scientists. It's because lone geniuses working out solutions to huge problems without ever talking to anybody else about it is really, really strange. That's not generally how science is done. Particularly in the cases of Perelman and Mochizuki, who basically dropped their (maybe) proofs on the community, and disappeared. It's also a major factor in the negative reaction to the Eric Weinstein theory-of-everything business a little while ago-- people in science react very negatively to supposed breakthroughs by outsiders in isolation in part because it's Not How We Do Things.

Returning to the question of who should be considered a scientist, to be honest, I think it's probably the second least interesting conversation it's possible to have about science. I'm pretty much happy to let anybody who follows the general practices of science call themselves a scientist. Sean's definition is pretty good, provided you add that final step. Without it, what you have isn't science, but alchemy.

(Of course, I have my own formulation of the basic process, and am working on a book built around that. And I'm currently working on chapters from the tell-everyone-about-it section, so I pretty much had to post this at some point...)

More like this

The book-in-progress (which is coming along, albeit slowly, thanks for asking) is built around making analogies between scientific discoveries and ordinary activities. This necessarily means telling a lot of historical stories, which is both good and bad. The bad part is that actual history is way…
Over in yesterday's communications skills post commenter Paul raises a question about priorities: I wonder to what extent good writers, public speakers and communicators are being promoted in science in place of good thinkers - people who can challenge prevailing dogma, invent promising novel…
NASA held a big press conference yesterday to announce that the Gravity Probe B experiment had confirmed a prediction of General Relativity that spacetime near Earth should be "twisted" by the Earth's rotation. A lot of the coverage has focused on the troubled history of the mission (as did the…
I'm doing edits on the QED chapter of the book-in-progress today, and I'm struck again by the apparent randomness of the way credit gets attached to things. QED is a rich source of examples of this, but two in particular stand out, one experimental and the other theoretical. On the experimental…

One of the things that helped fuel the scientific revolution was the invention of printing. With printing it was possible to share what you knew/discovered/developed with a lot of other people.

By Bill Gill (not verified) on 07 Jul 2013 #permalink

The shorthand I use for "doing science" is "observe, hypothesize, test, publish, refine." For a theorist, that sequence might be altered to "hypothesize, test, observe...", but the core elements are the same.

To my mind, what qualifies someone as a scientist is that they conduct theoretical or experimental work in a scientific field, and pubilsh their results successfully in the peer-reviewed journals in that field.

Generally this also entails having a graduate degree and working full-time in a field, usually in an academic institution or the research & development branch of a business corporation, but we shouldn't exclude those without such advantages: the work should be judged on its own merits.

The problem with "outsiders" breaking into a field, is of sorting wheat from chaff. The problem of kooks & cranks is well-known, but there are occasional real geniuses out there, and probably many ordinary people who could make diverse minor contributions if they had ways to contact working scientists for feedback and a bit of help getting published. (OTOH, perhaps blogs such as this are useful in that role? "Dear so-and-so, I have a couple of ideas I'd like to discuss, beginning with the following equation...")

Something else just occurred to me. The process I use in engineering (telephone switching systems, which are software-driven) goes "hypothesize, test, observe, refine" (I build systems for clients, so "publish" doesn't enter into it.).

Does that make me more of a theorist at work, even though philosophically I'm ferociously empiricist? This was something of an unexpected discovery;-)

Publication has been such an important part of the scientific process that it has even made its way into pop culture. I turn the microphone over to Tom Lehrer:

And then I write
By morning, night
And afternoon
And pretty soon
My name in Dnepropetrovsk is cursed
When he finds out I publish first

Not to mention the chain of friends by which the narrator of that song learns that the guy in Dnepropetrovsk has something worth scooping.

More seriously, if you are an experimentalist, there is a good reason not to work alone. Labs are full of lots of dangerous stuff like lasers, high voltage, dangerous chemicals, etc. If you have a buddy, then at least there is somebody to call the ambulance if things go wrong.

By Eric Lund (not verified) on 08 Jul 2013 #permalink

" It’s because lone geniuses working out solutions to huge problems without ever talking to anybody else about it is really, really strange. That’s not generally how science is done."

So, basically, you're saying that science can only be done by the rich, male and well-connected, and no one else need be taken seriously?

You're writing as if Stueckelberg *didn't want* his work disseminated. I would like to know more about the how and why of what he did with his work, because there's probably something to be learned there about the limitations of the scientific community and its inquisitiveness.

Everyone wants to bask in the glory of science but just as with military conquests it's the great leaders who matter most and who rightly deserve most of the praise since unlike foot soldiers they are not easily replaceable. Sure Einstein didn't work alone but without him physics of the early 20th century would be set back by decades.

You’re writing as if Stueckelberg *didn’t want* his work disseminated. I would like to know more about the how and why of what he did with his work, because there’s probably something to be learned there about the limitations of the scientific community and its inquisitiveness.

Virtually everything I know about Stueckelberg can be found in The Second Creation by Mann and Crease on Google Books. Start on page 140, if that link doesn't work.

Short version: a combination of bad luck and legal troubles trapped him in Switzerland during the critical years, and he let Pauli (who was relentlessly negative) talk him out of publishing some key ideas. He professed disinterest in getting credit for his work (in 1984 when they interviewed him), though he also had some psychiatric issues at one point, so this might not be wholly reliable. And while he got a number of important things right, he also devoted years to projects that were well outside the mainstream.

Sure Einstein didn’t work alone but without him physics of the early 20th century would be set back by decades.

I think you could make a case that general relativity might not have happened when it did without Einstein, but Poincaré and Lorentz were very, very close to getting all of Special Relativity by 1905; they had all the math, and were lacking only a couple of philosophical points, which somebody was bound to figure out sooner rather than later.

Einstein's heuristic model of the photoelectric effect was technically what got him the Nobel (it's the only specific result mentioned in his citation), and he's been quoted as calling it the only truly revolutionary thing he did. That did build on work by Planck, though, so the key idea was already out there, and some of the key figures in quantum theory spent years vehemently resisting the idea of "light quanta," too. So it's not clear to me that 1) the idea wouldn't've come up anyway, and 2) if Einstein hadn't been around to introduce it, that it would've significantly delayed the development of quantum theory.

So the question comes down to how important you think general relativity was (and how unique to Einstein-- I know Hilbert came close to scooping Einstein on it, but my vague impression is that Hilbert got interested in the problem because of Einstein, and probably wouldn't've developed it entirely independently) and how unique you think Einstein's light-quanta model of the photoelectric effect was. I doubt the loss of his specific contributions would've led to a decades-long delay in quantum physics, and I'm not entirely sure how seriously the lack of general relativity in 1915 would've set anything else back.

I'm not sure science without sharing is alchemy, as alchemy wasn't defined by its solitary-ness, but by its connections with mysticism. Certain alchemists shared results and techniques so I think the step of science they were neglecting was your step #2, in that they didn't look at the world as it actually is, but as some sort of mystical system that could be understood through ancient knowledge.

So, basically, you’re saying that science can only be done by the rich, male and well-connected, and no one else need be taken seriously?

That was a rather athletic jump to that conclusion. The point is that progress in science depends on people communicating their results with other scientists. While most scientists have a Ph.D. (or are in the process of obtaining one), every once in a while somebody off the Ph.D. track makes a significant contribution. The way the rest of the scientific world learns of it is that these people communicate their results to the rest of the world.

As a graduate student, I worked with one of the handful of living people to have a mountain named after him. He earned this honor by making a significant scientific discovery in Antarctica, which was duly communicated in one of the specialty journals of the field. At the time I met him, he did not even have a bachelor's degree (he earned one, in a non-technical field, while I was working with him). But his day job was as an engineer/technician in the lab, and his name appears on several technical publications from the group (including one of my papers).

By Eric Lund (not verified) on 08 Jul 2013 #permalink

Hindsight is 20/20.

Yes, I know all that usual story but people today forget that at the time there were many competing theories and ideas and while it's easy for us to see the right solution among them now that we know it it certainly wasn't the case back then. Those "couple of philosophical points" as you call them were absolutely crucial to make sense of all that mess and to make the jump from some puzzling equations applicable to a few esoteric experiments to the fundamental alteration of the very way we see space and time (not to mention convincing the majority of the need of it). Personally I consider it THE most profound and radical insight in all of the history of science and it certainly required a genius of the highest caliber. It's just amazing how easily people take it for granted nowadays.