Sunday, April 20, 2014

All science is social

One of the more surprising and profound things I learned in grad school is that science is an inherently social activity.  I'm not talking about societal impact of science - for example, pressure for and concern about the development of nuclear weapons.  Nor am I speaking of attempts by social institutions to control scientific knowledge, like the Church's prosecution leading to Galileo's forced recantation of heliocentrism.

Rather, my point is that science itself is social.  A lone experimenter might come up with some cool findings.  But if they aren't replicated to ensure they are not just a fluke, and if they don't become part of the public body of scientific knowledge, they are pretty trivial.  One of science's most important features - its ability to correct errors over time and come to successively better approximations of "reality" - can only occur within the context of a society of researchers, each interested in proving - or disproving - some specific piece.

And, in the long run, the society is in some ways more important than the question of whether an individual theory or researcher is "right."  Thomas Kuhn's seminal book The Structure of Scientific Revolutions made the point that an older generation may hang onto a paradigm in the face of contradictory evidence.  But that's not a bad thing.  This social/cognitive/emotional feature of science helps to ensure that any new theory challenging the status quo is well-elaborated and supported.  So, the social and personal context of science - with a range of diverse people involved - makes it richer and better.

A great example is the refutation of the 18th century investigation of "phlogiston" - believed to be an element that embodied fire, enabling things to burn.  Taking the Wikipedia article as an example, there is no overt mention of social elements of science.  Most of the discussion is around physical findings and theories.  But, take a look at the "Challenge and demise" section, The first paragraph starts as follows:
Eventually, quantitative experiments revealed problems... Robert Boyle burned magnesium in oxygen, and found the product, magnesium oxide, had more mass than the original magnesium. 
Let's read between the lines - the "quantitative experiments" were shared.  Robert Boyle obviously reported his findings. The findings get all the attention, but the process around them is only implied - as is all too often the case.  Without that process, there would just be a guy here who found something out, and a guy there.  How could there be a convergence of evidence that would lead to a change in the consensus view of fire?

Later in the article, we read the following:
Phlogiston remained the dominant theory until Antoine-Laurent Lavoisier showed that combustion requires a gas that has mass (oxygen) and could be measured by means of weighing closed vessels... These observations solved the mass paradox and set the stage for the new caloric theory of combustion.
Again, the social context is only implied. Lavoisier "showed" his findings - the verb "to show" implies an audience.  His findings "set the stage" for the new theory.  Someone's got to see the stage there, all set, and want to be in the show.  Else, there is no science.

For a more modern (counter?)example, consider Lysenko's rise and fall in the early USSR.  Propped up by the government, his theories spread throughout the country regardless of empirical support.  What was lacking in this case was an environment where ideas could be freely exchanged.  So,

OK, I've beaten this to death.  My point is, science can't exist without being embedded in a dynamic social give-and-take.  Publications and conferences are not just good vita-filler and travel opportunities.  Really, they represent the heart of science.  I think that is not as clear as it should be to the general public.

Friday, April 18, 2014

The straight signal

Here's an idea I've had for a while. You know how you come to a red light, and there are two lanes, and there's a car in the left lane... you get into the right lane because you don't know if that car will be turning left. Then someone comes up behind you wanting to make a right on red - too bad!

If we could rely on people to use their turn signals all the time, you could tell if the car in the left lane will be going straight - its turn signal wouldn't be on. But, alas, that is not reality. So, what we need is a straight signal. You get in the left lane, you turn on your straight signal, then people know they can get behind you.

The straight signal can be helpful in other situations - for example, in spots where most people turn, but you will be going straight. A couple of spots come to mind - for example, where, the road curves and a smaller road proceeds straight. There's always that one yahoo coming toward you who decides you're going to turn just like everyone else, so they'll cut in front of you. Turn on your straight signal to let him know you're coming!