The Science Algorithm

The Science Algorithm

The American Heritage Science Dictionary defines science as "the investigation of natural phenomena through observation, theoretical explanation, and experimentation, or the knowledge produced by such investigation. Science makes use of the scientific method, which includes the careful observation of natural phenomena, the formulation of a hypothesis, the conducting of one or more experiments to test the hypothesis, and the drawing of a conclusion that confirms or modifies the hypothesis." The Oxford English Dictionary likewise defines the scientific method (or "science algorithm") as a "method or procedure that has characterized natural science since the 17th century, consisting in systematic observation, measurement, and experiment, and the formulation, testing, and modification of hypotheses." 

Anything that subverts or undermines the scientific method is like a lie that once told, multiplies its effect exponentially. Like a soot cloud from a chimney, the effect swirls out, obscuring scientific truths, sometimes for years. 

Scientific misconduct comes in many flavors, including everything from ethical lapses, to fraud, negligence, deliberate bias, and premeditated dishonesty (trimming, forging, cooking--three terms were coined by computer pioneer Charles Babbage in 1830, which shows that we are not dealing with a new problem). Indeed, any such action that willfully compromises the integrity of scientific research introduces "glitches" or "bugs" in the algorithmic process of science. As science studies the ever more subtle aspects of nature, it is increasingly important to make sure that our scientific conclusions are as free from bias as possible. Of course all observations are biased by the transfer function of the measuring instruments and the unconscious bias of the observer, but that is a given.

A recent science controversy, Climategate, is part of a broadening debate about failure of the scientific peer review process (witness the WMO/IPCC's much maligned 2007 4th annual assessment report, that carelessly used dubious data pertaining to Himalayan glaciers). Other important recent science controversy concerns scientific fraud (i.e., manufacturing data and altering, fudging or biasing experimental results to favor a particular outcome) in research funded by commercial interests. Good old-fashioned bias (which can be additive or subtractive) seems to fall in the middle somewhere. 

This is a debate about a defective process in science, not defective people. The factors leading to various "sciencegates" (and to a general increase in scientific fraud generally, particularly in the biological science, but also Earth science) seem to be a result, not only of the Internet, but also of the need for researchers to please the powerful government or corporate bodies that fund them. Scientific fraud may also be committed to serve a particular viewpoint (e.g., Intelligent Design). In contrast to the nice, tidy, synthetic papers that grace so many scientific journal pages, many of which claim use of the scientific method, science is actually a very messy, ill-defined, process. Nevertheless, when the public spotlight falls on science and finds problems, true science suffers. The solution is to fix the problems, not turn off the spotlight! 

A debate during Climategate about whether certain emails were stolen, or whether they were actually obtained under the FIA, is simply a smokescreen. Email works like this: Everything you send in open (unencrypted) form across public networks is public. Whether you are a climate biaser or a bank robber, your emails are only a little more private than a small advertisement in the newspaper. Moreover, modern information tools and the Internet make it increasingly easy to commit fraud of any kind, not just scientific fraud. The Internet has created a more open society, and as a result we must all live to higher standards of accuracy and truth. This is good for society, and is good for science. It is also really good for the environment. Acts against the planet that went unchecked in the past are now on public view. This is our brave new world. We can like it or lump it; that is our choice. If we have lived in the past by the philosophy that some particular end justifies the means, now that viewpoint is much more likely to become fodder for public debate. I highly recommend a paper by Brian Martin, an Aussi academic, that discusses the prevalence of and sociological reasons for scientific fraud (http://www.bmartin.cc/pubs/92prom.html).  

Martin's paper is particularly interesting to me because decades ago when I worked as a scientist, I frequently came across a lot of this sort of thing, and (powered by the Internet) it seems to have only got worse since. To give an idea of the magnitude of the problem, school teachers now insist that essays and papers (written with the aid of the Internet) be submitted electronically so that they can be analyzed by an anti-plagiarization service; the teachers are fighting fire with fire. 

Science also needs to go outside the old algorithmic processes, and in particular address the negative influence of ideology-based government funding and also address the level power and influence of various politically-motivated organizations that would seem to operate under the peer-review process umbrella, but don't really. On this topic, it always comes to my mind how US Pres GW Bush once said to his captive scientists and analysts regarding WMDs, "That's the wrong answer, go back and look again!" 

Even blogs about science are themselves subject to bias, however feel free to comment, by way of an impromptu peer review... Now, I wish there were bias-free-blogger web cams in the arctic to see how the polar bears are doing...it's very difficult to know who to believe...