True, False, or Not Sure? Philosophy of Science in the 21st century

Whatever happened to objective reality? Why is anthropogenic climate change so hard to “prove”?

Many of us were brought up on the idea of an objective reality determined, at least provisionally, by experiments and analyses performed by a closed but international group of scientists, freely exchanging ideas, testing and refuting hypotheses. This loose federation of scientists formed Michael Polanyi’s “Republic of Science” [1] Michael Polanyi, “Republic of Science,” Minerva 1 (1962) 54-73[2019-10-19]. (1962) and they were guided by principles described by Robert Merton [2] Sociology of science: Wikipedia. Robert Merton(online).[2019-10-15]. (1942) To be “skepticism, universalism, communalism and disinterestedness.” Hypotheses tested widely enough and found not to be at variance with current evidence became theories. A crucial property of any theory was that it should logically be falsifiable by new evidence, as stated by Karl Popper [3] Falsifiability, [2019-10-15]. around 1960. At some point in time, evidence at variance with an accepted theory can lead to a revolution first and eventually a paradigm shift process described by Thomas Kuhn. While a discipline’s fundamental claims are shifting from one perspective to another, groups of scientists are attempting to disprove each other until one theory is generally agreed upon, a condition that must be fulfilled for science to progress. All theories are provisional because they are all based on observation and all observations have limitations, either in extent or in the accuracy of the measurements, which should always be expressed in terms of probabilities. The next revolution is always waiting around the corner. This is “Normal Science.”

Fast forward to 1993, when Funtowicz and Ravetz [4] S. O. Funtawicz and J. R. Ravetz, “Science for a post-normal age,” Futures 25 (2009) 739-755 defined “Post-Normal Science,” where “facts are uncertain, values in dispute, stakes are high and decisions urgent.” Post-normal science recognizes that multiple perspectives can and should be incorporated in order to solve complex problems. Scientists bear a much higher burden of responsibility for communication both amongst themselves and with a wide audience outside science. The continuing turmoil over climate change is a perfect example of post-normal science, but the same features are seen in the debate over energy supplies, GM foods etc. In some ways, individual engineers and medical professionals have always done “post-normal science.”

Let me contrast two recent “discoveries” to illustrate the differences. The first is the discovery of neutrino oscillations in the early years of this century (normal science) and the second is the discovery of anthropogenic climate change (post-normal science). In the first, I had a strong personal stake, which I shared with a few hundred other physicists, but almost no-one else. In the second, I also have a strong personal stake, but so do you, and so does everyone else on the planet.

Contrasting Normal and Post Normal Science

The effort to prove or disprove the existence of neutrino oscillations (you don’t need to understand what neutrinos are or why/how they oscillate to get the gist of this argument) followed Merton’s norms perfectly. Everyone involved was always skeptical, right to the end. The subject had universal implications, as neutrinos have played a part in every major event in the universe from the Big Bang to the explosion of supergiant stars that produced the heavy elements are fundamental to our makeup and our Earth. It was a big communal effort, involving many groups of physicists from around the world. We were disinterested, for although everyone has biases, we worked hard to remove those from our experiments and analyses (which were “blind,” as in medical trials), and ultimately, disproving the existence of neutrino oscillations would have been almost as big a deal as proving they do exist. The burden of proof, one way or the other, lay with the scientists, and no-one else: the closed republic of science. Time was not really of the essence, although politicians and funding agencies were often anxious about delays and costs. Failure to determine the answer one way or the other would have primarily been of consequence to the careers of the scientists involved. Either way, the world would have gone on pretty much the same way whatever the result. The universe may depend on neutrinos, but not of our understanding of them.

Now consider the discovery, if that is the right term, of anthropogenic climate change, the global disruption of the climate by human activities. To start with, no-one would claim this to be established as absolute truth or fact. The Intergovernmental Panel on Climate Change (IPCC) puts the probability that most of the observed warming in recent years is human caused at a minimum of 95% [5] IPCC, [2019-10-15]. . That the climate is changing significantly is as close to 100% certain as makes no difference (ask any polar bear or Australian farmer). I would be less surprised if the 50-year hunt for neutrino oscillations proved to be all in error, than if the climate was shown not to be changing. The human component is another matter. But the burden of proof is different because the implications are enormous and the timescale for action very short. This is not neutrino oscillations, because much more hangs on the outcome than the careers of a few physicists. Neither is it a court of law, where our society has decided that convicting an innocent person is a much worse error than allowing a guilty one to go free. Here it may be the other way around; taking action that proves ultimately not to be necessary would likely be much less costly than not taking necessary action. Reducing our footprint on this planet is a good idea for many reasons even if the climate proved to be stable. And because the implications are so enormous, everyone has to be involved: scientists, governments, businesses, individuals. And no-one, not even the scientists, is disinterested, because we have nowhere else to go and children born now will have to live in the last decades of this century. That makes the decision-making much more complicated, and it would be even if the facts were universally agreed upon. But they are not, because “facts” rely on making predictions about the future behaviour of a very complex system, which is intrinsically hard and depends on mathematical models, which will always be imperfect representations of reality. We don’t even have a good handle on the uncertainties, as the prospect of runaway positive feedback means that the outcomes could be well outside the means and standard deviations presented by the IPCC. (Even if we did know the probabilities of each outcome, we as a species seem to have trouble dealing with probabilities rather than “facts,” and scientists tie themselves in knots explaining them. The timescale is a further complicating factor. Because radical climate change seems to be occurring on a timescale of a few decades – shorter than one human life – we don’t have the luxury of taking the time rigorously to test the theory and to achieve worldwide consensus. Welcome to post-normal science.

For more details, ref. [6] Mike Hulme, “Why we disagree about climate change,” Cambridge 2009 is a great read.


Revised (CEW) 2019-10-15


1 Michael Polanyi, “Republic of Science,” Minerva 1 (1962) 54-73[2019-10-19].
2 Sociology of science: Wikipedia. Robert Merton(online).[2019-10-15].
3 Falsifiability, [2019-10-15].
4 S. O. Funtawicz and J. R. Ravetz, “Science for a post-normal age,” Futures 25 (2009) 739-755
5 IPCC, [2019-10-15].
6 Mike Hulme, “Why we disagree about climate change,” Cambridge 2009