Things I don't understand >

Science • Skepticism • Viewpoint • climate science > Scientific Alliance: Popper versus Kuhn on Climate Science

Scientific Alliance: Popper versus Kuhn on Climate Science

Science, Skepticism, Viewpoint, climate science August 6th, 2009

This just came in from the Scientific Alliance, and I thought it was so good it should be reproduced for posterity:

Time for a new paradigm on climate change?

There are two alternative ways to look at how science progresses. In one corner is the concept of the falsifiable hypothesis, credited to Karl Popper. Popper argued that all science is based on hypotheses, which must be tested to destruction. Sound evidence which does not fit with the hypothesis must logically cause it to be rejected. However, the other side of the same coin is that no hypothesis can ever be said to be proven. Over time, the body of evidence consistent with a successful hypothesis builds up to the extent that it becomes regarded as a theory, for example the theory of General Relativity, or Tectonic Plate theory.

At this stage, theories are treated, to all intents and purposes, as fact. However, even then, quite basic knowledge may, with time, be seen as merely provisional. A classic example is Newtonian mechanics, which fully describes the motion of bodies on the scale we are familiar with, but which breaks down both at the level of elementary particles (hence the development of quantum mechanics) and at a cosmological scale (where relativity comes into play).

Popper used the concept of falsifiability as his criterion for whether something is genuinely scientific or not. Thomas Kuhn, in the other corner of this contest, contributed a different view of how scientists work. He introduced the concept of “normal science” to cover the situation where scientists work on various topics within a central paradigm. In contrast to Popper, the Kuhnian view is that “wrong” results (ie, those which are in conflict with the prevailing paradigm) are considered to be due to errors on the part of the researcher rather than findings which jeopardise the consensus view. However, as conflicting evidence increases, a crisis point is reached where a new consensus view is arrived at: a so-called paradigm shift.

These two philosophical approaches represent the extremes of a spectrum. Popper is the purist, who describes how scientific progress ought to work in an ideal world. On the other hand, Kuhn’s description is more pragmatic and a more realistic view of what actually happens. When a hypothesis is first put forward, it would be quite easy to discard it if early experimental results falsified it. However, when a consensus builds up over time that a particular view is “correct”, it takes plenty of hard evidence to convince people they have been wrong. After all, scientists are only human.

The example often used of this happening in the fairly recent past was the derision which was directed at Wegener’s hypothesis of continental drift, when the prevailing scientific view was that land masses were immobile. Although there were some supporters of this view during the first half of the twentieth century, it was only in the 1950s that an understanding of plate tectonics led to the general acceptance that continents are not static. This was a revolutionary shift in thinking, but the paradigm took many years to change.

But Popper’s description was more nearly correct in the case of cosmology. In the 1950s, there were two competing primary models of the Universe: the Big Bang and the Steady State. By the mid-60s, the accumulation of evidence led most astronomers to accept that the Big Bang was the hypothesis which gave the better explanation of how the Universe behaves.

Coming now to the more topical and contentious case of climate change, it is clear that science is operating in a Kuhnian fashion. There are a number of observations which would apparently serve to falsify the hypothesised enhanced greenhouse effect. Not least of these are the missing signature of CO2-driven warming (an enhanced rate of warming in the upper troposphere relative to the Earth’s surface) and the lack of warming across the greater part of Antarctica. The response to this – from those who do not simply dismiss the evidence out of hand – is to point instead to evidence which is consistent with the AGW hypothesis and to introduce a range of fudge factors such as aerosols to account for the observed lack of correlation between atmospheric carbon dioxide level and average temperatures.

The behaviour of a great many researchers involved in climate change is far from Popperian. Rather than test their hypothesis by trying to falsify it, they look instead for evidence which supports it and, in a deeply unscientific manner, will often simply dismiss contrary evidence on the basis of minor flaws or criticism. This is research done according to prejudice rather than with an open mind. To compound the error, and because evidence can only be gathered by observation rather than experiment, increasing reliance has been placed on computer models.

Making headlines in the Guardian last week was a study not yet even published. Jointly written by Judith Lean of the US Naval Research Laboratory and David Rind of the NASA Goddard Institute for Space Studies and due to appear in Geophysical Research Letters, this is billed as the first analysis of the combined impact of human influences (including CO2 and aerosols), solar radiation, volcanic eruptions and ENSO (the El Nino Southern Oscillation) on global temperatures.

Their main conclusions are that anthropogenic global warming has been masked in recent years by reduced solar activity and a lack of a strong positive El Nino, but that a projected increase in solar activity will cause temperatures to rise at a rate 50% faster than projected by the IPCC. Many readers will of course remember that mainstream researchers have generally downplayed the role of variations in the Sun’s output as insignificant in terms of global temperatures, but there now seems to have been a reinterpretation to fit the facts.

But the main criticism of this paper (or at least, what has been reported prior to publication) is that it is not a scientific study but the output of a computer model. The study smacks of damage limitation, of a desire to find some rational explanation for the lack of temperature rise over the past seven or more years. The explanation is that well, yes, natural variation can be important, but that this is only creating a temporary masking effect, soon to disappear. Suspicions about the motivation for the paper are only increased by the Guardian headline: “New estimate based on the forthcoming upturn in solar activity and El Niño southern oscillation cycles is expected to silence global warming sceptics”.

Highly unlikely, as this is merely hypothesis and, crucially, it is not directly falsifiable. But what is important is that the authors are predicting the return of global warming in the next few years, and that the upward trend will be higher than before. If this does not occur, then we must conclude that their analysis is wrong. If they are wrong, it may be because the coming solar cycle will be a weak one, as many people are predicting. And, if so, the logical conclusion may be that natural cycles are more important than carbon dioxide emissions.

In the meantime, Henrik Svensmark and colleagues from the Danish National Space Centre have published a paper in the same journal which gives support for the hypothesis that cosmic rays, modulated by the solar wind, can indeed alter the degree of cloud cover and hence affect temperature (Svensmark et al; Cosmic ray decreases affect atmospheric aerosols and clouds; Geophysical Research Letters; Vol 36, L15101, doi:10.1029/2009GL038429, 2009). Their measurements indicate that cloud cover measured over oceans decreases to a minimum approximately a week after cosmic ray minima. The effect can take large quantities of liquid water out of the atmosphere. This hypothesis may or may not be right, but it remains a working possibility and should certainly not be dismissed lightly.

So, climate science, heavily influenced by global warming politics, continues to adhere to a central paradigm as described by Kuhn. Contrary evidence is clearly not going to be accepted as falsification. It will be fascinating to see what trends there actually are in climate over coming years and, if the predictions of renewed (and faster) global warming come to nothing, then what else will be necessary to cause the crisis which will lead to a paradigm shift. In the meantime, we have to hope that politicians do not take us too far down the road of trying to control the climate based on the current paradigm.