In this edition of TOKTalk I will talk about Karl Poppers contributions to the philosophy of science. What makes a theory scientific and what is the principle of falsification? Listen to find out!
Sir Reimund Karl Popper is considered to be one of the most influential philosophers of the 20th century, and of course it is difficult, if not impossible to summarize his contributions in a few minutes. Maybe I’ll start out by giving you a brief biographical overview over his life.
Popper was born in 1902 in Vienna, Austria and with the rise of Nazism in Germany and Austria he emmigrated to New Zealand. Later he went to Great Britain where he spent most of his life teaching at the London School of Economics. Popper contributed greatly to philosophy of science and published a notable book on this issue called “The Logic of Scientific Discovery” in 1934. He was also a political philosopher. One of his most well known books on the issue is “The Open Society and Its Enemies”, which was published in 1945. Popper died at the age of 92 in 1994.
Now what were his contributions for the philosophy of science? I think that I need to start at the beginning here. Popper was living at the the time when Albert Einstein formulated his theory of relativity. He was highly fascinated by this progress of science. Einstein formulated his Special Theory of Relativity in 1905 and then the General Theory of Relativity by 1915. At that time Einstein did not yet have experimental evidence for his theory. His theory made predictions, but the actual test to see if the predictions were correct was not performed yet. Einstein’s theory could be tested a few years later during the total solar eclipse of 1919. So Popper was of course excited to find out if the Einstein’s theory will be proven correct or not. If the outcome of the experiment confirms the theory, then the theory will be strengthened, if not, then Einstein’s theory will be abandoned.
Well, the experiment did confirm the theory, but this is was probably not the main point. Popper knew that the scientific community would have abandoned Einstein’s theory if the outcome would not have met the predictions. This is how science works, and how it should work. If the theory fails to make the correct predictions, then you have to replace the theory. This is what you call scientific progress. You replace old theories with better ones.
According to Popper scientific theories can not give you you certainty. The 1919 solar eclipse did prove Einstein’s Theory of Relativity correct, but maybe in the future another experiment will prove the theory wrong. In this case the Theory of Relativity will be falsified, and it is necessary to search for a new theory that is able to explain both the old and the new experimental results. As a matter of fact, even hundreds or thousands of proofs are not enough to ultimately prove a theory absolutely correct. You can never be certain, maybe the next experiment will prove the theory false. One single experiment is enough to falsify a theory, no matter how many times the theory has been proven correct before.
Naturally one has to be a bit careful here. Some theories are very well established already. And it is quite possible that experimental errors give the appearance that a theory is wrong. So it is not wise to jump to too early conclusions and to throw an established theory overboard to quickly.
I would like to illustrate this with an example. Our theory: all swans are white. Now if you go out to the lakes and rivers you can find plenty of evidence for this theory. With every white swan that you find your theory becomes stronger. Maybe you have already investigated all the 5000 swans in the area and came to the conclusion that your theory is indeed “correct”. But how many swans do you have to find to reach certainty? You have to investigate not only all living swans, but all swans of the past and of the future as well. Something that you can not do. But only one single swan of a different color will prove your theory wrong. Only one experiment showing the opposite is necessary to prove the theory false. And indeed, there are black swans in Australia. Your theory now has to be replaced by a newer, more general theory: “All swans are white, except the swan Cygnus atratus of western Australia”.
Now, does this mean that scientific theories are not always correct? According to Popper’s falsification principle, scientific theories are only correct until they are proven wrong. In his view making mistakes is a very important component of progress. Popper coined the term “falsification”. A good scientific theory must be falsifyable. This means that you must be able to come up with an experiment that tries to prove the theory wrong. The statement: “Tomorrow it will or it will not rain” is not scientific, because it is always correct, no matter what the weather is like. On the other hand the statement “The earth is flat” is indeed scientific, because it is testable, and in principle you can show that the statement is false, and as we know, this has already been shown.
I have experienced many times that people are afraid of formulating a clear falsifyable hypothesis when they are designing an experiment. Especially students are concerned that their hypothesis will be proven wrong by the experiment that they are to conduct. We remember, that a hypothesis is an educated guess. So what “hypothesis” do they come up with? Example: “When I mix substance A and B together, then either they will react or they will not react”. Now this is clearly not scientific and the hypothesis will always be correct, no matter what the outcome of the experiment is.
Some people confuse the terms falsifyability and false. A statement is falsifyable if there is the possibility to show that it is wrong, one can in principle conduct an experiment to test it. One could also say, that a falsifyable statement has the potential to be proven false. It does not have to be false. There seems to be also the misconception that only scientific statement have validity. The statement “I love my parents” is not falsifyable, I can not conduct an experiment to prove it wrong. It is therefore not scientific. But I assure you that it has a very high validity for me! Or what about the statement: “I love lemon ice cream”? This is also not falsifyable, you can not prove it wrong experimentally, and hence not scientific. But this does not mean that it does not possess validity, or that it is not correct!
Back to Einstein – it is possible that Einstein’s Theory of Relativity will be proven false sometime in the future? This possibility must of course exist. Otherwise the Theory of Relativity would not be scientific. And indeed such “theory replacements” did already occur in the past. Maybe a Unified Theory, a “Theory of Everything” will one day replace both the Theory of Relativity and Quantum Mechanics. The only thing I wonder is, if a theory that explains everything can be falsified…
And now to a question that some of you are burning to know the answer of: “Is the falsification principle falsifyable?” I once read, I don’t know if it’s true, that Karl Popper used to kick his students out of the classroom for asking such a question. The falsification principle is that what its name says, it is a principle and not a scientific theory. No, the falsification principle itself is not falsifyable, it is not scientific. It belongs to metascience, or philosophy. Don’t forget: the scientific method – hypothesis, experimentation, observation, conclusion – is also not falsifyable, still it is used by scientists on a daily basis.
I think our time is slowly running out again. And if time is running out, it is time for a quote. But I had somewhat of a difficulty deciding between two quotes… So I’ll just tell you both of them. The first one is by Karl Popper himself: “Good tests kill flawed theories; we remain alive to guess again.” And the second one is anonymous: “Theory guides. Experiment decides.”
Questions for Discussion:
- Is it conceivable that some statements are currently unfalsifyable (and thus unscientific) but maybe falsifyable in the future? Is it possible that new discoveries make unfalsifyable claims falsifyable, or are unfalsifyable statements always unfalsifyable? How do we know?
- There are many academic disciplines (history, social sciences, arts, etc. ) in which scientific experiments can not be conducted to gain further knowledge. Still, people working in these disciplines often use techniques that are also used in the natural sciences (establishment of cause and effect relationship, literature research, establishment of explanations etc.) Try to come up with other activities and tasks that characterize “academic work”.
- To what extent are falsifyable experiments also a part of the work of a historian or a psychologist? To what extent can these disciplines be separated from the natural sciences?