Prof Tai Kai NG
Department of Physics, Hong Kong University of Science and Technology

After learning how to measure time by a clock, Kit decided to use his new knowledge to do some scientific researches. After recording the time of sunrise for 7 consecutive days, and taking into account that there are uncertainties in his measurement caused by weather, reading clock, etc., Kit proposed his theory of sunrise:

The sun rises every morning in HK at 6:30am ± 3 minutes!

Question: Can Kit’s work be qualified as scientific research? And can K’s assertion be qualified as a scientific theory? Why?

To answer this question one has to know what do a scientific research and a scientific theory mean. May be you can do a small research here.

Ask your teachers, classmates and friends about what they think science is, tell them the above short story, and ask whether they think Kit is doing a scientific research.

- Discussion on Scientific Statement (Interview with Dr. Tong Shiu Sing) (Chinese Version Only)

You should have gathered now a lot of opinions about what people think of what is science. But, what do scientists think? Is there a definition of science? Does science refer to a particular set of knowledge like physics, chemistry and biology? Or to a particular method of investigation? Or both? You may have identified the following criteria for scientific research:

If you think about it, Kit’s work in fact satisfied all the above criteria! But is that enough for a “scientific theory”?

To the scientific community science is distinct from other subjects by a unique character called falsifiability. We refer to Kit’s story for the explanation on falsifiability.

You may be surprised that what Kit has done qualifies as a scientific research, albeit a very primitive one. Besides satisfying the above 3 criteria, his statement about sunrise is also falsifiable – meaning that there exist criteria to show that his theory can be wrong. In the above example, Kit may find that his prediction remains right for a month or so. However, as time goes by, he will find that the time of sunrise changes significantly from his prediction, thus falsifying his theory.

The requirement of being falsifiable is a unique and very important concept in scientific research. A theory is not scientific if it is not falsifiable, or if there is no way to prove that the theory is wrong even if it is wrong. Let me give an example of ‘non-scientific’ theories: “Newton’s Law has to be wrong somewhere, although I don’t know where it is”. The statement may be right or may be wrong. However, it is impossible to prove that it is wrong because to prove that the statement is wrong, we have to show that Newton’s Law is correct everywhere (in both time and space) in our universe! This is practically impossible even if Newton’s Law is everywhere correct! The requirement of falsifiability imposes a very strict constraint on scientific theories beyond simple true and false – a scientific theory must make non-trivial predictions, something that we cannot recognize immediately as either true or false. But what it says must be testable, in the sense that there must exist a criterion which, if the theory fails the criterion, it is wrong!

The concept of falsifiability often sounds “strange” to people when they see it for the first time because, naively we expect that after performing appropriate tests, we should know with certainty whether a scientific theory is truth or false. In particular, we always hope to assure that a scientific theory is right before it is applied to our technology in daily life! This is at least what is advertised on the TV.

The problem is that this is often impossible!

Let me illustrate by considering the following simple example:

Theory: The sun rises every morning from the East!

This is a theory that originates from our daily life experience. But let’s see whether we can prove it. Well, to prove this statement we have to assure that the sun rises from east not only today, tomorrow, or one hundred years from now, but everyday as long as our universe exists! There is just no way human being can “prove” this! However, one day, if the sun does not come out from the east, we know that the statement is wrong. We don’t have to wait till the end of the universe. The bottom line is:

It is often much easier to prove that something is wrong in science, but very often impossible to prove that it is correct!

This is why the idea of falsifiability is introduced! Very often we cannot prove that a theory is correct and the only thing we can do is to try our best to test the theory by falsifying its predictions.

Good scientific theories are those, which makes highly nontrivial predictions and yet survived millions of tests on their validity. The validity of the theory would be questioned even if one of the tests has shown result in disagreement with theory. To see how hard it is to have a “good” scientific theory, let me take Newton’s Law as an example. Newton’s Law is the theory of force and motion of macroscopic objects. It has led to a lot of important inventions since its fundamental equations were established: including the construction of buildings, cars, airplanes, and the modeling of weather and climate. Imagine how much it is against our common sense to predict that it is possible to have tons of metal flying in the air, supported by only two wings that do not seem too large or strong. Yet airplanes are almost as common as cars nowadays! Scientists also know nowadays that Newton’s Law describes correctly only the macroscopic world. At molecular or atomic levels, Newton’s Law is replaced by a different set of physical law called Quantum Physics. Thus despite all its great success, Newton’s Law is still a scientific theory of limited validity!

For a long time, philosophers have been struggling with the question of what is meant
by knowledge (or truth)? In particular, the argument of whether knowledge has an objective meaning is a central issue in philosophy. A popular point of view is that all human knowledge are artificial (human) constructs and all have limited historical and cultural validity. Putting in mathematical terms, all knowledge are built under a set of self-consistent human invented assumptions or logical structures which cannot be a prior justified, and knowledge built under the category of Science is no exception. In particular, it was claimed that the 'objectiveness' of Science is just an illusion and is unjustified.

In fact, Scientists agree that knowledge in Science does depend on a set of self consistent human invented assumptions as in other subjects. However, science subjects are more 'objective' than others in the sense that scientific knowledge are required to be falsifiable, i.e. in science we admit that our theory may have limited validity and we have a set of criteria to test the validity of pieces of knowledge. The requirement of falsifiability is not enforced in other systems of knowledge and marks the most important difference between science and other subjects.

Let me give an example. We often see on TV finance commentators predicting or explaining the outcome of Stock Markets. Very often they explain that if the Stock Market rises, it's because of reason (a), if it drops, it's because of reason (b), and so on. You may be impressed by the ability of these experts to explain everything, if anything. You may believe that the conclusions drawn by these experts are supported by strong scientific analysis and thus are "scientific conclusions".

From the standpoint of falsifiability the conclusions drawn by finance commentators are often not scientific because they don't make predictions that can be falsified. The bottom line is that a theory which can only explain but not predict is not considered to be a scientific theory.

Another good counter example to science is religion (Christian for example). Unlike science where knowledge is built on "doubt" and tests of knowledge are required all the time (falsifiability), religion is built on the concept of faith or "believe". We have to have faith first before we can see and understand the beauty and truth of the knowledge behind it. This is an opposite approach to Science and the two approaches are complimentary to each other.

[This essay is an extract from a lecture given by the author at HKUST.]