The importance of irrefutable theories in the light of falsification
Karl R. Popper achieves with his work on the subject of the character of science a tool which helps to classify the scientificalness of research. The famous and simple concept of falsification, in which the irrefutability of a theory in an conceivable event is claimed to render a theory unscientific, appears to make it easy to classify the truthfulness and credibility of a theory.
At first this statement might appear logical and comprehensible but there are a number of cases which do not live up to the strict limits of the theory but nevertheless may picture the truth. These theories are often called 'metaphysical' and cannot be falsified by empirical evidence although observations might support the theory. The interesting aspect of this category is the possibility of developing into 'scientific' theories over the time. Karl Popper gives a few examples for these kinds of theories related to historical physical explanations of nature phenomena. The conclusion is, theories in the context of physics to be sometimes 'unscientific' due to the lack of refutability but depict observations and are therefore based on physical principles. Only the limitation in comprehension of the context or the technological equipment at this certain point in time leads to the classification as 'non-scientific'. With additional knowledge as well as an improvement in technology, these theories, which do not need to be untruthful, might become testable and refutable and therefore 'scientific'.
Although most of the examples for these kinds of theories can be found in history because of the initiating research for comprehending nature, it is sometimes possible to find these issues in modern times as well. In the following, there are two examples for recent ( in the face human history, the atomic model by Rutherford from 1911 is considered to be recent in this text) scientific theories which help to exemplify the limits of significance of falsifiability in scientific theories.
The probably most common example for this is the concept of the Higgs-Boson which is an elementary particle predicted by the British physicist Peter Higgs in the 1960s. Because of the immense amount of energy and luminosity needed to create an Higgs-Boson in an particle accelerator, the possibility of verification of the theory has been impossible for a long time. Therefore, the theory needed to be classified as 'non-scientific' in the sense of falsification. The point is not that criteria of falsification e.g. 'the refutability by any conceivable event' (Popper, 1963) are violated because of the lack of people to imagine a test which is able to produce substantiated results, but the inconceivability of a process to perform the contemplated task. When the Higgs-mechanism was originally published in the 1960s, the technological realization was impossible which eventually results in the testability not to be existing at this time and even further not to be developable in the foreseeable future. Technically speaking, the theory must had to be classified as 'non-scientific' for people who didn't believe in the development of a technology to create the needed circumstances. As a consequence, the falsifiability of this theory solely depended on the faith of people in the feasibility of the testing method.
It still needs to be mentioned that even though the testing of the theory is possible nowadays and even if there are signs of the existence of the Higgs-Boson, the concept of falsification prohibits its existence to be accepted. This is due to the famous concept of hypotheses only to be disputable and not affirmative.
When you look back in history, applying the concept of falsification leads to assessing most of the physical theories we know to have been 'non-scientific' for the people of the period in time because of the inconceivability of testability.
Another example to explain this is the atomic model by Thomson and Rutherford. Initially, it is to mention that both models are still used to exemplify simple principles of behavior of materials (often in school). Of course, there are better descriptions of the atomic and molecular composition and behavior today.
Originally, the atomic model of Rutherford replaced the model of Thomson when an experiment was executed by Geiger, Marsden and Rutherford in 1909 (gold foil experiment) which led to results, the old model from Thomson could not explain. Therefore, in the sense of falsification, the model of Thomson must have had become 'scientific' the moment, Rutherford or someone else took the fact into consideration that the old model lacks the explanation for a phenomenon and thought of a way how the expected behavior can be tested. Or, to put it differently, to conceive of an experiment targeted on disproving the model of Thomson. The results of the experiment led to the new theory of Rutherford whereas the Thomson's theory was proved to be wrong.
From the point of view of people in this time, the testability of Thomson's theory had been developed with die 'gold foil experiment' which makes it 'scientific'. The fact, that the test didn't support the assumptions of the model made it untruthful but nonetheless 'scientific'. Rutherford's model, on the other hand, proved to be more accurate than its predecessor but, due to no test to be available which targeted on disproving the theory and the key aspect of an hypothesis, it needed to be classified as 'non-scientific'. Strictly speaking, not until a new way of testing the Rutherford model was invented, it should have been 'non-scientific' but nonetheless state of the art.
These two examples show that testability, the key concept of falsification, highly depends on the time the theory is formulated as well as the status of knowledge about the phenomenon the theory shall explain and the technological advancements. Therefore, the status 'non-scientific' does not necessarily represent the validity of a theory but rather expresses the status of science to be able to investigate a theory based on phenomenological observations.
( The text was inspired by Karl R. Popper's 'Conjectures and Refutations' (1963). An excerpt can be found on "http://www.stephenjaygould.org/ctrl/popper_falsification.html")