Karl Popper on Demarcation and Induction
An overview of Popper’s criterion for distinguishing between science and pseudo-science, and his take on the problem of induction
Introduction to Karl Popper
An Austrian-British philosopher, academic and social commentator, Sir Karl Raimund Popper is considered one of the most influential 20th century philosophers of science. He was influenced by Albert Einstein, Rene Descartes and Immanuel Kant, and he influenced Imre Lakatos and Paul Feyerabend.
Contributions: The topics that he significantly contributed to are induction and the scientific method. He introduced the criterion of falsifiability. The Logic of Scientific Discovery is his most notable work.
Popper on the Problem of Induction
The Problem of Induction threatens the cogency of all scientific knowledge. While Popper accepts the Humean critique of induction, he disagrees that it necessarily leads to skepticism. Additionally, he rejected the Newtonian and Baconian beliefs that ‘observation’ is the initial step in science.
He claimed that the scientific methodology is not inductive and that the principle of induction is not logical in nature. He gave the following justifications for this claim — while using the inductive method, one goes from singular propositions to universal statements. Since such inductive statements are synthetic statements (descriptive statements that are possibly, but not necessarily true), their negation is logically possible. Therefore, he concludes that science has mistakenly believed that induction is its basis.
Positivists and the Problem of Induction
“The history of science is everywhere speculative” — Popper
Popper was against the positivist view of science being capable of being reduced to a formal, logical system. His perception of a scientific theory is that of an invention/creation stemming from the empirical data from the scientists’ intuition.
Popper believes the main epistemologist/positivist motivation for having faith in the inductive method is their belief that it is through induction alone that they can demarcate between science and non-science.
In line with his rejection of inductive logic, Popper rejects this belief too.
He criticizes Wittgenstein by pointing out how his criterion of meaningfulness coincides with the inductivist criterion of demarcation (which is flawed). This criterion fails to draw a line of distinction between scientific and metaphysical systems as it gives them an equal status (that both systems have meaningful pseudo-statements).
Elimination of Psychologism
Popper was against the psychologist’s approach to epistemology. He claimed that such an approach would only lead to skepticism. Therefore, he refused to accept Hume’s psychological definition of induction and criticized him for presenting the problem of induction as a psychological one rather than a philosophical one.
He explained how decisive scientific arguments need to be purely logical in nature. Hume tries to justify the use of induction in our daily life by appealing to repetition and similarity. But such ‘similar’ cases do not have a perfect same-ness and hence can only be called cases of ‘probable similarity’. They have no logical basis as they are mere psychological anticipations/expectations.
Deductive Testing of Theories
Popper believed that science begins with the formulation of a hypothesis. The aim of the scientific method, then, is not to accumulate inductive proof in support of the hypothesis, but to try to falsify it.
After a hypothesis is developed, predictions are deduced from it. These predictions are then tested through experiments. If a counter-instance is found, then the hypothesis is discarded. If not, then it is considered unfalsified and un-negated.
Therefore, confirmation here refers to those hypotheses that are undefeated conjectures, or whose falsification attempts have been unsuccessful.
Criticism of Popper’s Solution for the Problem of Induction
- The problem of induction can be understood to be the problem of basing judgments about the future on the basis of past evidence. By emphasizing the belief that scientific theories are conjectures, and that there is no rational basis for believing their predictions, Popper denies the possibility of rational judgments being made about the future. Therefore, he does not solve the problem of induction, he merely refuses to recognize it.
- Popper fails to give an account of the basis of our expectations for the future. Much as he tried to avoid it, he ended up becoming a skeptic. Unlike Hume, he did not try to provide a solution to the problem of induction or claim that on being unable to do so, the method should still be used. He argued that scientists should abandon its use altogether.
The Problem of Demarcation
Popper claims that instead of confirming theories, science is principally about falsifying them. It should not be understood as a system of well-established theories that give us the possession of unfalsifiable truths.
There exists a line of demarcation between science and non-science (or pseudo-science).
The first step of scientific theories is tentative guesswork. In other words, science does not begin unless a hypothesis has been formulated. But so do non-scientific theories. Also, both are formed on the basis of empirical evidence. Hence, these are not points of distinction between them. So what makes science different from mere superstition?
As per Popper, scientific statements, unlike non-scientific ones, are falsifiable, unlike the latter in which ‘faith’ plays a role in the acceptance of uncontested statements. It is important to note that ‘falsifiable’ does not mean ‘false’. When applied to a statement, it refers (in principle) to the property of it being falsified (shown to be false) through observation/experimentation.
Thus, a theory can be called scientific only if it can be falsified. Thus, it is falsifiability that distinguishes science from non-science.
As per inductive logic, all empirical scientific statements must be conclusively decidable. This implies that they cannot be falsified. If this is accepted to be science, then scientific theories would not be empirically testable. It is while keeping this in mind that Popper introduces falsification as the criterion of demarcation, as it eliminates the possibility of induction being used in scientific methodology.
Defining the Problem
The problem of demarcation is the problem of finding the criterion that would enable one to distinguish between empirical sciences and metaphysical systems.
Popper’s Intentions
What Popper intended was to formulate a suitable characterization for empirical science. He did not want to distort the issue and relapse into the positivist dogma, so he did not try to justify his proposals as the true/essential aims of science. Instead, he argued rationally in support of them by analyzing their logical consequences. By doing so, he tried to emphasize on their ability to elucidate the problems of the theory of knowledge.
Popper was guided by value judgments and predilections and believed that scientific discovery would be impossible without faith in ideas, even if these ideas are metaphysical (entirely speculative) in nature. This points to the fact that Popper was not anti-metaphysical.
As per him, the following were the ‘first tasks’ of logic of knowledge:
- To present a concept of empirical science in order to make linguistic usage as definite as possible.
- To make the demarcation between science and non-science clear.
Experience as a Method
The intention of empirical sciences as a system is to represent only one thing — the real world (the world that we experience).
The three requirements of an empirical theoretical system are:
- It needs to be synthetic so that it can represent a world that is non-contradictory
- It must fulfil the criterion of demarcation. In other words, it cannot be metaphysical if it is to represent a world of possible experiences.
- As a system, it must have distinguishing traits that separate it from other systems.
Popper elaborated on the third point by saying that such a system is distinguished by the virtue of it having stood up to tests. It can be distinguished by the application of the deductive method to it for analysis and description.
Here, experience would be the distinctive method by which one theoretical system could be distinguished from another. Therefore, empirical science is characterized not only by its logical form but also by its distinctive method.
Therefore, the theory of knowledge (whose purpose is the analysis of method/procedure belonging to empirical science) and the theory of the empirical method (the theory of experience) are synonymous.
Falsifiability as a Criterion of Demarcation
The attribute of being capable of being refuted is the criterion of demarcation that characterizes statements of empirical science. This means that verifying and falsifying them should be logically possible.
“If there is no possible way to determine whether a statement is true then that statement has no meaning whatsoever. For the meaning of a statement is the method of its verification.”— Friedrich Waismann
Popper does not acknowledge the method of induction. He believed inference to theories (from particular statements that are ‘verified by experience’) as being logically inadmissible. This means that theories are never empirically verifiable. Therefore, a criterion which lets the statements which are not capable of being verified into the domain empirical science needs to be chosen.
For Popper, the scientific system should have a logical form which allows for it to be singled out through empirical tests (in a negative sense). Thus, it is necessary that there be a possibility for an empirical scientific system to be refuted by experience.
Non-empirical statement: The new batman comic will or will not be in the comic bookstore tomorrow.
Empirical statement: The new batman comic will be in the comic bookstore tomorrow.
Objections
These are the objections to falsifiability as a criterion of demarcation, and their respective defense.
Objection 1
It might seem unnatural to suggest that science, which is expected to provide us with positive information, should be characterized as satisfying the negative requirement of refutability.
Reply: Chances of clash are more likely when a theory’s logical character is positive as it would then be a positive singular statement (unlike laws that prohibit).
Objection 2
The same objections that Popper raised for verifiability as a criterion for demarcation can be raised against falsifiability.
Reply: Popper’s criterion is based on the asymmetry between falsifiability and verifiability, stemming from the logical form of universal statements. Thus, deductive inferences (through modus tollens) make possible inferring the falsity of universal statements from the truth of singular ones.
Objection 3
Even if asymmetry is accepted, it is still not possible to conclusively falsify any theoretical system.
Reply: Popper asserts that the empirical method shall be characterized as that method which excludes those ways of escaping falsification which are logically possible. In other words, it is its manner of exposing to falsification in every possible way that characterizes an empirical method. It is because of this intense struggle for survival that the best suited system is found.
Solution of Hume’s Problem of Induction
Popper claimed that the resolution of the problem of demarcation through the principle of demarcation also resolves the problem of induction. The contradiction between the central thesis of empiricism (that only experience can decide the truth or falsity of scientific statements) and Hume’s discovery of the unacceptability of the inductive method as the scientific method. Popper points out that this contradiction arises only when it is assumed that being ‘conclusively decidable’ is necessary for all empirical scientific statements. In other words, both their verification and refutation must be possible (at least in principle).
But if we refuse to accept this, and let those statements that are at least falsifiable (even if not verifiable) be accepted as empirical, scientific statements, then the contradiction would disappear.
Therefore, the method of falsification does not presuppose any inductive inference, but only allows those tautological transformations of deductive logic whose validity is not disputed.
