Verification and falsification

Within science conclusions are based on observations, or on previous hypotheses that are also based on observations.

Sometimes also a hypothesis, e.g. an attempt to generalize the result or formulate a cause for the observation, is proposed.

The hypothesis, or consequences from the hypothesis, may become additionally verified or falsified by additional observations.

A verification of a hypothesis implies that an observation, or another verified hypothesis, is in accordance with the hypothesis.

A falsification of a hypothesis implies that an observation, or another verified hypothesis, opposes the hypothesis.

The falsification hence implies that an observation (or another verified hypothesis) demonstrate that the hypothesis was erroneous. It hence verifies a negation of the hypothesis.

A verification of a hypothesis increases our belief in the hypothesis. A falsification of a hypothesis decreases our belief in the hypothesis.

A special importance of falsification has been claimed within philosophy of science.

I have, however, not been able to find a strict definition of the term in this context. Therefore a definition is given below:

A negation is a logical operator. Hence there is no fundamental epistemological difference between verification and falsification:

To falsify a hypothesis is equivalent to
verify the negation of the hypothesis

Both verification and falsification are based on observations and cannot be used in claims that a hypothesis is eternally "true" or "false".

This is discussed under Epistemology, section induction.

There is no logical difference between verification and falsification:

"All ravens are black" is a statement that is

• theoretically possible, but hard in reality, to verify.
• possible to falsify.

"Not all ravens are black" is a statement that is

• possible to verify.
• theoretically possible, but hard in reality, to falsify.

The example demonstrates that verification and falsification are fundamental identical activities. In order to verify, alternatively falsify, any of the two statements, all ravens should be inspected, or we must find one non-black raven.

Statements corresponding to "Life exists on other planets" or "Hitler was a paedophile" can be verified using only a few observations, but they both require induction to be falsified.

A very trustworthy statement like "A released stone falls" cannot be falsified.

Philosophers of science like to include when the discovery (verification) of black swans falsified the statement "all swans are white". These historical events are discussed at kunskapsteori.se.