Science and truth

“Überzeugungen sind gefährlichere Feinde der Wahrheit als Lügen.”, (Beliefs are more dangerous enemies of truth than lies.), Friedrich Nietzsche, I, Aph. 483

Truth scepticism

In addition to scepticism about the reference objects of statements, there are also doubts about the necessity or content of the concept of truth. A distinction must be made here between formal and factual truth.

The formal aspects of truth are dealt with, for example, in mathematical model theory¹. Formal truth refers only to abstract statement structures and not to specific worldly objects. It is about relationships between formal languages and the structures that interpret them.

A connection to reality only arises in model-theoretical semantics (the study of meaning), which links formal syntax with meanings.²

There is considerable controversy about the relevance of factual truth, which has recently received particular media attention through terms such as ‘post-truth’ or ‘post-factual’.³

Concept of truth

The predicate ‘true’ can be circumvented linguistically in many situations by simply saying ‘it is the case’, ‘it is valid’ or ‘it is a fact’. However, this does not eliminate the factual concept of truth, but merely packages it in alternative linguistic constructs.

The redundancy or deflationary theories of truth consider the predicates ‘true’ and ‘false’ to be superfluous and reduce the statements ‘A is true’ and ‘A is false’ to ‘A’ and ‘not A’ respectively.⁴ Nevertheless, the truth values ‘true’ and ‘false’ remain indispensable in propositional logic. In everyday speech, it is also helpful to be able to talk about true and false statements.

The concept of truth is linked to classical propositional logic, which is indispensable in both everyday life and science. In a criminal trial, for example, the question may be whether a person A shot a victim or not. According to classical logic, A cannot have both shot the victim and not shot the victim. Therefore, the statement that A shot the victim can only be either true or false. To find out which is the case, the court also consults scientific forensics. Certain facts, such as blood spatters of a certain pattern on A’s clothing matching the genetic profile of the victim, together with other evidence, lead to the conclusion that A did indeed shoot the victim. However, this conclusion is only valid if the theory of genetic code is also accepted as true. This means that a criminal trial that uses scientific forensics must assume the truth of this science.

What is ‘factually true’?

There are different views on what it means for a statement to be factually true:

1. In the correspondence theory, statements are considered true if they correspond to facts in objective reality.
2. The coherence theory considers a statement to be true if it is in a maximally logically consistent relationship with other statements.
3. The consensus theory of truth states that a statement is true if it is recognised by a majority or consensus of people.
4. In relativism, truth is not absolute, but relative to certain perspectives, contexts and framework conditions.
5. The standpoint theory is related to relativism, in which truth depends on an individual or social standpoint.

Multiple truths?

Relativistic approaches confuse truth with concepts such as opinion, interest, perception, state of mind or background experience. The latter are, of course, dependent on the individual. For example, a person under the influence of drugs perceives the environment differently than someone who is not under the influence of drugs. However, the aim is to eliminate subjective influences rather than simply accepting multiple contradictory statements.

There is also the danger of self-contradiction. For if relativism is to be absolutely valid, it has thereby refuted itself. Ultimately, relativism is a declaration of bankruptcy on the part of science, since it abandons the quest for truth.

There can also be no ‘knowledge justice’ between several competing and incompatible systems of statements, because knowledge is - regardless of our certainty about it - more or less objectively valid. It is not knowledge itself, but its application or relation to reality that can be judged as just or unjust in the light of certain ethics.

Consensus and coherence?

Whether a statement is true or false can only be determined on the basis of criteria that are all fallible. A broad consensus on the truth of a statement is one such indicator. However, the view that truth is popular agreement or a social construct is highly fallible and unsuitable as a scientific concept. Although the consensus and agreement achieved in a broadly argued discourse play an important role in science, empirical observations and analytical definitions are nevertheless indispensable.

According to the criterion of coherence, the truth of a statement depends primarily on whether it fits seamlessly into an existing system of statements or beliefs without creating contradictions. Logical coherence is also an important aspect of the truth of a system of statements. However, logic alone is not sufficient. This understanding of the term lacks an explicit reference to empiricism. Another problem is that there can be several internally consistent systems that contradict each other. A conspiracy theory can be logically coherent but still false.

Truth and error

One form of radical scepticism is the view that today’s truth will likely be tomorrow’s error. Immanuel Kant wanted to prove Newtonian physics to be absolutely true. However, unlike the theory of relativity, it produces incorrect results when large masses or high relative velocities are involved. Classical physics also provides incorrect predictions when it comes to tiny quantum particles.

So is there nothing but error? This far-reaching pessimism is not necessary, however. It turns out that new, usually more fundamental theories converge on older theories or contain them as approximations under the conditions under which the older theories were valid. For example, the special theory of relativity provides almost identical results to Newtonian mechanics for speeds that are much smaller than the speed of light. For large quantum numbers or macroscopic systems, quantum mechanics approaches the classical description (e.g. via the Ehrenfest theorem). This is referred to as the correspondence principle.

This situation arises in very general statements. A mere reference to ‘true’ or ‘false’ is often no longer sufficient here. Instead of understanding truth as an all-or-nothing principle, it is better to speak of partial or gradual truth in such contexts.⁵ Of course, this does not eliminate the possibility of error, because even the assessment of partial truths remains fallible in principle.

Correspondence theory

The goal of science is to reconstruct objective reality as best as possible - independently of subjective perceptions, cultural conventions or political objectives. It does not base its claim to knowledge on virtual constructs such as the ‘matrix’, but on what exists even without observers. It is not concerned with mere consensus or the production of arbitrarily many consistent theoretical worlds. Personal perspectives or moral assumptions are not its yardstick, but rather the subject of its investigation. Scientific integrity is precisely demonstrated by its willingness to bow to reality - not the other way around.

Within the framework of this scientific programme, the correspondence theory of factual truth therefore appears to be the most appropriate concept of truth. Although scientific statements remain fallible as a result, competing approaches either cannot rule out errors or do not even attempt to attain objective knowledge, tending instead towards dogmatic or subjectivist approaches.