A hypothesis ( from the ancient Greek ὑπόθεσις about the late Latin hypothesis literally for " imputation " ) is a statement whose validity is believed to be possible, but not proven or verified. For hypotheses, it is common that the conditions are given under which they are to be valid.

In positivist epistemological currents, the hypothesis is a precursor of a theory to which it can be verified by observation. Unlike in common parlance called "theory" in science so in contrast to the hypothesis, a statement or an interdependent group of statements that have been confirmed by empirical evidence (eg, theory of relativity, gravitational theory, theory of evolution ). An assertion or a guiding principle to be proved or disproved by scientific evidence, ie thesis (eg " economies in itself is devoid of moral content ").

However, critical - rational approaches considered, theory, speculation and hypothesis are equivalent because theoretical assumptions generally can not be verified (the principle of falsifiability ).

Empirical Studies

In the empirical sciences hypotheses have the status of an assumption, which are usually reviewed by the deductive- nomological model. Thereby observed data is applied to the hypothesis and compared so that, if the hypothesis is consistent with the actually observed events. If agreement given the hypothesis was confirmed and the underlying theory, which also includes the assumptions and the experimental arrangement in addition to the hypothesis has proven its worth. Hypotheses that contradict any observation sentence shall not be considered empirically (see falsifiability ).

Hypotheses serve either the Declaration of already known facts or assumptions as general empirical statements of law. Since empirical laws can only be confirmed by a finite number of observations (see induction), such statements may not prove to be definitive, but should be viewed only as proven.

If hypotheses formulated as preliminary assumptions in order to check the possible basic assumptions of a yet undefined theory, one speaks of a working hypothesis.

Can several hypotheses to explain an event, a hypothesis of a rival hypothesis can be distinguished by the conclusion that the best explanation.

At the end of the scientific work or at the beginning in the Management Summary is found then in addition to the summary of the work usually the answer to the question whether the hypothesis could be tested positive or not (See Poincare, Henri: Science and Hypothesis, 4th Edition, Xenomoi Verlag, pp. 152-154 ).


In a logical conversation, a hypothesis is the premise of an argument whose truth is first excluded. This hypotheses act as implications that serve the defense of a thesis. Formal:

Is the thesis ( the consequence of A) is valid under the assumption of hypotheses, each hypothesis must be checked.


In mathematics, the term originally referred to the hypothesis unproven foundations or general principles from which mathematical propositions are derived. Since these principles are used as axioms, does not apply the criterion of truth for them. You are set. The costs associated with them are deductive conclusions.

Brain research

One approach to artificially replicate the brain, describes it fancy and save hypotheses and compare them with perception. Not applicable further hypotheses would be discarded until the remaining is applicable or stored new ( learned ) is.

Statistical hypotheses

The statistical hypothesis that can be falsified with a certain probability, one applies statistical tests. See also: statistics, error 1st and 2nd kind

Historical Aspects


Plato addressed the issue of multiple hypothesis, so also in his dialogue Phaedo ( 100a ):

Abduction with Charles S. Peirce

Charles S. Peirce described the conclusion of an event to a case under the assumption of a rule, as a hypothesis, which is a separate circuit way besides induction and deduction. Do I look as smoke and go by the rule of " Where there's smoke, there's fire ," said I come to the conclusion " there is fire." The final way the hypothesis is logically inconclusive. (see also abduction).

Types of hypotheses according to Poincaré

The philosopher and scientist Henri Poincaré distinguished three kinds of hypotheses:

Illustrative quotes

" Hypotheses, yet vibrant of its own questioning position, rarely come already to the practical evidence to the technical, social change, as of succeeding. They remain in the mere attempt by the mere telling; this fails, they remain true well within the knowledge, they are no longer even in limine outside her how the abstraction, but they wander into the experimental history of the realization of the detected errors. "