from Wikipedia, the free encyclopedia

A hypothesis (from ancient Greek ὑπόθεσις hypóthesislate Latin hypothesis , literally 'assumption') is an assumption formulated in the form of a logical statement , the validity of which is believed to be possible, but which has not yet been proven or verified . The hypothesis must by their conclusions verifiable, while depending on the result either proved or checked or refuted be would . When formulating a hypothesis, it is common practice to state the conditions under which it is supposed to be valid. With clear logical relationships, this happens in the following form:

Whenever ..., then ... "

In positivist epistemological trends the hypothesis is the precursor of a theory to which it verified by observations may be, it is possible assuming anyone to refute them unique. In scientific language , the term "theory" , in partial contrast to general usage , denotes a group of related logical statements that are at least partially confirmed by empirical evidence (e.g. the theory of relativity or the theory of evolution ). An assertion or a guiding principle ( observation sentence ) that is to be proven or refuted by scientific evidence is called a thesis (eg: " Doing business in itself is free of moral content").

Critical-rational approaches, on the other hand, take the view that theory , speculation and hypothesis are equivalent, since theoretical statements cannot be verified, but at most falsified (principle of falsifiability ).

Empirical Sciences

In the empirical sciences, hypotheses have the status of an assumption that usually has to be checked according to the deductive-nomological model . Observed data is applied to the hypothesis and examined whether the hypothesis and the observed events match. If there is agreement, the hypothesis is confirmed.

Testing a hypothesis is often supported by one's own empirical studies. Since empirical investigations can only cover a finite number of observations, such statements are not conclusively proven , but only regarded as proven (so-called proven hypothesis ).

If hypotheses are formulated as preliminary assumptions on the basis of which further work is planned, one speaks of a working hypothesis . If errors in the assumptions emerge in the further course or if more precise findings emerge, the working hypothesis is adjusted. In contrast to the by a scientific experiment to be tested hypothesis before the experiment set are not changed in the course, otherwise the increased risk of that occurring in the experiment random correlations incorrectly as actual ( causal interpreted) effects ( "a Shoot an arrow at a wall and then paint the target around it ").

At the end of the scientific work or at the beginning in the management summary, in addition to the summary of the work i. d. Usually the answer to the question whether the hypothesis could be tested positively or not.

If several hypotheses can explain an event, by inferring the best explanation, one hypothesis can be distinguished from a rival hypothesis, but this does not mean that the hypothesis thus distinguished must also be the correct one.

With probability (Probabilitätsaussagen) statements of empirical science noisy hypotheses, for example, on a metric scale :

The more ..., the more / less ... "


In a logical conversation, a hypothesis is the premise of an argument whose truth is initially factored out. Hypotheses act as implications that serve to defend a thesis. Formally:

If the thesis (the consequence ) is valid under the assumption of the hypotheses, the individual hypotheses must be checked.


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

Statistical Hypotheses

The statistical hypothesis uses the opposing pair null hypothesis and alternative hypothesis. If a hypothesis has a certain probability statement as its content, it is checked by statistical tests on:

If probabilistic hypotheses are formulated on a nominal scale , the hypothesis form is:

If… then it is more likely that…. "

Historical aspects


Plato treated the theme of the hypothesis several times, including in his dialogue Phaedo (100a):

“I always base my investigation on an assertion that I consider particularly strong; and that which I then have the impression that it is in harmony with it, I call true; what, on the other hand, does not seem to be in line with it, I call untrue.

Abduction in Charles S. Peirce

Charles S. Peirce called the conclusion of an event assuming a rule for a case as a hypothesis that represents its own conclusion in addition to induction and deduction. For example, if I see smoke and use the rule “Where there is smoke, there is fire”, I come to the conclusion “There is fire”. The conclusion of the hypothesis is not logically clear.

Hypothesis as a form of judgment in Kant

In his Critique of Judgment, the philosopher Immanuel Kant differentiates between four approaches for theoretical proofs:

  1. logical-strict conclusions (e.g. through deduction )
  2. Use of analogies
  3. the likely opinion
  4. Hypothesis as a "possible explanation"

In his view, the hypothesis is initially only one of four possibilities to find a proof, but he makes the minimum requirement of a hypothesis, "at least the possibility of it must be completely certain".

Types of hypotheses according to Poincaré

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

  1. Natural hypotheses. These are hypotheses of a very general kind that are reflected as background knowledge. Examples are the existence of the outside world and the principal recognizability of things. Natural hypotheses can only be given up with difficulty because they are constitutive for the scientific community.
  2. Indifferent hypotheses. These hypotheses function as conventions , since one could also assume the opposite hypothesis for a theory and this would only complicate the theory, but not refute it.
  3. Generalizing Hypotheses. Only this type of hypothesis can be confirmed or refuted by experience. They are obtained by induction connections .

Illustrative quotes

“Hypotheses, still vibrating from their own questioning, rarely come to practical evidence of technical, social change as more successful. You remain in the mere attempt at mere explanation ; if this fails, they certainly remain within the knowledge , they are no longer in limine outside of it, like abstraction , but they wander into the experimental history of knowledge, of the errors recognized . "

- Ernst Bloch : Über Fiktion und Hypothese, 1953, in: Gesamtausgabe Vol. X, pp. 21–26, p. 25

See also


Web links

Wiktionary: Hypothesis  - explanations of meanings, word origins, synonyms, translations

supporting documents

  1. ^ Duden Online: Working hypothesis .
  2. Archive link ( Memento of the original from October 12, 2011 in the Internet Archive ) Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice.  @1@ 2Template: Webachiv / IABot / www.ib.hu-berlin.de
  3. Poincare, Henri: Wissenschaft und Hypothese, 4th edition, Xenomoi Verlag, pp. 152-154.
  4. Critique of Judgment , page 466 ff., In: Das Bonner Kant-Korpus , Documentation, Electronic Edition.
  5. in limine means something like on the threshold