Gettier problem

from Wikipedia, the free encyclopedia

The Gettier problem arises from an objection to the so-called classical analysis of knowledge (KAW or JTB for Justified True Belief). This defines knowledge as an opinion (belief) that is justified and true. This is countered by the fact that a justified and true opinion can also be true by “ chance ”. This in turn would mean that this opinion is then no knowledge, as the KAW would however claim, which is why the KAW is wrong. The Gettier problem now is to improve the KAW so that it can deal with such cases of accidentally true justified opinion, or to replace it with another, better analysis. The problem goes with the famous three-page essay Is Justified True Belief Knowledge? (1963) (Eng. Is justified, true opinion knowledge? 1987) by the American philosopher Edmund Gettier . The Gettier problem is one of the main problems of current epistemology .

The standard analysis of knowledge as justified true opinion

Up until the publication of Gettier's paper, most epistemologists assumed that knowledge could be analyzed as justified, true opinion. More accurate:

A subject S knows that P , if and only if:

  • (i) S believes that P ,
  • (ii) P is true , and
  • (iii) S has good reason to believe that P .

According to the first condition (condition of belief) one can only know what one also believes. (Here, as is often the case in current epistemology, “believe”, “think”, “consider to be true”, “be convinced” are used synonymously.) According to the second condition (truth condition) one can only know what is actually the case is. According to the third condition (justification condition), the opinion must not be guesswork or the like; the believer must be able to give reasons for his belief. What exactly is meant by the third condition is controversial. However, Gettier only presupposes two assumptions about justification that are comparatively uncontroversial (cf.Gettier 1963: 121):

  1. Fallibilism: A justified opinion can be wrong.
  2. Deductive unity: If one derives another opinion from a justified opinion in a logically correct manner, then the second opinion is also justified. The Gettier problem should therefore apply to all those variants of the standard analyzes of knowledge that share the two assumptions about justification.

Gettier's counterexamples

Gettier gives two counterexamples for justified, true opinions that are nevertheless not knowledge. The three conditions are met in these cases, but it is not a question of knowledge. The three conditions of the standard analysis are therefore not sufficient.

First counterexample

Smith and Jones applied for a job. Smith has strong reasons to believe the following:

  • (a) Jones is the one who will get the job, and Jones has ten coins in his pocket.

Smith's reasons for (a) are that the boss assured him that he would end up picking Jones, and that he, Smith, counted the coins in Jones' pocket ten minutes ago. From (a) Smith draws the following conclusion:

  • (b) The person who gets the job has ten coins in their pocket.

Smith believes (b) only because he sees the logical conclusion from (a) to (b). In this case, Smith is justified in believing (b).

Now, however, without Smith knowing, Smith gets the job, and Smith, too, without Smith knowing, has ten coins in his pocket. (b) is therefore true, although (a), from which Smith (b) inferred, is false.

In our example, the following applies:

  • (i) Smith believes that (b),
  • (ii) (b) is true,
  • (iii) Smith is justified in believing that (b).

But it is also clear that Smith does not know that (b) is true. Because (b) is only true because Smith got the job and had ten coins in his pocket, not because Jones got the job and had ten coins in his pocket. Smith bases his opinion that (b) on counting the coins in Jones' pocket and wrongly assuming that Jones would get the job.

Second counterexample

In Gettier's second counterexample, Smith assumes that Jones drives a Ford. He has strong reasons (or evidence ) for this assumption: Jones has always driven a Ford since he met Jones, and Jones took Smith out for a drive while sitting in a Ford. In addition, Smith has another friend named Brown who has no knowledge of Smith's current whereabouts.

With the strong evidence for the proposition

  • (c) "Jones owns a Ford"

in the back, Smith concludes the further proposition (d):

  • (d) Jones owns a Ford or Brown is in Barcelona.

Now it turns out completely different: In truth, Jones doesn't own a Ford at all, he just lent it. Hence, proposition (c), for which Smith had strong evidence, is wrong. Coincidentally, however, Brown is actually in Barcelona, ​​which proves proposition (d). Again, Smith has a true justified opinion, but still no knowledge.

General characteristics

In general, Gettier problems are characterized and characterized as such that a justified belief comes true in a different way than expected. A new analysis of knowledge must tackle this epistemic accident.

Problematization of different approaches

Eliminate false assumptions

In the two examples, the justified true opinion comes about through a conclusion from wrong premises . However, it would be premature to think that the standard analysis could now simply be improved by adding an additional clause:

  • (iv) The opinion that P is not based on a conclusion from a false assumption.

Further examples have shown that this modified standard analysis is not sufficient either (see Feldman 1974), such as the following: Smith enters a room and sees Jones. He immediately forms the justified opinion

  • (e) Jones is in the room.

In fact, Smith had not seen Jones, but a faithful replica of Jones. Now, as luck would have it, Jones is really in the room - although Smith of course hasn't seen him at all.

Smith's opinion that (e) is a justified true opinion that is not based on false assumption, but still not knowledge.

Demand for causality

The idea of demanding a causality between the fact and the belief in it, that is

  • (iv) The opinion that P , is causally related to  P .

solves the basic problem and also the above example of the simulation, since there is no causal connection between the presence of Jones and Smith's belief in it. However, a causal connection can also be added here if one assumes that the replica is only in the room because Jones created it there. In this version, the new definition of knowledge is therefore also inadequate. The objection that there is no adequate causal connection here lacks a definition of the term “adequate”.

Requirement of a probability of success

Another possibility would be to require that epistemic success, i.e. This means that the statement believed to be true was also likely. Because then the highly improbable cases of Gettier problems that have occurred would be excluded from the concept of knowledge.

However, the belief that you will not win the lottery would also have to be knowledge if it proves to be true, as this event is very likely. To claim that you not only believe but know that you have not won, on the other hand, seems absurd.

The epistemic difference between opinion and knowledge

A German refutation of Gettier problem is found in Steen Olaf Welding in the epistemic distinction between opinion and knowledge: There are reasons for the opinion of P that are not consistent with the knowledge that P . Thus, the respective assessment of the reason (s) for P is decisive: If Q is taken as a sufficient reason for P , then Q could be the reason for a person to claim that he knows that P , and if Q as an inadequate reason for P is judged, then you could Q be a reason for them to think or to believe that P . Since there are no generally valid criteria for the sufficient or insufficient reasons for the truth of a statement P independent of the judgment of a subject, it is not possible to define the concept of opinion or that of knowledge.

Historical classification

Often Plato is already credited with understanding knowledge as a justified true opinion. In Dialog Menon, for example, the traditional definition of knowledge (episteme) can be found as a correct opinion bound by justification (orthe doxa): Instead of slipping away from being momentarily considered, the correct idea is permanently retained by a justification. In the Gorgias , too , knowledge and belief are defined by the fact that knowledge always includes truth, but belief not necessarily; Likewise, Plato speaks in the Politikos of “true opinion with security” (alethes doxa meta bebaioseo).

However, Plato questions this analysis in Theaitetus : He is just negating that knowledge (episteme) is "true opinion [about x] with knowledge of a difference [from what distinguishes x from all relevant alternatives of the x-like type], a reason." or an explanation "would be (doxa orthê meta epistêmês diaphorotêtos: logou […] proslêpsis). The definition of knowledge as “true opinion with justification” is rejected, since the justification of an opinion would again have to be justified, as well as the justification of the justification, which would lead to an infinite recourse . Rather, there should be an unfounded beginning of all justification. The justification of an opinion must therefore be based on existing knowledge in order to turn the true opinion into knowledge. However, the definition “knowledge is a true opinion based on knowledge” cannot be valid either, since the term to be defined is contained in the definition and this would lead to a circular argument . The dialogue ends aporetically .

In the first half of the 20th century, however, it was widely rumored that Plato analyzed knowledge as a true, justified opinion and should therefore be seen as a pioneer of standard analysis.

See also

Web links


Primary literature

Secondary literature

  • Gerhard Ernst, Lisa Marani (Ed.): The Gettierproblem. A balance sheet after 50 years. Mentis, Münster 2013, ISBN 978-3-89785-840-4
  • Richard Feldman: An Alleged Defect in Gettier Counterexamples. In: Australasian Journal of Philosophy , 52, 1974, pp. 68-69.
  • Alvin Goldman : A Causal Theory of Knowing. In: The Journal of Philosophy , 64, 1967, pp. 335-372
  • Martin Grajner: Epistemology. In: Breitenstein, Rohbeck (ed.): Philosophy. Metzler, Stuttgart / Weimar 2011, p. 147 (149–152)
  • Keith Teacher, Thomas Paxson: Know ledge: Undefeated Justified True Belief. In: The Journal of Philosophy , 66, 1969, pp. 1-22.
  • Robert Nozick : Philosophical Explanations. Cambridge / MA 1981, ISBN 0-674-66479-5
  • Marshall Swain: Epistemic Defeasibility. In: American Philosophical Quarterly , 11, 1974, pp. 15-25.

Individual evidence

  1. Steen Olaf Welding: The epistemic difference between opinion and knowledge. In: Ders .: Where am I? Some essential problems in philosophy. Meiner, Hamburg 2016, pp. 37–44
  2. ^ Plato, Menon 98a: dêsê aitias logismô.
  3. Plato, Gorgias 454d.
  4. Plato, Politicus 309c.
  5. On interpretation: Timothy Chappell:  Plato on Knowledge in the Theaetetus. In: Edward N. Zalta (Ed.): Stanford Encyclopedia of Philosophy . , Kelly L. Ross: Knowledge , 2007
  6. ^ Plato, Theaetetos 210a-b.
  7. Plato, Theaetetos 203 cd.
  8. ^ For example, in the commentary by Lewis Campbell: The Sophistes and Politicus of Plato . Oxford 1867, p. 184: “'real true opinion with confirmation:' i. e. knowledge, as defined in Theaet. sub fin. and Meno 98 a, b; Phaedo 76; Tim. 51 d, e; Legg. 2, 653 b ”. Hans Henning Raeder, Platons Philosophische Entwicklungselung , Teuber 1905, p. 347, points out that at the end of the Theaetetus only the doxa is related to real objects, but a higher status is reserved for knowledge. Another Rainer Enskat: Authentic knowledge. What epistemology can learn from Platonic Socrates . In: Amicus Plato magis amica veritas. Festschrift for Wolfgang Wieland on his 65th birthday . Berlin / New York 1998, pp. 101-43, 103f. says Plato had in Theaitetos given the "most sophisticated working definition of knowledge term" seamlessly imposed at the time Gettier.