Back    All discussions

Different kinds of knowledge have different justifications and definitions

The theory of knowledge is concerned with the production and acquirement of knowledge. But what is the definition of knowledge? Traditionally, knowledge is defined as justified true belief (JTB). In 1963, Gettier presented examples in which the subject has a justified true belief which, intuitively, fails to count as knowledge. Although more conditions are added into the definition of knowledge, there still exist counterexamples.

I think the Gettier problem has revealed that validity of statements depend on validity of their justificaion. Take the Gettier problem as example, the statement is right just by chance. Just like I say "it will rain tomorrow". It maybe happen. But the justification is wrong, thus the conclusions is wrong. It offer an new understanding and solution for Gettier problem.

Different kinds of knowledge have different justifications. Based on different justifications, what is known could be classified into six categories: fact propositions, analytical consistent propositions, material consistent propositions, deducted propositions, inducted propositions and analog propositions.

In the cases of facts, they may be defined as "true beliefs". But in the cases of deducted statements, they will be defined as "justified beliefs".
validity of deducted statements depend on validity of their premises and reasoning rules, while validity of inducted statements depend on the fact that there is not exceptional case.

We have no need to seek the universal definiton of knowledge. Alternatively, Different kinds of knowledge have different justifications and definitions. That is what the Gettier problem tells us.
I have written a paper "on the definition of knowledge" in Chinese. I think I have solved the Gettier problem.

Different kinds of knowledge have different justifications and definitions
In the cases of facts, they may be defined as "true beliefs".

Isn't this exactly what Gettier argued against?  Smith, in both cases has a true belief ("the man with 10 coins in his pocket will get the job" and "Either Jones owns a Ford or Brown is in Barcelona" are both true statements of fact that are believed by Smith) but in neither case does it constitute knowledge. 

Different kinds of knowledge have different justifications and definitions
Hi Nigel, My name is Trey...I'm a grad student at USM. Nice to see a fellow Mississippi grad student, what a coincidence.


There may be an argument against a universal definition of knowledge, but l don't think you have it here. What you really seem to be suggesting is that there are different sorts of justification for different sorts of knowledge claims. I think that is fairly uncontroversial. It seems to me that I can fairly well be justified in the belief that it will rain tomorrow -- if I felt like doing the work, I could mount up enough evidence for a justified belief. I personally would shy away from thinking of future contingents in terms of, 'knowledge'...but even if we did....if we allow something less than certainty...then why not if I have strong evidence that it will rain..and it is true that it will rain tomorrow? It seems you are conflating 'chance' a bit here...if 'it will rain tomorrow' is truly a matter of objective chance, then we don't require certainty for those kinds of knowledge claims anyway (I take it we probably have all sorts of probabilistic knowledge of which we are not certain), but if the chance is epistemic uncertainty, but there is a fact of the matter...then why couldn't we have knowledge?

It seems like the only kind of knowledge you want to accept is knowledge of deductive validity...that seems pretty stringent to me.

Different kinds of knowledge have different justifications and definitions
Hi Chaohui
Your suggested approach seems to be a Contextualist one. You might want to look at other  approaches along those lines. 

Unfortunately there are some problems with what you have. Firstly I think you've slightly missed the fundamental issue behind Gettier's examples, guessing it will rain tomorrow would not actually be a Gettier example because it's not sufficiently justified. To make it in to a Gettier example we need to have an initial justification that is rendered false, but a conclusion which turns out to be true anyway. Moreover the original Gettier examples were instances of deductive reasoning so your distinction there doesn't change the problem all that much.


Different kinds of knowledge have different justifications and definitions
Thanks for all the responses.
The example I mentioned is not a suitable example.

The Gettier example is a deducted propostional. In my theory, validity of deducted statements depend on validity of their premises and reasoning rules.
(Statement 1)"The man who will get the job has ten coins in his pocket, because Jones is the man who will get the job, and Jones has ten coins in his pocket."

(Statement 2)"The man who will get the job has ten coins in his pocket, because Smith is the man who will get the job, and Smith has ten coins in his pocket."

Statement 1 is different with statement 2, altough they have the same conclusion. In this example, we say that statement 1 is false, but statement 2 is true.

For decuted propositions, they will be examined together with their justificaions.

Different kinds of knowledge have different justifications and definitions
Hi all,
I am Erickson from Meridian Junior College (Singapore), just a high school student but I took a subject on epistemology during my high school years. Hence my apology if there is lack of understanding on my part which I hope can be remedied here through professional expertise of everyone here.

As far as I know, Gettier problem posits the fact that justified true belief is necessary but insufficient conditions for knowledge. That is, while it is relatively easy for us to admit that justified true belief are necessary conditions, but Edmund Gettier suggested that they are simply insufficient. This can be shown from the examples that everyone has provided (the John-Smith coin problem).

Another example I heard was about a person A entering a room B, and sees B in the room. However, B is in fact hiding in a cupboard such that A does not actually see B, and what A actually sees is hologram of B capable of deceiving A in the process. In this case, A believes B is in the room, A is justified (as he sees the B as a hologram, which to him is a correct justification since it is indistinguishable from B himself), and there is truth value in this since B is in fact in the room (namely, the cupboard).

Hence, in this case, I interpret it as either of these;
1. The problem lies with the fact that the justification is wrong - as A falsely thought that he has justification of seeing B in the room. In particular, JTB is fulfilled because of luck/chance (B is in the room). Therefore, we can say that the justification is wrong and thus A actually has no knowledge since JTB is not satisfied.
2. The problem lies with the fact that in most experiential cases (i.e. cases that involves empirical experience, such as the act of seeing the room) we cannot in practice have the 'Truth', i.e. the T in the JTB trinity. In this case, A cannot access the truth T as B was hidden in the cupboard, so he is in no way think that the B he saw was wrong if it is indistinguishable. I myself as a narrator about this problem concerning A and B can know this truth T because I am outside the context and the fact that B is inside the cupboard is known to me. However, if this is the case, then it is technically impossible for A to declare that he did not see B since he cannot in practice know that B is inside cupboard unless he is some sort of extreme skeptics which also doesn't help.

By virtue of my interpretation, I honestly think that Gettier problem can only be resolved if and only if one can guarantee that a person can have the truth T in all cases in which we are to obtain a particular piece of knowledge, or that a person can guarantee that his justification is indubitable. However, as far as empirical world is concerned, indubitable justification is impossible (unlike deductive statements in which the justifications are certain e.g. mathematical proofs if the premises are true), and we as the one acquiring knowledge cannot detach ourselves from contexts (like a narrator of a story who knows everything, e.g. narrator of the A and B in the example above), so somehow I do not think it is practically solvable.

Any thought on this? Thanks. 

Different kinds of knowledge have different justifications and definitions

I believe that there is a conflict inherent in the justified true belief conditions, and that addressing this conflict is the best way to bring the definition on knowledge in line with our usage of the term. The conflict is between the "true" and the "justified" requirements. When we judge if something is true, this is an absolute judgment, which makes use of all available information. For example, in hypothetical situations we can define certain facts as true or false, and judge them to be true or false for this requirement with access to perfect information. In general usage, it's easier to say if something is true or false with access to more information, and we'll have more confidence in our judgment the more people are in agreement. For this requirement we make use of the maximum amount of information, and as many perspectives, as are available. The "justified" requirement has another specific set of requirements, and we recognize a different amount of information in making the judgment. A possible definition of justified is:

We can say that S is justified in believing p if: p can be logically concluded from all of S's observations and/or appeals to authority that they believe are true.

Inherent in that definition are two limitations. One is that we're relying on only the information available to S, as opposed to our judgment of "true" where we make access of all available information. The other critical limitation is that we accept that S can make use of false or incomplete information if they believe it's true, we are explicitly allowing false information to be used in making the judgment. Many proposals to avoid Gettier style problems will add a condition such as "no false premises" or "causality" which attempt to avoid this issue. However, I think that as long as the definition of knowledge includes a justification requirement there will be conflicts with at least one other requirement, adding additional requirements won't resolve the issue.

The first step therefore is to remove the "justified" requirement, but we'll have to add back in a requirement that captures the important features we're now missing. We should also make note of several assumptions we're making. One, that S can use their observations, and that they can trust their observations. Allowing access to observations should be fairly easy to accept as that is the primary way we gain knowledge about the world. The other assumption, that we can trust our observations, shouldn't be understood as "that looks like a rose, therefore it is a rose" but as "that looks like a rose, I'm not dreaming or hallucinating, therefore that's something that looks like a rose." We also assume that we can make use of experts or authorities, but once we accept the other two assumptions this one is no longer explicitly necessary. An appeal to authority can be understood as "observing that someone is an expert" and "observing them" for example by listening to them, or reading something they wrote. An appeal to authority is therefore just a different kind of observation.

Taking these assumptions in to consideration we can simplify the "justified" requirement as a "logical" requirement, so that we have a definition of knowledge as a "logical true belief."

We can say that S knows p if and only if:

  1. S believes p
  2. p is true
  3. It's possible to logically conclude p from all of S's true observations  

This change removes two key problems. First, by recognizing that we're only going to use true observations we allow ourselves access to the same information for the "logical" requirement that we allow ourselves for the "true" requirement. Two, we're also recognizing that while it's S's observations we're judging, we’re not relying on their perspective to judge if they're true. This means that the definition avoids any conflicts of the amount of information available between the different requirements. If we create a hypothetical situation, and allow ourselves perfect information we can make a perfect judgment if a fact is true or not for both requirements. This should avoid any Gettier or Gettier style problems. For example:

Case 1

"Smith has applied for a job, but, it is claimed, has a justified belief that "Jones will get the job". He also has a justified belief that "Jones has 10 coins in his pocket". Smith therefore (justifiably) concludes that "the man who will get the job has 10 coins in his pocket".
In fact, Jones does not get the job. Instead, Smith does. However, as it happens, Smith (unknowingly and by sheer chance) also had 10 coins in his pocket. So his belief that "the man who will get the job has 10 coins in his pocket" was justified and true. But it does not appear to be knowledge."

However, the justified belief that "Jones will get the job" cannot be logically concluded from true observations, and therefore any conclusions based on it would not count as knowledge under the new definition.

Case 2

"Matthew drives through an area that appears to have many barns. In fact it contains a great many realistic barn facades, perhaps made to help shoot a Hollywood movie 'on location'. When Matthew looks at the one real barn along his route, he forms the allegedly justified true belief, 'There's a barn over there.' But if he follows the strong requirement for justified belief, then his thought process will follow the previous mentioned steps exactly."

However, the justified true belief "There's a barn over there" cannot be logically concluded from the true observation "that looks like a barn." Barns aren't defined by what they look like, they are defined by how they're used. Therefore it would not count as knowledge under the new definition either. It would be possible to conclude: "that looks like a barn" which would be judged correctly as knowledge. Or to inspect the barn carefully to see how it's used, and conclude that "this is a barn since it's used like a barn" which would also be accepted as knowledge.

In both cases the "logical true belief" definition of knowledge not only avoids calling something knowledge which we wouldn't ordinarily call knowledge, it also makes it easier to see exactly where the gap in knowledge is, where the information available falls short of allowing us to know a certain fact. 


This is my first attempt at putting all these thoughts down on one page. I've also posted a slightly longer version here Which includes some footnotes, other slightly off topic thoughts and a printer friendly link. I'll try and update that with any corrections or edits that come up here.

Different kinds of knowledge have different justifications and definitions
I share the position (with Tristan Cunha) that Gettier has sufficiently refuted the JTB model, but for very different reasons.  I will return to these reasons below, but should first state that in my view an error is made in the way the question is introduced.  The "Gettier problem", if he's correct, is not Gettier's, but is a problem with the JTB version of the general correspondence model of knowledge which he strives to refute.  It functions roughly in the same way as Kant's dynamical Antinomies do for the claims Reason makes on the Understanding where unconditioned by possible experience:  to reliably constrain comprehensive claims.  Accordingly, the associated problem is not to repair the problematic criterion (as introduced by Zhuang on 2010-01-20 and attempted also by Tioa on 2010-01-30), but to look for another, if any exist, such as, for example, a "genuiness" criterion, where truth is not determined by correspondence to a knower, but rather a special variety of "being" a knower.  

With respect to my view of the results of the argument, it shows that some propositional contents of given beliefs could be true (correspond correctly under strict bi-valence) without being justified by any relationship those contents have to the belief-grounds by which those contents are held or asserted.  I interpret this part of the argument to be brought out by Zhuang in his 2010-02-12 response, annoting the truncation pointed out of a claim's truth from its tertiary grounds, so that justification of a given belief can have at most a mere accidental relation to the objective truth of its contents.  Therefore my support for Gettier's refutation is based on the radical disjunction of sufficient justification from the bi-valence of truth-claims, restoring a more cautious version of Hume's criteria of consistency of relations between habits of expectations and "bundles of perceptions", to repeat the famous phrase.  This in turn is by default a recommendation for a non-correspondence model altogether, where the relation between belief-justification and truth-claim-correctness is not accidental, but identical in individual cases, such where one makes an assertion about a "true work of art", for example.  

Chunha's attempt to repair the model by 1) collapsing appeals to authority into observations in general, as one variety thereof, and 2) subsequently distinguishing between observational and technical claim-contents (i.e. objects defined by how they look and how they are used) does not directly address (for propostions) Gettier's basic point:  the bi-valence of claim-contents maintains no non-accidental relation to the belief-justification.  By Chunha's argument the unconditionality of the relation is restated at the cost of elimination of observational contents from possible knowledge.  And besides, the use/observation distinction does not hold up at more minute levels, when one is learning a new skill, for example.

Different kinds of knowledge have different justifications and definitions
The fact that the truth criterion eludes practicality makes it almost impossible to formulate satisfactory model that can explain what counts as knowledge and what does not count as one.

Since knowledge differs from truth in the sense that it needs, in principle, a knower, its attainment must inevitably depend on how much information the potential knower can have in principle and in practice. So long as the information available in principle and in practice do not tally, it is impossible to prevent Gettier cases from denying some instances of knowledge acquisition.

Taking the same example of Barn façade case, one realizes that unless the driver actually verifies in practice whether the façade represents true barns or not, otherwise the driver most likely will conclude, according to all the information available through his observation and reasoning, that he sees nothing but true barns all along. Truth criterion spells out that there is only one true barn and the rest are mere façades - which in principle knowable only for a subject with perfect information - while in actuality the person involved in this particular case always have imperfect information.

According to the driver, he is justified in concluding that because that is all the information he has, unless in all knowledge formation, he strives to be a subject with perfect information, which is practically impossible. The 'J' in JTB presupposes the idea that one may not have sufficient ingredients to know something is true or not, while 'T' in JTB requires truth to be attained before we claim to have knowledge. J and T are, therefore, contradictory in their existence as a model of knowledge, because they both cannot be fulfilled at the same time: if J can ever be unquestionably fulfilled, that would make attainment of J no different from attaining T, and thus J would be equivalent to T.

If there is a satisfactory model of knowledge at all, it must necessarily NOT be JTB, nor can the model subsume it if it were to avoid the same problem at all.