Epistemology: The Definition of Knowledge

Since Plato, most consider a definition as necessary & sufficient conditions ie. being unmarried and a man are necessary and sufficient conditions for being a bachelor. Thus, we must find the necessary and sufficient conditions of knowledge in order to define it.

We can test a claim/denifition by showing counter examples; that one condition is not necessary or that the conditions aren't sufficient.

A good definition is not circular (a bachelor is a bachelor), nor is it negative (defined in terms of what it isn't) if possible and it will help explain a concept, providing more information.

Otherwise known as JTB

A person has knowledge that x if, and only if:

• x is true
• they believe that x
• they have justification for believing x

These are necessary and suffcient conditions.

Necessary conditions are needed as part of the definition - without them, the definition would not be correct. Sufficient conditions, when taken together, encompass fully what x thing is; nothing else needs to be added to make the definition accurate. A condition can be necessary without being sufficient.

Gettier Cases were scenarios that showed people having JTB (justified, true belief) without necessarily having knowledge, since it seems more accidental.

Smith, Jones and the job. Smith has reason to believe Jones will get the job they're applying for (perhaps told by the employer), and he knows that Jones has 3 coins in his pocket as he counted them earlier. He concludes the person with 3 coins in his pocket will get the job. Smith is the one that gets the job, however he also had 3 coins in his pocket. Hence is conclusion is a JTB, yet not knowledge.

There is also the barn county example: https://www.iep.utm.edu/gettier/ (under 4. Other examples)

Brown in Barcelona - Smith has justified belief that Jones owns a Ford. He randomly "guesses" his friend, Brown, is currently in Barcelona. So he concludes "EITHER Jones owns a Ford OR Brown is in Barcelona". His conclusion is true because Brown is in Barcelona, but Jones sold his Ford and is currently using a rental one, thus he doesn't own one. - there is nothing that connects what justifies his belief and his belief's being true. He inferred his conclusion from a false belief. 

-Justification is not a necessary condition: sometimes justification isn't needed at all. If I ask "Do you know who wrote the Meditations?" I do not ask for justification when you reply "Descartes". This is more practical. - However, this does not provide necessary and sufficient conditions for theoretical knowledge rather than practical. Furthermore we can still object this is just a lucky guess rather than knowledge. -HOWEVER, true beliefe may not be sufficient for knowledge but this does not follow that justification is necessary.

-Truth is not a necessary condition of knowledge: Could knowledge simply be justified belief? Relativism about the truth argues that the truth is relative - for example, the earth being flat was true to people in the very very distant past, who had good reasons for believing this "fact" at the time. Therefore, it was true to them. Relativism rejects the unqualified "truth" for "true for".

-However; this could mean that simply believing something to be true would make it true, which is not the case in reality. In addition, it seems wrong for two contradicting justified beliefs to both be "true for" groups of people.

Belief is not a necessary condition: (Plato) a) It is possible to know something without believing it b) knowledge is never a form of belief.

a) An example of this would be a nervous student giving the correct answer on a test, which he knows with justification of revising it but - due to his self doubt/nerves - does not believe he has the right answer. -However, we could say he doesn't have the answer because he did not believe himself and thus 'fully commit' to it. Also, perhaps he did believe - it was just unconscious.

b) Belief and Knowledge involve different faculties and different objects. Knowledge is infallible - you can't know what is false but beliefs aren't - they can be mistaken. They have different powers. Knowledge is about the real world, ignorance is not knowing about the real world - belief is between the two. They have different objects.

Zagzebski: belief may not be always true like knowledge but they can be the same faculty

It is JTB + No False Lemmas. A person has knowledge if they have:

• A belief that x
• Their belief that x is true
• Their belief that x is justified
• They did not infer their belief that x from anything false

This therefore avoids the issues of Gettier cases, as they involve the person infering a proposition from something false ie. that Jones will get the job. 

Fake Barn Country

Henry is driving though Fake Barn Country; he doesn't know that the barns are all fake. When looking at barns, he'd conclude they are real. They are not. He is false, and does not have knowledge.

However, one day, he sees a barn and concludes it is a barn and it is a barn. So he had justified (saw it with his own eyes), true belief that did not come from anything false (such as an incorrect premise). Yet he still does not have knowledge.

Clearly, No False Lemmas is not a full definition of knowledge.

A person knows that x if:

• x is true
• they believe that x
• their belief that x is caused by a reliable method.

However, this doesn't overcome the fake barn country example - Henry used a reliable method; his sight and knolwedge of what barns look like.

Nozick modified reliablism as so:

A person knows that x if:

• x is true
• they believe that x
• If in the same/similar situation, if x were false, they would not believe x,
• If in the same/similar situation, if x were true, they would believe x.

This solves fake barn country, as if the barn were fake, he'd still believe it was real.

Some dislike Nozick's tracking theory of knowledge as it entails externalism (https://en.wikipedia.org/wiki/Internalism_and_externalism#Externalism) - the view that a person does not need to have any internal access or cognitive grasp of the facts that make their belief justified. 

Reliabilists argue this is how people operate normally. We aren't aware of the processes involved with seeing things - we just see them.

The generality problem - the processes involved in the 'reliable' method of sense-experience are highly varied and may not all be reliable. Which process should be considered the reliable one?

For a belief to count as knowledge, it must be justified in such a way that it is certain. They must be 100% impossible to doubt in any way, shape or form.

In all Gettier cases there is not certaintyof the justification - Smith may have misheard about Jones getting a job, or is being tricked by an evil demon etc. 

Arg for Infallibilism:

P1: No one can know what is false
C1: Therefore if I know that P, I cannot be mistaken about P
C2: Therefore, for justification to secure knowledge, it must guarantee truth
C3: Therefore, if I am justified in believing that P, I can't possibly be mistaken
C4: Therefore, if it is possible that I'm mistaken, then I can't be justified in believing that P
C5: Therefore, infallibilism is true.

Infallibilism is too strict. Under this view, hardly anything is truly knowledge (mainly analytic knowledge) and so no one has knowledge of most things. This is impractical.

Zagzebsky: As long as the 3rd conditon is independent of truth like this (T+B+?), no matter how we add to the conditions for knowledge, we will always be able to construct Gettier Cases that show the proposed definition is incorrect. There is a 'recipe' for making such cases (Q represents the conditions added to true belief):

1: start w/ a belief that is Q but false as a result of 'bad luck'
2. Now add some 'good luck' so belief is true after all.
3. This true belief will be Q, as it is exactly like the false belief that is Q but happens to be, by luck, true. But this luck means it is not knowledge.
4. For any theory where Q is independent of truth, knowledge is not true belief + Q.

We need a definition that demonstrates both how and why truth and the third condition (Q) are connected & not merely added together. This connection will have something to do with why knowledge is good in a way that mere true belief is not, leading to virtue epistemology. 

You know that P if and only if:

• p is true
• you believe that p and -
• your belief is a result of you exercising your "epistemic" or "intellectual" virtues.

this focuses on the person and what they do in informing their beliefs.

Zagzebski's version: - you believe that p, - your belief that p arises from an act (or acts) of intellectual virtue.

-A virtue is a state of a person that is good by way of helping the person achieve some good purpose or goal. Intellectual virtues aim at intellectual goods. She says it has 2 components:

1-virtue motivates us to pursue what's good. ie. intellectual virtue motivates us to pursue truth
2-involves component that enables us to be successful; ability to be reliable in forming beliefs.

Knowledge is when a claim is true because of intellectual virtue - not in spite of it!!!

-We need to know more about what it is for a belief to be true because it arises from acts of intellectual virtue before we can reach a verdict in cases like Barn Country. Did Henry reach his correct conclusion as a result of intellectual virtue or not? 

-This view may restrict knowledge to adult human beings, excluding babies & animals. Zagzebsky says that we define virtue more broadly to include relatively automatic, unconscious actions like looking or remembering. This can be enough to know the truth.

-Could we a say a person has knowledge if they exercised intellectual virtue to find out a piece of knowledge, despite not being intellectually virtuous as a person? Is their belief still 'accidentally' true in this case? Should knowledge only be granted to people who have intellectual virtues? 

-Are virtues necessary? We don't seem to need motive in gaining knowledge as long as one discovers the truth reliably