Epistemology(n.): The theory of knowledge, from Greek episteme (knowledge) + -logy (speaking, discourse, treatise, doctrine, theory, science)1.

Disclaimer

This article is an elaborate form of quiz practice, so I am making a conscious effort to only focus on what was discussed in my course. If this feels incomplete, consider reading the Internet Encyclopedia of Philosophy and the Stanford Encyclopedia of Philosophy. The best way to learn philosophy, though, is to talk with others.

Introduction

What does it mean to know something? How has the answer to the previous question changed over time? How can we defend against the challenge of skepticism, which claims that we cannot know anything?

The Standard Analysis of Knowledge

Plato (428-423 B.C.E — 347-348 B.C.E.) states necessary and sufficient conditions for knowledge in his dialogue Meno2 3.

The Standard Analysis of Knowledge: For any subject x and any proposition P, x knows that P iff:

(i) x believes that P,
(ii) P is true, and
(iii) x is justified in believing that P.

(i): To have knowledge of something, you have to be aware of it. In other words, knowledge is a mental state of some kind. The belief forms the foundation of what could constitute knowledge, and we need to later check if this belief is true and justified. For example, I could have the belief that an AWP costs $4800 dollars. We will see in the next point that I do not have knowledge of this proposition.

(ii): If I were to tell you something, and there is some black-box method to determine that it is false, then I do not know about that thing. For example, if I told you I know that an AWP costs $4800 dollars, and you looked at the price in the buy menu and saw it was $4750 dollars, then it is fair to say I do not know what the price of an AWP is.

(iii): If I were to randomly fill an answer on a multiple choice test and I got it correct, then I had a true belief regarding the answer. However, I would argue that I would not have knowledge about the answer, as there was no justification process. I just got lucky.

We can immediately start questioning what is meant by each of these conditions, and if they are individually necessary4 and jointly sufficient5 for knowledge. As we will find out, many objections have been raised to this definition.

A Brief Note on Justification

The question of what constitutes proper justification for knowledge is quite relevant. This section introduces three schools of thought: foundationalism, coherentism, and reliabililism.

Foundationalism: Some beliefs are basic, and whether they are justified does not depend on whether any other belief is justified.

For any subject x and any belief that P, x is justified in believing that P iff:

(i) x accepts it as basic, or 
(ii) x derives it from basic ones by suitable means.

I certainly have no issue just accepting things as true (not when I was a kid asking why? repeatedly, though) and then working with them to produce other things (every single course I have ever taken relies on this).

Coherentism: A system of beliefs is doxastically coherent to the degree to which the beliefs it is made of deductively entail one another.

For any subject x and any belief that P, x is justified in believing that P iff:

(i) The coherence of x's system of beliefs will increase if x believes that P.

A flat-earther would be justified in believing additional beliefs about the flat earth because it increases their doxastic coherence, whereas if they were to believe in a round earth, it would not be justified according to this system.

Reliabilism: A reliable method is a method of forming beliefs that reliably produces true beliefs.

For any subject x and any belief that P, x is justified in believing that P iff:

(i) x's belief that P is caused by a reliable method.

This theory attempts to strengthen the original idea behind justification, which was to highlight the underlying cognitive processes involved in creating knowledge. For example, following the scientific method to publish in a good journal would be a reliable process and thus an acceptable source of justification for the beliefs in the journal.

Modern Approaches to the Standard Analysis of Knowledge

Suppose I look at my watch and I see it is 10 PM. I have a justified true belief about what time it is. However, I did not know that my watch battery died exactly 12 hours ago. Do I have knowledge about what time it is?

Many arguments of the above form have been raised, but modern epistemologists recognize Edmund Gettier as the first to fully articulate the relation between these special cases and the Standard Analysis in his 1963 article Is Justified True Belief Knowledge?6. He argues that the three conditions are not jointly sufficient for knowledge7.

Gettier’s Case:

Suppose that Smith and Jones have applied for a certain job. And suppose that Smith has strong evidence for the following conjunctive proposition:

(d) Jones is the man who will get the job, and Jones has ten coins in his pocket.

Smith’s evidence for (d) might be that the president of the company assured him that Jones would in the end be selected, and that he, Smith, had counted the coins in Jones’s pocket ten minutes ago. Proposition (d) entails:

(e) The man who will get the job has ten coins in his pocket.

Let us suppose that Smith sees the entailment from (d) to (e), and accepts (e) on the grounds of (d), for which he has strong evidence. In this case, Smith is clearly justified in believing that (e) is true.

But imagine, further, that unknown to Smith, he himself, not Jones, will get the job. And, also, unknown to Smith, he himself has ten coins in his pocket. Proposition (e) is then true, though proposition (d), from which Smith inferred (e), is false. In our example, then, all of the following are true:

(e) is true, Smith believes that (e) is true, and Smith is justified in believing that (e) is true.

But it is equally clear that Smith does not know that (e) is true; for (e) is true in virtue of the number of coins in Smith’s pocket, while Smith does not know how many coins are in Smith’s pocket, and bases his belief in (e) on a count of the coins in Jones’s pocket, whom he falsely believes to be the man who will get the job.

Here is a summary of common responses to the above case:

  • Gettier Cases do not meet the justification condition and the Standard Analysis holds.
  • Gettier Cases do, in fact, count as knowledge and the Standard Analysis holds.
  • A fourth condition is needed to ensure the Gettier Cases do not count as knowledge (commonly referred to as an Anti-Gettier condition)
  • The justification condition needs to be amended or replaced to ensure the Gettier Cases do not count as knowledge.

One example of a justification replacement is given by Robert Nozick in his 1981 article Knowledge and Skepticism8.

Counterfactualism:

For any subject x and any proposition P, x knows that P iff:

(i) x believes that P,
(ii) P is true, and
(iii) x's belief that P tracks truth:
    a. If P were true, x would still believe that P.
    b. If P were not true, x would no longer believe that P.

    or

    a. In the closest logically possible situation to P, x would still believe P.
    b. In the closest logically possible situation to not P, x would no longer believe that P.

Intuitively, the goal of this replacement condition, known as truth-tracking or sensitivity, is to eliminate the “luck” that happens in Gettier cases.

Consider the broken-clock case mentioned earlier. I glance at my watch and it reads 10 PM, and it just so happens to actually be 10 PM. I have a justified true belief about what time it is. However, under Nozick’s analysis, I do not have knowledge. Why? If it were not 10 PM, would I no longer believe it is 10 PM? No. Because my watch is frozen, I would believe it was 10 PM regardless of what the actual time is. My belief does not track truth, and so by counterfactualism, it does not count as knowledge.

Consider another Gettier-style case known as the Barn County Case. Jon is driving through a county that is secretly filled with barn facades: convincing flat cutouts that look exactly like barns from the road. He happens to glance at the one real barn in the county and forms the belief P: “that is a barn.” His belief is true, and his perceptual process seems perfectly ordinary. On the Standard Analysis, Jon knows that P. However, under Nozick’s analysis, Jon does not have knowledge. If Jon did not happen to look at the one barn in the whole county (i.e., in the closest possible situation where Jon is not looking at a barn), would Jon no longer believe it was a barn? Also, no.

Brains in a Vat

Finally, we will consider the skeptical hypothesis.

Skeptical Hypothesis: For any subject x and any proposition P, the proposition X knows that P is false.

This hypothesis offends our sapient sensibilities as beings capable of having knowledge, so it needs to be addressed so we can strengthen our definitions of knowledge. We will consider Hilary Putnam’s 1981 article describing the Brain in a Vat (BIV) scenario as our form of a skeptical hypothesis, which is a modern take on the Cartesian Evil Demon scenario9 10 11. It supposes that we are disembodied brains suspended in vats of nutrients, wired up to a supercomputer that feeds us a perfect simulation of reality. This skeptical hypothesis can be used to construct an argument that knowledge of the external world is impossible.

An Argument For Skepticism:

P1. Dr. Liz is not justified in believing that she is not a BIV.
P2. If Dr. Liz is not justified in believing that she is not a BIV,
    then Dr. Liz is not justified in believing any proposition P
    about the external world.
C1. Dr. Liz is not justified in believing P. 
P3. If Dr. Liz is not justified in believing P,
    then Dr. Liz does not know P.
C2. Dr. Liz does not know P.

When I first heard this argument in high school (paraphrased as “Can you prove you are not a BIV?”), I was quite upset, because I couldn’t come up with a defense against it.

Putnam’s Semantic Externalism Response

Putnam responds to this argument by rejecting P1. In other words, the proposition Dr. Liz knows that she is not a brain in a vat is true.

An Argument Against Skepticism:

P1. All logically possible situations can be divided into those where BIV and those where -BIV.
P2. For any P, P is a tautology iff P is true in all logically possible situations where BIV and 
    in all logically possible situations where -BIV. 
P3. In all logically possible situations where BIV, B(-BIV) is true.
P4. In all logically possible situations where -BIV, B(-BIV) is true.
C1. B(-BIV) is a tautology. 
P5. Dr. Liz believes she is not a BIV.
P6. For any P and any x, if x believes that P and P is a tautology, then x knows that P.
C2. Dr Liz. knows that she is not a brain in a vat.

Putnam assumes that if you believe something that is a tautology (true in all logically possible situations), then you know it. He removes the justification condition entirely and uses tautology as the accepted form of truth.

Putnam appeals to Semantic Externalism. The key idea is that many concepts, especially what he calls natural kind concepts, are not defined purely in your head. Instead, they are defined in terms of two things: examples from your environment and the deep features of those examples. For instance:

water: x falls under water iff x shares deep features with a*
(pointing at some particular clear liquid in your environment)

This means that the same-sounding word can pick out entirely different things depending on what the world around you is actually like. Your concept of “water” is partly constituted by the physical stuff you grew up pointing at.

Now apply this to the BIV case. Consider Dr. Liz’s belief that she is not a brain in a vat, in each of the two situations:

-BIV: Dr. Liz believes it's not the case that she shares deep
      features with c, where c is a real brain in a real vat.

BIV:  Dr. Liz believes it's not the case that she shares deep
      features with d, where d is a machine-made illusion of a
      brain in a vat.

In both situations, Dr. Liz’s belief is true. In -BIV, she is right that she is not a real brain in a real vat. In BIV, she is right that she is not a machine-made illusion of a brain in a vat. The content of the belief shifts to match the world she is in, and the belief is always true. Putnam concludes that “I am not a brain in a vat” is a tautology, and since Dr. Liz believes it, she knows it. P1 is false.

Two notable objections have been raised against Putnam. First, the BIV-since-yesterday scenario: suppose someone grew up in the real world and was only just turned into a brain in a vat last night. They would still possess natural kind concepts formed from their previous real-world experience, meaning their belief “I am not a brain in a vat” could be straightforwardly false. Putnam’s argument does not seem to cover this case. Second, the two beliefs objection: the content of Dr. Liz’s belief is different in -BIV and in BIV, as Putnam himself establishes. But then what she believes is not really a tautology; it is just a belief that cannot be false in the circumstances where she holds it. Compare this to the belief “I am alive,” which cannot be false whenever I believe it.

Nozick’s Truth-Tracking Response

Nozick offers a different strategy. The skeptic’s argument can be restated in terms of knowledge directly:

P1. Dr. Liz does not know that she is not a BIV.
P2. If Dr. Liz does not know that she is not a BIV,
    then Dr. Liz does not know any proposition P about the external world.
C1. Dr. Liz does not know P.

Nozick’s response is to accept P1 and reject P2.

Accepting P1. Does Dr. Liz know she is not a BIV? Apply the tracking conditions. The sensitive condition asks: if she were a BIV, would she no longer believe she is not a BIV? No. If she were a brain in a vat, the computer would still be feeding her exactly the same experiences, and she would still believe that she is not in a vat. The truth-tracking condition fails, and Nozick agrees with the skeptic on this point: we do not know that we are not brains in a vat.

Rejecting P2. The skeptic needs P2 to get their conclusion, and it looks plausible because of the following principle:

Principle of Deductive Closure:

For any subject x and any propositions P and Q:

P1. x knows that P implies Q
P2. x knows that P.
C1. x knows Q. 

or

P1. K(P)
P2. K(P→Q)
C1. K(Q)

We should note that this is NOT the same as Modus Ponens, as we are applying the concept of knowledge to the propositions.

The argument for P2 would go: Dr. Liz knows that if she has hands, then she is not a BIV. If she knew she had hands, she would therefore know she was not a BIV by closure. Since she does not know she is not a BIV, she must not know she has hands either.

Nozick argues that the Closure Principle is false. Beliefs are not automatically closed under known implication, as we are not perfect rational agents. If Jon believes today is Monday, and believes that if today is Monday then Garfield is unhappy, it does not follow that Jon has actually formed the belief that Garfield is unhappy. Similarly, a subject’s knowledge does not automatically extend to everything their knowledge logically entails.

Once closure is rejected, P2 loses its support. It becomes possible to know ordinary propositions about the external world, like that you have hands, even while failing to know the skeptical hypothesis is false. Your belief that you have hands tracks truth in the ordinary sense: if you did not have hands, you would not believe you did. The fact that this belief does not extend to ruling out all skeptical scenarios does not undermine it.

Of course, rejecting the Principle of Deductive Closure is an extreme measure, but so is the skeptical hypothesis.

Both Putnam’s and Nozick’s arguments have had considerable response for/against them.

  1. Etymonline ↩

  2. In Meno, Plato records a dialogue between Socrates and Meno, where Socrates convinces Meno that a true belief is not enough for knowledge, and that justification is needed. In some quick Google searches, it seems as though Plato himself never strongly advocated for the JTB view, but it is attributed to him. Read through the full dialogue here: Plato ↩

  3. I am sure he mentioned it in other places as well, but Gettier pointed me here. ↩

  4. If at least one of the three conditions is not met, the proposition “x knows that P” is false. ↩

  5. If all of the three conditions are met, the proposition “x knows that P” is true. ↩

  6. Read the article (Only 2.5 pages!) here: Gettier ↩

  7. There are cases where all of the three conditions are met, and the proposition “x knows that P” is false. ↩

  8. Read the article here: Nozick ↩

  9. Read the article here: Putnam ↩

  10. The Matrix was inspired by the BIV scenario, not the other way around. ↩

  11. This is the origin of the phrase “I think, thereform I am”. Descartes argued that the mind would still exist regardless of if the senses were being manipulated. ↩