From Great Issues in Philosophy, by James Fieser


Copyright 2008, updated 1/1/2021



A. Skepticism

Radical Skepticism

Criticisms of Radical Skepticism

B. Sources of Knowledge

Experiential Knowledge

Non-Experiential Knowledge

Rationalism and Empiricism

C. The Definition of Knowledge

Justified True Belief

The Gettier Problem

D. Truth, Justification and Relativism

Theories of Truth

Theories of Justification

What’s so Bad about Relativism?

E. Scientific Knowledge

Confirming Theories

Scientific Revolutions


Study Questions


Some years ago, 39 members of an organization called Heaven's Gate committed suicide in the belief that they were shedding their earthly bodies to join an alien spaceship that was following the path of a comet. The centerpiece of the cult's belief system was that there are superior beings out there in the universe that exist on a higher level than we do on earth. They have perfect bodies, roam the galaxy in spacecrafts, and have mastered time travel. Occasionally these aliens send an away team to earth, who temporarily take the form of human beings, and assist willing human students in transforming to this higher level. Jesus and his disciples, they believed, were an earlier away team of alien teachers. Another away team appeared in the late twentieth century. After the right training, human students would need to kill their physical bodies, which would release their spirits into the atmosphere. Nearby alien spacecrafts would then retrieve the spirits and provide them with perfect bodies.

             It's one thing to have an interesting idea about how the universe runs. It is quite another thing to know that the idea is true. Heaven's Gate members indeed claimed to know the truth of their views. That knowledge, they explained, begins when superior aliens implant special wisdom in the minds of select human students; this knowledge is further refined as they study under their alien teachers. But does this count as genuine knowledge? The Heaven's Gate cult is a frightening example of how badly things can go if you get your knowledge claims wrong. One of the central concerns of philosophy is to understand the concept of knowledge, which might help us distinguish between the convictions of a Heaven's Gate believer and the convictions of, say, a scientist. We want to see what precisely it means to "know" something, and what the legitimate avenues are for gaining knowledge. We'd also like to know how to respond to skeptics who say that all knowledge claims, including scientific ones, are just as uncertain as the views of Heaven's Gate believers. These are the primary concerns in the philosophical study of the concept of knowledge, which goes by the name epistemology, from the Greek words episteme (knowledge) and logos (study).

             There are two ways that we normally use the term "knowledge." First, I might say that I know how to do some task, like fix a flat tire on my car or run a program on my computer. This is procedural knowledge, which involves skills to perform specific chores. Second, I might say that I know some proposition, such as that "Paris is the capital of France." This is propositional knowledge, which is knowledge about some fact or state of affairs in the world. As important as procedural knowledge is in our daily lives, it is propositional knowledge that interests philosophers and is the focus of this chapter.




According to Heaven's Gate believers, their knowledge about the superior alien race came principally from the aliens themselves who took the form of human teachers. However, once becoming human, the alien teachers were stripped of their previous memories and knowledge. All that remained for them was a hazy image of the higher level, which they struggled to convey to their human students. Ironically, they explain, the aliens purposefully imposed this knowledge restriction on themselves since "too much knowledge too soon could potentially be an interference and liability to their plan." Immediately we should be suspicious about the belief system of the Heaven's Gate cult since the sources of their knowledge are so shaky. Not only must the human students blindly trust the statements of their supposed alien teachers, but the alien teachers themselves have no clear memories of their previous alien lifestyle. Genuine knowledge must have some evidence to back it up, which we don't see here. It's thus pretty natural for us to be skeptical about cults like Heaven's Gate that make extravagant claims with little concrete evidence. If we didn't have this built-in suspicion we'd be suckered into every hair-brained scheme that came along.

             But how far should our skepticism go? As long as there have been philosophers on this planet, there have been skeptics who have cast doubt on even our most natural beliefs, such as my belief that the table in front of me actually exists. One ancient philosopher, for example, believed that everything in the world changed so rapidly that, when someone spoke to him, he could not trust that the words meant the same thing by the time they reached his ears. He thus wouldn't verbally reply to anyone, but would only wiggle his finger indicating that he heard something. While this is an extreme reaction, it vividly illustrates the notion of philosophical skepticism, which is the view that there are grounds for doubting claims that we typically take for granted.


Radical Skepticism

There are many kinds of philosophical skepticism, and one distinguishing factor involves the extent of the skeptic's doubt. Limited skepticism focuses on a particular claim, such as the belief that God exists, or that there is a universal standard of morality, or that there is intelligent life elsewhere in the universe. In each case, the skeptic would argue that we should doubt the specific claim in question. Many of us are limited skeptics about at least some beliefs, yet at the same time we fully believe other claims. For example, I might be a religious skeptic about God's existence, but not be a moral skeptic about a universal standard of morality. Then there is radical skepticism, which maintains that all of our beliefs are subject to doubt. For any belief that we propose, we cannot know with certainty whether that belief is true or false. This is the type of skepticism that has attracted the most interest among philosophers. A couple centuries ago traditional thinkers argued that this kind of skepticism is a danger to everything that we hold sacred and it threatens to set civilization adrift on an ocean of chaos. The foremost task of philosophers, they argued, should be to combat radical skepticism and establish the certainty of our most important beliefs. Time has shown, though, that this was a false alarm: radical skepticism has not done any apparent damage to society. In fact, radical skeptics have maintained that there is a special benefit to skepticism: it can make us more tolerant of others when we realize that we ourselves cannot claim to have superior knowledge.

             There are three general strategies for defending radical skepticism, each named after its originator. The first is Pyrrhonian skepticism, which was inspired by the ancient Greek philosopher Pyrrho (c.365-c.275 BCE). While Pyrrho wrote nothing, through his teachings he started a skeptical tradition whose aim was to suspend belief on every possible issue. The Pyrrhonian position is this: for any so called fact about the world, there are countless ways of interpreting it, none of which we can prefer above another; we should thus suspend belief about the nature of that thing. Take, for example a red ball that's in front of me. My eyes tell me one thing about it, but my sense of touch tells me an entirely different thing. To someone else who is color blind or has chapped hands, it will have a different set of features. To a dog it will appear even more differently. Suppose that someone was shrunk to the size of a molecule and sitting on the ball: the ball's surface would seem flat, not round. Suppose someone else expanded to the size of a mountain and was looking down on the ball: it would appear to be a speck with no recognizable features at all. We get used to the way that we perceive things like a red ball, and we assume that the ball actually has the features that we perceive. According to the Pyrrhonian skeptic, there's no basis for preferring our individual perspective over any other one. Arguments supporting any claim to knowledge will always be counter-balanced by opposing arguments, thus forcing the suspension of judgment about the original knowledge claim. Thus, views of the physical world, God, morality and everything else are all merely a matter of perspective, and the wisest course of action for us is to abstain from believing those views. Doubting everything, Pyrrhonians argue, will give us a sense of peace since we will no longer be pulled back and forth in controversies about science, God, morality, politics, or anything else.

             The second approach to radical doubt is Humean skepticism, defended by the Scottish philosopher David Hume (1711-1776). According to this view, the human reasoning process is inherently flawed, and this undermines all claims to know something. The problem is that when we list the reasons for our various beliefs about the world, we'll find that many of the explanations are contradictory. For example, if I follow one course of reasoning, I'll come to the conclusion that the ball in front of me really is round. But if I reason in another way, I'll conclude that the ball's roundness is just a matter of my perspective. Maybe the ball really is round; then again, maybe it's not. It makes no difference what the truth of the matter is since we now cannot trust anything that human reason tells us. It's like being tested on a math problem: it makes no difference if you accidentally come up with the right answer. Once you've made a mistake in your calculations, your solution to the math problem is wrong, and you get no partial credit. Similarly, human reasoning is defective, and it's irrelevant whether it accidentally leads us to the truth.  After exposing a series of contradictions within the human reasoning process, Hume makes this dismal assessment:


The intense view of these manifold contradictions and imperfections in human reason has so wrought upon me, and heated my brain, that I am ready to reject all belief and reasoning, and can look upon no opinion even as more probable or likely than another. [Treatise of Human Nature,]


Thus, for Hume, everything that we reason about is based on faulty mental programming, and we need to regularly remind ourselves of this before we get too confident about what we claim to be true.

             The third approach to radical doubt is Cartesian skepticism, named after French philosopher René Descartes (1596-1650). On this view, our entire understanding of the world may just be an illusion, and this possibility casts doubt on any knowledge claim that we might make. Descartes himself was not a skeptic, but he tentatively used a compelling argument for radical skepticism as a tool for developing a non-skeptical philosophical system. Descartes speculates: what if he was just a mind without any body, bobbing around in the spirit realm, and everything he perceived about the world was implanted in his mind by a powerful evil demon. Everything he assumes about the world, then, would be false. He describes this scenario here:


I will suppose that ... some evil demon with extreme power and cunning has used all his energies to deceive me. I will consider that the sky, the air, the earth, colors, shapes, sound, and all other external things are nothing but deluded dreams, which this genius has used as traps for my judgment. I will consider myself as having no hands, no eyes, no flesh, no blood, nor any senses, but falsely believing that I have all these things. [Meditations, 1]


Descartes did not actually believe that he was being manipulated by an evil demon. His point is that this is a theoretical possibility that undermines all of our knowledge claims. I look at a ball in front of me; while it seems to really be there, I can't know this for sure since my experience might be an illusion imposed on me by the evil demon. 

             Of the three approaches to radical skepticism, the Cartesian version has captured people's imagination the most. Science fiction movies play off this theme. For example, in the film The Matrix, people's bodies are suspended in tubs of goo and their brains are wired into a massive computer that generates an artificial reality. Similarly, a commonly used example in contemporary philosophy is that of the brain in a vat: a mad scientist puts a person's brain into a glass jar and wires it to a supercomputer that creates an artificial reality. Whether the mechanism is an evil demon, the Matrix, or a mad scientist, the victims' experiences are so convincing that, from their perspective, it's impossible to tell that the perceived reality is fake.


Criticisms of Radical Skepticism

With arguments as extreme as these, traditional philosophers wasted no time trying to stamp out the fire of radical skepticism. Four arguments were commonly used. First, even if there are ample reasons for me to doubt everything, there is still one truth that is irrefutable: my own existence. For, even if I say "I doubt that I exist," I must still be present to do the doubting. The act of doubting itself requires a doubter, and so my own existence will always be immune to skeptical doubts. This was the criticism that Descartes himself made of radical skepticism, which he encapsulated in the expression "I think, therefore I am." But radical skeptics have not been impressed by this maneuver. The problem with Descartes' solution is that it assumes too many things about what the "I" is behind all those doubts. Most importantly, it takes for granted that the "I" is a unified, conscious thing that continues intact as time moves on. But, according to the skeptic, this conception of the "I" relies too heavily on memory. I assume that I'm the same person now that I was a few moments ago because that's how it seems in my memory. But memory is a very easy target of doubt. Imagine that, every half second, an evil demon wiped clean all of my memories and gave me entirely new ones. One moment I think I'm a farmer, a half-second later a caveman, a half-second later a frog. For all I know, says the skeptic, that is actually happening to me right now, and in that situation it would seem pretty meaningless for me to assert that "I exist".

             A second common attack on radical skepticism is that we cannot behave as skeptics in our normal lives. Sure, there is the occasional eccentric skeptic, like the finger-wiggling ancient philosopher described earlier. But if we persistently doubted everything, then we wouldn't eat when hungry, move from the path of speeding cars, or a thousand other things that we do during a typical day. We would hesitate and question everything, but never act. Radical skeptics have not been impressed with this argument either. According to Hume, we have natural beliefs that direct our normal behavior and override our skeptical doubts. As legitimate as radical skepticism is, nature doesn't give us the option to act on it. He makes this point here:


Most fortunately it happens, that since reason is incapable of dispelling these clouds [of skepticism], nature herself suffices to that purpose, and cures me of this philosophical melancholy. . . . I dine, I play a game of backgammon, I converse, and am merry with my friends; and when, after three or four hours' amusement, I would return to these speculations, they appear so cold, and strained, and ridiculous, that I cannot find in my heart to enter into them any further. [Treatise, 1.4.7]


Thus, according to Hume, we waver back and forth between skepticism and natural beliefs. When we realize how philosophically unjustified natural beliefs are, we are led down the path of skepticism. The doorbell then rings, and we're snapped out of our philosophical doubts and back to our normal routines and natural beliefs.

             A third attack on radical skepticism is that the skeptic's position is logically self-refuting. The skeptic's main point is this:



Let's call this "the skeptic's thesis." However, if I put forward the skeptic's thesis, then I am implying that I know it with certainty. It is like saying this:



The skeptic's thesis itself seems to be an exception to the very point that it is making. Thus, the skeptic's thesis is logically inconsistent with itself and we should reject it. But skeptics have a response to this criticism, which they sometimes explain using the metaphor of a digestive laxative. We take laxatives to rid our digestive system of unwanted stuff. But as the laxative takes effect, the laxative itself is expelled from the digestive system along with everything else. The skeptic's thesis, then, is like a laxative: we take it to rid our minds of all unjustified beliefs, and in the process we expel the skeptic's thesis itself. It is a higher level of skepticism in which we set aside everything, including even the skeptic’s thesis.

             A fourth criticism of radical skepticism is that it rests on an unrealistically high standard of evidence. There are two basic levels of evidence: complete and partial. The skeptic assumes that genuine knowledge requires complete evidence, but complete evidence is not achievable. Try as we might, says the skeptic, we can never prove any assertion with absolute certainty and some skeptical argument will cast doubt on that assertion. The solution is to this skeptical challenge is to reduce the qualifications for knowledge and be content with partial evidence. To illustrate, suppose I want to gather enough evidence to support the claim that "I know that there is a ball in front of me." I first get evidence through my senses: I perceive the ball with my eyes. I could then get supporting evidence by having others stand in front of the ball and report whether they see it too. I could get even stronger evidence by using scientific equipment that would measure the ball's density and detect the light spectrum reflected off the ball. This may seem extreme, but even then the evidence is still not complete. I could bring in a team of physicists to study the ball and write up an exhaustive report. I could hire a second team to do more tests. But even this is not complete since there are always more tests that I could run. Complete evidence is not possible, and the radical skeptic knows this. What, though, if we lower the requirements for what counts as knowledge? We could allow partial evidence, but not require the evidence to be complete. In the case of the ball, it might be enough to simply rely on the evidence that I gain about it through my senses, as incomplete as it is. This would put radical skepticism to rest. The problem with this solution, though, is that it doesn't refute radical skepticism, but surrenders to it. It concedes the impossibility of ever having genuine knowledge with absolute certainty. What we are left with is a version of knowledge that's so diluted that it doesn't count for much more than a personal conviction. After viewing the ball with my eyes, I may as well say "I have a partially supported belief that a ball is in front of me." Inserting the word "knowledge" here would add nothing.

             While radical skepticism seems excessive, it nevertheless poses a challenge to genuine knowledge that cannot be easily combated. It may well be impossible to ever refute radical skepticism, and so it might forever remain the archenemy of knowledge. While attempts to destroy this villain may ultimately fail, struggling with the issue helps illuminate the nature of knowledge itself. It's much like research into seemingly incurable diseases: even if scientists cannot discover a cure for cancer, the investigation still gives them a greater insight into human physiology. As we move on to explore the concept of knowledge in more detail, skepticism will always be lurking in the background, often forcing us to reject some theories and revise others.




We claim to know a lot of facts, for example, that fire is hot, that George Washington was the first U.S. president, and, in the case of Heaven's Gate believers, that superior aliens are roaming the galaxy. Our knowledge claims vary dramatically, and frequently we claim to know something that we really do not know. One way of understanding the concept of knowledge is to look at the different ways in which we acquire knowledge.

             Philosophers have traditionally maintained that there are two types of knowledge from two entirely different sources. First, there is knowledge through experience: seeing something, hearing about something, feeling something. This goes by the Latin term a posteriori which literally means knowledge that is posterior to, or after experience. Second, there is knowledge that does not come from experience, but perhaps instead is intuitively supplied from reason itself, such as logical and mathematical truths. This is called a priori knowledge, which, from Latin, literally means knowledge that is prior to experience.


Experiential Knowledge

Experiential (a posteriori) knowledge is of many types, four of which we will look at here. The first and most obvious kind of experiential knowledge is sensory perception. Each of our five senses is like a door to the outside world; when we throw them open, we are flooded with an endless variety of sights, sounds, textures, smells and tastes. When I look at a cow in front of me and say "I know that it is brown," the source of this knowledge rests upon my visual perception of the brown cow. While sensory perception is perhaps the dominant source of experiential knowledge, it immediately raises a critical question: when I look at a cow, do I perceive the actual cow itself, or just a mental copy of it that is processed through my visual system? A theory called direct realism holds that we see the actual cow itself. This is what we ordinarily assume when we open our eyes and perceive anything around us: I perceive the real table, chair, car, or whatever. We have a natural confidence in our senses, and we simply assume that what we see is what is actually there. But, according to a rival theory called indirect realism, this could not possibly be true, since there are countless instances where we know that the actual objects do not match what we perceive. An Object appears to get larger as I move closer to it, while the object itself clearly remains the same size. There are optical illusions, such as a stick which appears bent when in water; there are mirages, such as the appearance of water puddles on hot roads. According to indirect realism, then, I do not see the actual cow itself when I look at it, but only a copy or visual representation of it, almost as if I am viewing the cow on a television screen. There is, then, a big gap between what the cow itself and the image of it that appears to me, and for all I know the real cow may be vastly different than how it is represented. Thus, as much as we rely on sensory perception to gain knowledge, its main problem is that it is unreliable: uncertainty is built into the very act of perception.

             A second source of experiential knowledge is introspection, which involves directly experiencing our own mental states. Introspection is like a sixth sense that looks into the most intimate parts of our minds, which allows us to inspect how we are feeling and how our thoughts are operating. If I go to my doctor complaining of an aching back, she'll ask me to describe my pain. Through introspection I then might report, "Well, it’s a sharp pain that starts right here and stops right here." The doctor herself cannot directly experience what I do and must rely on my introspective description. Like sensory perception, the problem with introspection is that it is not always reliable. When surveying my mental states, I may easily misdescribe my feelings, such as mistaking a feeling of disappointment for a feeling of frustration. Other mental states seem to defy any clear descriptions at all, such as feelings of love or happiness.

             A third source of experiential knowledge is memory. My memory is like a recording device that captures events that I experience more or less in the order that they occur. I remember my trip to the doctor and the pain that I described to her at the time. This recollection itself constitutes a new experience. Again, the problem with experiential knowledge through memory is that it is not always reliable. For example, I might wrongly recollect that there's pizza in the refrigerator, completely forgetting that I ate it all last night. Also, sometimes overbearing people like police investigators can make us think that we remember something that never happened. And then there is the phenomenon of deja vu, the feeling that we've encountered something before when we really haven't.

             A fourth source of experiential knowledge is the testimony of other people. Take, for example, my knowledge that George Washington was the first U.S. president. Since Washington died centuries before I was born, I could not know this through direct sensory perception. Instead, I rely on the statements in history books. The authors of those books, in turn, rely on accounts from earlier records, and eventually it traces back to the direct experience of eyewitnesses who personally knew George Washington. A large portion of our knowledge rests on testimony, facts about people we've never seen our places we've never been to. While it is convenient for us to trust the testimony of others, the problem is that there is often a high likelihood of error. This is particularly so with word-of-mouth testimonies, where we're often sloppy in the accounts that we convey to others. Testimonies from written sources are usually more reliable than oral sources, but much depends on the integrity of the author, publisher, and the methods of fact-gathering. With oral or written sources, the longer the chain of testimony is, the greater the chance is of error creeping in.

             Sensory perception, introspection, memory, and testimony: these are the four main ways of acquiring knowledge through experience. Did we leave any out? There are a few contenders, one of which is extrasensory perception, or ESP for short. For example, you might telepathically access my mind and know what I am thinking. Or, through clairvoyance, you might be aware of an event taking place far away without seeing it or hearing about it. If ESP actually worked, we might indeed classify it among the other sources of experiential knowledge. But does it? Typical studies into ESP involve subjects guessing symbols on cards that are hidden from view. If the subject does better than a chance percentage, this is presumed to be evidence of ESP. However, the most scientifically rigorous experiments of this sort have failed to produce anything better than a chance percentage. While we regularly hear rumors of people having ESP, we have little reason to take them seriously. The safe route, then, would be to leave ESP off the list of sources of experiential knowledge.

             Consider next religious experiences. Believers sometimes say that they receive prophecies from God, or are guided by him, or know something through faith. Christian theologian John Calvin even spoke of a sense of the divine that we all have, which informs us that God exists. Might any of this count as experiential knowledge? The question is a complex one considering the wide range of religious experiences that believers report. Let's narrow the question to two representative types: knowledge through faith and prophetic knowledge. Regarding faith, as typically understood, faith involves belief without evidence, such as faith that God exists, or that the bodies of the dead will be resurrected in the future, or that our souls will be reincarnated in different bodies. These faith beliefs may be important for in our personal religious lives, but there is a problem when we to claim to know something through faith. One of the chief requirements for something to count as “knowledge” is that there is evidence to support it, as we will see more clearly in the next section. But since faith is belief without evidence, then, technically speaking faith would not qualify as knowledge. Prophetic knowledge faces the same challenge as ESP: are prophecies any more successful than educated guesses? Imagine an experiment that we might conduct in which half of the subjects were prophets, and the other half non-prophets. We then asked both groups to make predictions about the upcoming year; at the end of the year we then checked the results. How would the prophets do? The odds are slim that we could even conduct the experiment since prophets would say that they cannot prophesize on demand: it is a unique and unpredictable revelatory experience. They might also say that their revelations from God are not the sort of things that can be confirmed in the newspaper. If prophetic experiences are genuine sources of knowledge, the burden of proof seems to be on the believer. In the meantime, it would be premature to include it among the normal sources of experiential knowledge.


Non-Experiential Knowledge

Turning next to non-experiential (a priori) knowledge, this source of information is more difficult to describe. Some philosophers depict it as knowledge that flows from human reason itself, unpolluted by experience. We presumably gain access to this knowledge through rational insight. Usual examples of non-experiential knowledge are mathematics and logic. Take, for example, 2+2=4. Indeed, I might learn from experience that two apples plus two more apples will give me four apples. Nevertheless, I can grasp the concept itself without relying on any apples. I can also expand on the notion in ways that I could never experience, such as with the equation 2,000,000 + 2,000,000 = 4,000,000. Logic is similar; take for example the following argument:


All men are mortal.

Socrates is a man.

Therefore, Socrates is mortal.


When we strip this argument of all its empirical parts – men, morality, Socrates – the following structure is revealed:


All X are Y

Z is an X

Therefore Z is a Y


This logical structure is something that we know independently of experience. In addition to mathematics and logic, there are other truths that we know non-experientially, such as these:


·       All bachelors are unmarried men.

·       A sister is a female sibling.

·       Red is a color.


In each of the above cases, the truth depends entirely on the concepts within these statements. In the first, "unmarried men" is part of the definition of "bachelor"; the statement is thus true by definition, irrespective of our experiences.

             Traditional philosophers have argued that two concepts are central to non-experiential knowledge. First is necessity: non-experiential truths are necessary in that they could never be false, regardless of how differently the world was constructed. 2+2 would equal 4 in every conceivable science fiction scenario of the universe. Even if no human being ever existed, it would still be true that "All bachelors are unmarried men" based on the meaning of the words themselves. Experiential knowledge, though, is different in that it is contingent, as opposed to necessary: it could be false if the world had unfolded differently. Take the statement "George Washington was the first U.S. president," which is an item of experiential knowledge. It is of course true as things stand now. But we can imagine a thousand different things that might have prevented Washington from becoming president. What if he was sent to an orphanage for chopping down the family cherry tree? What if he choked to death on his wooden teeth prior to his inauguration? The truth of all experiential knowledge hinges on the precise construction of the world as it currently is.

             The other concept embedded in the notion of non-experiential knowledge is that of an analytic statement: a statement that becomes self-contradictory if we deny it. Take, for example, the statement "All bachelors are unmarried men." Its denial would be this:


It is not the case that all bachelors are unmarried men.


This is clearly self-contradictory since it would be like claiming that there exists some bachelor who is married, which is impossible. Denying mathematics or logic would also produce self-contradictory statements. Experiential knowledge, on the other hand, is synthetic rather than analytic: denying it will not produce a self-contradiction. Take again the statement "George Washington was the first U.S. president," which we know is true from experience. Its denial would be this:


It is not the case that George Washington was the first U.S. president.


While this statement is false as things actually stand, it is not self-contradictory since, if the world had unfolded differently, the U.S. might well have had a different first president.


Rationalism and Empiricism

An important philosophical war took place in the seventeenth and eighteenth-centuries between two schools of thought. First there were rationalists from continental Europe who were critical of sense experience and felt that genuine knowledge was acquired non-experientially through reason. The leaders on this side were René Descartes, Benedict Spinoza, and Gottfried Leibniz. Second there were empiricists from the British Isles who felt that non-experiential reasoning would give us nothing, and experience was the only path to knowledge. John Locke, George Berkeley and David Hume were the leaders on this side. The war finally ended when Immanuel Kant proposed a compromise: true knowledge depends on a mixture of experiential and non-experiential knowledge. We need both, Kant argued, otherwise our whole mental system will not operate properly.

             Let's look at this historical battle in more detail, starting with Descartes, who championed the rationalist position. Sense experience, Descartes argued, is seriously flawed and cannot be the source of important ideas that we have. Take, for example, the idea of a triangle. Look around the world and you'll never see a perfect triangle, whether it is a shape that we draw on a piece of paper or the side of a pyramid in Egypt. On close inspection, they will all have irregular lines. The fact remains, though, that we do have conceptions of perfectly-shaped triangles. Rationalism, according to Descartes, offers the best explanation of how we get those perfect ideas. There are two central components to the rationalist position: innate ideas and deductive reasoning. Innate ideas, according to Descartes, are concepts that we have from birth that serve as a foundation for all of our other ideas. While they are inborn, we only become aware of them later in life, when we reach the "age of reason" some philosophers called it. Innate ideas are in a special class of their own: we know them with absolute certainty, and it's impossible for us to acquire them through experience. While rationalists were reluctant to offer a complete list of innate ideas, the most important ones include the ideas of God, infinity, substance and causality. Regarding deductive reasoning, Descartes held that from our innate ideas we deduce other ideas. It is like in geometry where we begin with foundational concepts of points and lines, and deduce elaborate propositions from these about all kinds of geometrical shapes. Descartes was in fact inspired by the deductive method of geometry and maintained that we deduce ideas in the same way. Through deduction, the certainty that we have of innate ideas transfers to the other ideas that we derive from these. Mistakes creep in only when our deductions become so long that they rest on memory. All knowledge, he argued, including scientific knowledge, proceeds from innate ideas and deductive demonstration.

             Turn now to empiricism, particularly Locke's version. Locke's first task was to challenge the theory of innate ideas: none of our concepts, he argued, are inborn. Our mind is from birth like a blank sheet of paper, and it is only through experience that we write anything on it. One problem with innate ideas, he said, is that we can explain the origin of each one of them through experience. The idea of God, for example, is not innate as Descartes supposed, but comes from our perceptions of the world around us. There is thus no reason to put forward the theory of innate ideas when experience explains these notions just fine. Locke also found fault with the rationalist position that we do not become aware of innate ideas until later in life. It is not clear how such ideas can linger in our minds for so many years before we can be conscious of them. And by that time our minds have been flooded with experience, and a late-blooming innate idea wouldn't contribute anything to our knowledge of the world. Empiricists also challenged the rationalists' emphasis on deductive demonstration. We don't expand our knowledge by deducing new concepts from foundational ones, as mathematicians do. Geometry is the wrong model to follow. Instead, we acquire new knowledge through induction, such as making generalizations from our experiences. I hit ten light bulbs with a hammer and each breaks; I generalize from this that all similar light bulbs that I hit with a hammer will also break. We first perceive, then we generalize. We perceive some more, then generalize some more. That is how we push knowledge forward.

             Then Kant came along, the great mediator in the rationalism-empiricism debate. Kant was sympathetic with empiricism but thought that it suffered from a serious problem: it does not offer a good explanation for how we acquire non-experiential knowledge, such as mathematics and logic. Complex mathematical formulas in particular could not come from sense perception. There is a quality of self-evidence and certainty that they have, which fallible experience could never produce.  Kant's solution was not to resurrect the old theory of innate ideas. Instead, he argued that there are innate organizing structures in our minds that automatically systematize our raw experiences, sort of like a skeleton that gives shape to flesh. For example, as I watch someone hit a light bulb with a hammer, raw sensory information rushes in through my eyes. My mind immediately reconstructs this information into a three-dimensional image and puts it on a timeline. My mind then imposes other organizational schemes on the sensory information. It makes me see the hammer and light bulb as separate things, rather than just a single blob of stuff. It then makes me see the hammer as the cause of the light bulb breaking. My experience of the world, then, is a fusion of innate structures and raw experience. The innate structures are a concession to rationalism, and the experience part a concession to empiricism. 

             Rationalism and empiricism in their original forms are outdated theories today, in part because of Kant’s insights. Nevertheless, they still are useful for depicting two fundamentally different ways in which we assess the sources of knowledge. Rationalism will continue to be attractive whenever we have knowledge that cannot be easily explained by experience. Empiricism will be attractive whenever the claims of innateness are doubtful.




Throughout our discussion of knowledge so far, certain concepts have repeatedly appeared.  There is the question of the truth of a claim. There is also the matter of our personal belief conviction for a claim. There are also issues about the evidence or justification that we have for a claim. Tradition has it that these are the three key elements to knowledge: truth, belief and justification. For example, when I say "I know that Paris is the capital of France", this means


It is true that Paris is the capital of France.

I believe that Paris is the capital of France.

I am justified in believing that Paris is the capital of France.


For short, contemporary philosophers call this definition of knowledge justified true belief, often abbreviating it "JTB". The crucial point about this definition is that all three components must be present: if any one of the three is absent, then it does not count as knowledge.


Justified True Belief

Let's now examine each of the three elements of the JTB definition of knowledge. First is that a statement must be true if we claim to know it. For example, I can't "know" that Elvis Presley is alive if the truth is that he is dead. Knowledge goes beyond my personal feelings on the matter and involves the truth of things as they actually are. A critic of the JTB definition of knowledge might question whether truth is always necessary in our claim to know something. For example, the critic might argue, scientists in the middle ages claimed to know that the earth was flat based on the available evidence of the time. Even though we now understand that it is round, at the time they had knowledge of something that was false. Didn't they? In response to the critic, it may have been reasonable for scientists back then to believe the world was flat, but they really didn't know that it was. Their knowledge claims were premature in spite of how strong their convictions were. This is a trap that we fall into all the time. While talking with someone I may say insistently, "I know that Joe's car is blue". When it turns out that Joe's car is in fact red, I have to apologize for overstating my conviction. Truth, then, is an indispensable component of knowledge.

             Second, I must believe the statement in order to know it. For example, it's true that Elvis Presley is dead, and there is enormous evidence to back this up. But if I still believe that he is alive, I couldn't sincerely say that I know that he is dead. Part of the concept of knowledge involves our personal belief convictions about some fact. A critic of the JTB definition of knowledge might argue that belief is not always required for our claims to know something. For example, according to the critic, I might say "I know I'm growing old, but I don't believe it." In this case, I have knowledge of a particular fact without believing that fact. In response to the critic, if I utter the previous sentence, what I actually mean is that I'm not capable of imagining myself getting old or I haven't yet emotionally accepted that fact. I am just making my point more dramatically by saying "I don't believe it." Instead I really do believe it, but I don't like the thought of it.

             Third, I must be justified in believing the statement insofar as there must be good evidence in support of it. Suppose that I randomly pick a card out of a deck without seeing it. I believe it is the Queen of Hearts, and it actually is that card. In this case I couldn't claim to know that I've picked the Queen of Hearts; I've only made a lucky guess. To "know" that I've picked the Queen of Hearts, I would need some evidence for the claim. Again, a critic might question whether evidence is really needed for knowledge. For example, a store owner might say "I know that my employees are stealing from me, but I can't prove it!" Here the store owner has knowledge of a particular fact without any evidence for it. In response to the critic, the store owner is really saying that he strongly believes that his employees are stealing from him, but doesn't have enough evidence to press charges. Evidence, then, is indeed an integral part of knowledge.


The Gettier Problem

For centuries philosophers took it for granted that knowledge consists of justified true belief. In 1963 a young philosophy professor named Edmund Gettier published a three-page paper challenging this traditional view. He argued that there are some situations in which we have justified true belief, but which do not count as knowledge. This was dubbed "The Gettier Problem" and discussions of it quickly dominated philosophical accounts of knowledge. Gettier's actual illustrations of the problem are complex, but a simpler example makes the same point.

             Suppose that a ball in front of me appears to be red. First, I believe it is red. Second, I'm justified in this belief since that's how the ball appears to me. Third, it's also true that the ball is red. I thus have a justified true belief that the ball is red. However, it turns out that the ball is illuminated by a red light which casts a red tint over it, which is a fact that I'm unaware of. Although the ball in reality is red, under the light it would appear red to me even if the ball was white or some other color. Consequently, I can't claim to know that the ball is red even though I have a justified true belief that it is. I was fooled by the effects of the red light, but made a lucky guess anyway. Again, the point of this counterexample is to show that some instances of justified true belief do not count as genuine knowledge. This suggests that the traditional JTB definition of knowledge is seriously flawed.

             What can we do to rescue the JTB account of knowledge from Gettier's criticism? A common response is to add a stipulation to the definition of knowledge that would weed out counterexamples like the red ball. Most of the Gettier-type counterexamples involve a case of mistaken identity. In our current example, I mistake the appearance of a red-illuminated ball for an actual red ball. Perhaps, then, we can stipulate that knowledge is justified true belief except in cases of mistaken identity. More precisely, we can add a fourth condition to the definition of knowledge in this way:


I know that the ball is red when,

(1) It is true that the ball is red;

(2) I believe that the ball is red;

(3) I am justified in my belief that the ball is red;

(4) There is no additional fact that would make my belief unjustified (for example, a fact about a red light).


According to the above, my belief about the red ball would not count as knowledge since it would not pass the fourth condition. That is, there is indeed an additional fact regarding the red light that would make my initial belief about the ball unjustified. That additional fact undermines, or "defeats" my original justification. We've thus saved the JTB definition of knowledge, although cluttering it a little with a fourth condition. This strategy is called the no-defeater theory (also called the indefeasibility theory). A problem with this strategy, though, is that there are possible counter examples even to this, that is, situations in which we have undefeated justified true belief that don’t count as true knowledge. This, in fact, is a problem with most proposed solutions to the Gettier problem: if we get creative enough, we will likely find a new counter example that defies the solution.

             What does the Gettier problem mean for the concept of knowledge? In some ways, the problem seems like an overly picky point about the traditional definition of knowledge, and if it has taken centuries before someone discovered it, then it can't be that big of a deal. Skeptics, though, see it otherwise. For thousands of years skeptics have been telling us that knowledge is unachievable, and it should not surprise us now to learn that the concept of knowledge itself defies a simple and coherent definition. It may well be that nature has hardwired the concept of "knowledge" into our human thinking process, and we have no choice but to regularly apply that notion as we conduct our normal lives. It also well may be that the JTB definition is the specific conception of knowledge that nature has imposed on us. If so, nature has done this for the benefit of our own survival. Nevertheless, the skeptic argues, we should not mistake a natural survival hack for a pipeline to the truth. Nature takes shortcuts to assure that organisms operate efficiently. The concept of knowledge she has embedded within us is a shortcut that on the whole points us in the right direction with the limited evidence that we have. What we lack in evidence nature has supplemented with an attitude of insistence that we "know" things. The Gettier problem is a reminder that knowledge is as much about attitude as it is about our access to the reality of things.




Truth and justification, we have seen, are two of the key components of knowledge. They are also concepts that need some explanation themselves. Let’s first look at the notion of truth.


Theories of Truth

The concept of truth has many possible meanings. We talk about having true friends, owning a true work of art, or someone being a true genius. In all of these cases the word "true" means genuine or authentic. In philosophy, though, the notion of truth is restricted to statements or beliefs about the world, such as the statement that "My car is white" or "Paris is the capital of France". While we all have gut feelings about what it means for a statement to be true, philosophers have been particularly interested in arriving at a precise definition of truth. Here's one suggestion from a song:


"What is truth?" you ask and insist,

"Correspondence to things that exist?"

The answer, you fool, requires no sleuth:

Whatever I say is the truth.

Want proof of the truth? I say so! So there!

Purveyors of falsehood beware:

I'm sick of your lies, and, truth be told,

I am the truth, behold!


The above account of truth is clearly satirical since no one would seriously grant that the truth of all statements is grounded in the assertions of one individual person. But what are the more serious alternatives for definitions of truth? As usual in philosophy, there's much disagreement about what the correct definition is. We will consider the three leading candidates here.

             The first and most famous definition of truth is the correspondence theory: a statement is true if it corresponds to fact or reality. This is the most commonsensical way of looking at the notion of truth and is how standard dictionaries often define the concept. A true statement simply reflects the way things really are. Take the statement "My car is white.” This statement is true if it conforms to how the world actually is, specifically whether my car is in fact painted white. As compelling as the correspondence theory of truth seems, skeptics immediately see one major flaw with it: we do not have access to the world of facts. In spite of my best efforts to discover the way things really are, I’m at the mercy of my five senses, which, we have seen, are unreliable. While my senses tell me that my car is white, the color receptors in my eyes may not be working properly and my car may be a shade of yellow. For that matter, I may be living in a world of hallucinations and do not even own a car. The sad fact is I can never reach beyond my perceptions and see the world as it really is.

             With trivial issues, such as the truth concerning the color of my car, I may be willing to simply pretend that I have direct access to the world of facts and blindly trust my senses. This may serve my immediate needs perfectly well. It is not so easy to pretend, though, when I investigate the truth of more serious statements, such as whether “Bill murdered Charlie.” Even if I have a mountain of evidence that implicates Bill, such as fingerprints and eyewitness testimony, it’s impossible for me to turn back the hands of time and directly access the scene of Charlie’s murder. I only have hints about what the reality is. Similarly, if I’m investigating the truth of the statement “God exists,” I cannot directly access the reality of an infinitely powerful deity, even if God did exist and stood right in front of me. The best I would have is some imperfect evidence that the mysterious being standing before me was indeed God. Thus, the correspondence theory would not permit us to say either that “It is true that Bill murdered Charlie” or “It is true that God exists.”

             A second famous definition of truth is the coherence theory, which aims to address the shortcomings of the correspondence theory. According to the coherence theory, a statement is true if it coheres with a larger set of beliefs. Rather than attempting to match up our statements with the actual world of facts, we instead try to see if our statements mesh with a larger web of beliefs that support them. For example, the statement “my car is white” is true if it coheres with a collection of other beliefs such as “many cars are painted white,” “I perceive that my car is white,” and “other people invariably report that my car is white.” With the coherence theory, we avoid skeptical obstacles such as the unreliability of our senses and the possibility that we are hallucinating. What matters is our web of beliefs, which we all have access to, in contrast with a hidden world of facts that is blurred by the limits of our sensory perceptions. We also can even investigate statements such as “It is true that Bill murdered Charlie” or “It is true that God exists.” What matters here is whether these statements consistently fit with other beliefs that we have, beliefs about the pieces of evidence against Bill, and beliefs about the evidence regarding a divine being.

             Unfortunately the coherence theory faces criticisms, the most important of which is that it is relativistic. That is, it grounds truth in the changeable beliefs of human beings, rather than in an unchanging external reality. According to the coherence theory, the standard for all truth is the larger web of beliefs that people hold, beliefs about white cars, criminal evidence, evidence for God’s existence, and countless other issues. The problem is that belief systems come and go.  Take beliefs about criminal evidence as just one example. Many cultures throughout history based criminal convictions on the evidence of supernatural omens: prophetic visions, the flight path of birds, patterns in the guts of sacrificed animals. That was their belief system which they relied on. In other cultures, the testimony of one eye witness is sufficient to prove guilt. In our culture today we have fingerprints, DNA samples and psychological profiles which all contribute to our belief system about criminal guilt. The statement “Bill murdered Charlie” could cohere with some belief systems, but not with others. We typically think about truth as being absolute: either Bill murdered Charlie or he didn’t. But if truth hinges on a changeable belief system, then truth is no longer absolute, but relative to some particular belief system. This type of relativism is precisely what the skeptic's position is on truth inquiry. Thus, if you don't want to travel down the skeptic's path, you will not be happy with the coherence theory of truth.

             The problems with the correspondence and coherence theories are so serious that many contemporary philosophers have abandoned both. In fact some philosophers have even abandoned the concept of “truth” as being completely unnecessary. This brings us to our third theory, the deflationary theory of truth: to assert that a statement is true is just to assert the statement itself. Compare these two statements:



What is the difference between the two? Nothing of substance. The phrase “it is true that” seems to be just repeating something that is already assumed in the phrase “my car is white.” In that sense, I am being redundant if I use the phrase “it is true that.” At times it may be rhetorically helpful to use the phrase “it is true that” in an effort to convince someone of my belief. Suppose you say to me “I don’t believe that your car is white.” I might respond by saying, “You’re wrong: it’s absolutely true that my car is white”. Again, I’ve not added anything of substance by injecting the notion of truth into my response; I’ve just stood up to you more forcefully. In short, according to the deflationary theory, the quest for a clear conception of truth, such as correspondence or coherence, will not succeed because it is ultimately a quest for something that doesn’t really exist.

             But the deflationary theory also faces problems, one of which is that the notion of truth is built into our normal expectations of what we assert. When I say that “my car is white” you have an expectation that what I’m saying is true. Occasionally I do say something that is false, but when that happens, we all recognize that I am doing something that is incorrect. The normal expectation, then, is that my assertion will be truthful. This creates a problem for the deflationary theory: by eliminating the notion of truth, it cannot adequately account for our normal expectation of truthfulness.


Theories of Justification

Of the three components of knowledge, justification is the one that has attracted the most attention among contemporary philosophers. For centuries most philosophers followed a theory of justification called foundationalism. On this view, our justified beliefs are arranged like bricks in a wall, with the lower ones supporting the upper ones. These lowest bricks are called “basic beliefs”, and the ones they support are “non-basic” beliefs. Take this example:


• My car is white (non-basic belief)


This belief rests upon some supporting ground-level basic beliefs, including these:


• I recognize the car in front of me as my car (basic belief)

• I remember what white things look like (basic belief)

• The car in front of me looks white (basic belief)


There are two distinct elements to this foundationalist theory of justification. First, our ground-level basic beliefs are self-evident, or self-justifying, and thus require no further justification. When we have such beliefs, we cannot be mistaken about them, we cannot doubt them, and we cannot be corrected in our beliefs about them. For example, if I am perceiving the color white, then my belief that I am perceiving white is self-evident in this way. Even if I am hallucinating at the moment, my belief that I am perceiving the color white cannot be called into question. The second element of foundationalism is that justification transfers up from my foundational basic beliefs to those non-basic beliefs that rest upon them. Think of it like the mortar between bricks that begins at the very bottom level, locks them solid, and moves upwards to lock the higher bricks into place. For example, if I have the three basic beliefs about my car and whiteness listed above, then I am justified in inferring the non-basic belief that “my car is white.”

             While foundationalism holds a respected place in the history of philosophy, it faces a major problem: it is not clear that there really are any self-evident basic beliefs that form the foundation of other beliefs. Foundationalists themselves have mixed views about what exactly our lowest-level foundational beliefs are. Descartes, for example, argued that there is only one single brick at the foundation of my wall of beliefs, namely, my belief that I exist. Every other belief I have rests on this. Locke, on the other hand, held that our most foundational beliefs are simple perceptions such as blue, round, sweet, smooth, pleasure, motion. These combine together to make more complex ideas. Contemporary philosophers resist both Descartes’ and Locke’s depiction of our most foundational beliefs. Some offer examples such as “I see a rock” (a basic belief about one’s perception), “I ate cornflakes this morning” (a basic belief about one’s memory), or “That person is happy” (a basic belief about another person’s mental state). But even these are questionable since they seem to rely on beliefs or perceptions that are more ground-level. If there really are ground-level foundational beliefs that are self-evident or self-justifying, you would think that philosophers would have agreed long ago about exactly which ones they are. But there is no such agreement.

             An alternative to foundationalism is coherentism: justification is structured like a web where the strength of any given area depends on the strength of the surrounding areas. Thus, my belief that my car is white is justified by a web of related beliefs, such as these:


• I recognize the car in front of me as my car

• I remember what white things look like

• The car in front of me looks white


These, though are not foundational, but instead depend on another web of beliefs related to them, which includes these:


I remember purchasing my car

People seem to agree that I use the term “white” properly

Nothing is abnormally coloring my vision, such as a pair of sun glasses


Each of these, in turn, rests on an ever-widening web of related beliefs. At no point do we reach a bottom-level foundation to these beliefs; the justification of each belief rests on the support it receives from the surrounding web of beliefs that relates to it. Coherentism is closely associated with the coherence theory of truth. With truth we determine that a proposition is true if it coheres with a larger web of beliefs. With justification, we determine that a belief is justified if it is supported by a larger web of beliefs. To use another metaphor, it is similar to how each entry in the dictionary consists of words that the dictionary also defines. It is a self-contained system of definitions that is not reliant on foundational notions outside of itself. Coherentism’s similarities with the coherence theory of truth make it vulnerable to the same fundamental charge of relativism: not everyone’s belief system is the same, so a particular belief might find justification within your larger web of beliefs, but not within mine. Your belief system might justify the belief that “Bill killed Charlie,” that “God exists,” or that “abortion is immoral,” while my belief system might not justify any of these. By embracing coherentism, then, we are accepting yet another version of skeptical relativism. We would like to think that justification is a bit more universal and not dependent on the peculiarities of a particular person’s belief system.

             Given the liabilities of both foundationalism and coherentism, many contemporary philosophers hold a third position called reliabilism: justified beliefs are those that are the result of a reliable process, such as a reliable memory process or a reliable perception process. It is sometimes called an “external” theory of justification since it depends on factors external to the believer’s understanding. It is like how we depend on a reliable clock to tell us what time it is. As long as we have confidence in the clock mechanism itself, then that’s all we need in order to trust the time that it tells us. We don’t have to inspect the internal gears of the clock and see how they relate to the movement of the clock’s hands. Similarly, to justify my beliefs, I don’t need to inspect how each belief connects with surrounding beliefs that are beneath them or next to them; I just trust the reliability of my mental process that gives me the belief. If my memory process is on the whole reliable, then I’m justified in my belief that I ate cornflakes this morning for breakfast. If my perceptual process is on the whole reliable, then I’m justified in my belief that my car is white. By contrast, unreliable processes might be if I believed my car was white because of mere wishful thinking, or through ESP, or from indoctrination by the Society for the Advancement of White Cars. What matters is the reliability of the larger processes upon which my beliefs rest, not my other beliefs that border them. According to reliabilism, the fault with both foundationalism and coherentism is that they rely too much on introspection: presumably, with our mind’s eye, we can see the strength of our specific beliefs and how they gain support from other beliefs that are connected to them (either like bricks in a wall or strands in a web). But, says the reliabilist, this approach places too much confidence in our ability to internally witness the connections between our specific beliefs. Introspection, we have seen, is notoriously unreliable, and our standards of justification should not depend on what our mysterious mind’s eye internally perceives, but, instead, upon more external standards and mental processes that we know are reliable through our life experiences. I am justified in believing that I ate cornflakes for breakfast because that is what I remember, and I trust my memory since it is a reliable process of supplying me with information about the past.

             Reliabilism is an appealing theory since it dispenses with the untrustworthy mechanism of introspection and has us place our confidence in our normally reliable mental processes. In ordinary situations, such as justifying that my car is white, reliabilism may work fine, since I can just rely on the trustworthiness of my normal perceptual process. But there is a problem with reliabilism in extraordinary situations. If I am an eye witness to a crime, how can I tell for sure what I really saw? If I am having a mental health examination, how can I tell whether my behavior is rational or irrational? If I am documenting an historical event in my life, how can I tell whether it is accurate? If I am engaged in a theological or political debate, how can I tell whether my convictions are valid? Simple appeal to the reliability of my mental processes may not be good enough. In these situations what is at issue is the very structure of my belief system, and I am forced to investigate how my convictions rest upon other beliefs, and those upon still more beliefs. It thus may not be that easy to set aside either foundationalist or coherentist approaches to justification.


What’s so Bad about Relativism?

Twice so far the issue of relativism has raised its ugly head, and how we assess theories of truth and justification hinges greatly on how we feel about relativism. The relativist position in general is that knowledge is always dependent upon some particular conceptual framework (that is, a web of beliefs), and that framework is not uniquely privileged over rival frameworks. The most famous classical statement of relativism was articulated by the Greek philosopher Protagoras (c. 490–c. 420 BCE), who said that “Man is the measure of all things.” His point was that human beings are the standard of all truths, and it is a futile task to search for fixed standards of knowledge beyond our various and ever-flexible conceptual frameworks. Knowledge in medieval England depended on the conceptual framework of that place and time. Knowledge for us today depends on our specific conceptual frameworks throughout the world and throughout our wide variety of social environments.

             Our initial reactions to relativism are usually negative. “The truth is the truth,” I might say, “and it shouldn’t make any difference what my individual conceptual framework is. Some conceptual frameworks are simply wrong, and others may be a little closer to the truth.” But is relativism really so bad that it warrants this negative reaction?

             The first step to answering this question is to recognize that there are different types of relativism, some of which may be less sinister than others. The most innocent and universally accepted type is etiquette relativism, the view that correct standards of protocol and good manners depend on one’s culture. When I meet people for the first time, should I bow to them or shake hands? If I make the wrong decision, I might offend that person, rather than befriend them. Clearly, that depends on the social environment that you are in, and it makes no sense to seek for an absolute standard that applies in all situations. Etiquette by its very nature is relative. There is also little controversy regarding aesthetic relativism, the view that artistic judgments depend on the conceptual framework of the viewer. We commonly feel that there is no absolute right and wrong when it comes to art, and it’s largely a matter of opinion. I might enjoy velvet paintings of dogs playing cards, while that might offend your aesthetic sensibilities. In many cases, perceptual relativism is also no big issue: one’s sensory perceptions depend on the perceiver. Something might appear red to me but green to you. There are people known as “supertasters” who experience flavors with far greater intensity than the average person, so much so that they need to restrict themselves to food that you or I would find bland. How we perceive sensations depends on our physiology, which we readily acknowledge may differ from person to person.

             The types of relativism that we often resist, though, are those connected specifically with the two components of knowledge that we’ve discussed above, namely, truth and justification. These are the ones most associated with skepticism. Truth relativism is the view that truth depends upon one’s conceptual framework. This amounts to a denial of the correspondence theory of truth and acknowledges our inability to access an objective and independent reality. Justification relativism is the view that what counts as evidence for our beliefs depends upon one’s conceptual framework. This is a denial of foundationalism and an acknowledgement of coherentism. The German philosopher Friedrich Nietzsche (1844-1900) boldly embraced truth and justification relativism, as we see here:


Positivism stops at phenomena and says, “These are only facts and nothing more.” In opposition to this I would say: No, facts are precisely what is lacking, all that exists consists of interpretations. We cannot establish any fact “in itself”: it may even be nonsense to desire to do such a thing. . . .  To the extent to which knowledge has any sense at all, the world is knowable: but it may be interpreted differently, it has not one sense behind it, but hundreds of senses. “Perspectivity.”  [Will to Power, 481]


For Nietzsche, then, there are many perspectives from which the world can be interpreted when we make judgments. Some justification relativists even go so far as to deny the universal nature of so-called laws of logic; even these, they maintain, are grounded in mere social conventions.

             A standard criticism of truth and justification relativism is that it leads to absurd consequences that no rational person would accept. By surrendering to relativism, we abandon any stable notion of reality and place ourselves at the mercy of cultural biases, fanatical social groups, and power-hungry tyrants who are more than happy to twist our conceptual frameworks to their benefits. Everything, then, becomes a matter of customs that are imposed on us, even in matters of science. Scottish philosopher James Beattie (1768–1790) makes this point in a fictional story where he describes a crazy scientist who attempts to put relativism into practice:


[The scientist] was watching a hencoop full of chickens, and feeding them with various kinds of food, in order, as he told me “that they might [give birth to live offspring and] … lay no more eggs,” which seemed to him to be a very bad custom. . . . “I have also,” continued he, “under my care some young children, whom I am teaching to believe that two and two are equal to six, and a whole less than one of its parts; that ingratitude is a virtue, and honesty a vice; that a rose is one of the ugliest, and a toad one of the most beautiful objects in nature.” [James Beattie, “The Castle of Skepticism”]


According to Beattie, if we took the relativist’s position seriously, we’d be forced to accept absurd views like “it is just a matter of custom that chickens lay eggs,” or that “it’s possible that 2+2=6.” Thus, even if we acknowledge a certain level of relativism with etiquette, aesthetics and perception, we need to draw the line when it comes to standards of truth and justification.

             How might the relativist respond to this criticism? One approach is to hold that not all conceptual schemes are on equal footing, and some indeed are better than others. Nietzsche argues that there are competing perspectives of the world, and the winner is the one whose conceptual framework succeeds the best:


It is our needs that interpret the world; our instincts and their impulses for and against. Every instinct is a sort of thirst for power; each has its point of view, which it would gladly impose upon all the other instincts as their norm. [Will to Power, 481]


Nietzsche presents the conflict as a kind of power struggle among competing conceptual frameworks, where the winner takes all. A more gentle approach, though, would be to hold that the winner is the one that best assists us in our life’s activities and allows us to thrive. If people today held that “it is just a matter of custom that chickens lay eggs,” or “it is possible that 2+2=6”, their underlying conceptual framework would not enable them to succeed very well in the world. For that matter, such a conceptual framework would not have allowed people to thrive very well in medieval England or any other pre-modern period of human history. While there may be an underlying objective reality that molds our conceptual frameworks in successful ways, that possibility is irrelevant since, according to the relativist, we could never know such an objective reality even if it existed. What we do know is how our conceptual frameworks enable us to succeed in the world, and that is the real litmus test for truth and justification.

             Thus, with many ordinary life beliefs, relativist theories of truth and justification work reasonably well, without leading us down the path to absurd consequences. What, though, of more scientific theories? In medieval times people thought mental illness was caused by demon possession; today we think that it is caused by physiological brain disorders. The medieval theory worked well in its own day. Does that mean that it was true back then, supported by a medieval web of beliefs, but not now? In scientific matters, people feel uncomfortable with relativism and instead believe that our knowledge of physics, chemistry and biology has a fixed and objective reference point. We will next examine the issue of scientific knowledge in more detail.




Every child knows the tale of Isaac Newton’s inspiration for his views on gravity. While sitting beneath a tree he saw an apple fall, which prompted him to wonder why things always fell downward rather than sideways or upward. In time Newton formulated his theory of universal gravitation, which described the attraction between massive bodies. Less known is the rival theory of intelligent falling, devised by the satirical newspaper The Onion. According to this view, things fall downward “because a higher intelligence, 'God' if you will, is pushing them down.” As proof for their view they cite a passage from the Old Testament book of Job: “But mankind is born to trouble, as surely as sparks fly upwards.” Accordingly, a defender of intelligent falling theory remarks, “If gravity is pulling everything down, why do the sparks fly upwards with great surety? This clearly indicates that a conscious intelligence governs all falling.” The theory of intelligent falling is obviously not a real theory, but rather a parody of the religiously-based intelligent design theory. Nevertheless, we can ask the basic theoretical question, why is universal gravitation a better account of natural events than intelligent falling? The task of science is to explain how the natural world works, to give us knowledge of the underlying mechanics of natural phenomena. That knowledge does not come easy, though, and it seems that science has to wrench nature’s secrets out of her. As scientists put forward rival theories, how do we determine which are closer to the truth?


Confirming Theories

The starting point for discussion is to distinguish between three related scientific concepts: a hypothesis, a theory, and a law. The weakest of these is the scientific hypothesis, which is any proposed explanation of a natural event. It is a provisional notion whose worth requires evaluation. Newton’s account of gravity began as a humble hypothesis, and even the theory of intelligent falling qualifies as a hypothesis. While hypotheses may be inspired by natural observations, they don’t need to be, and virtually anything goes at this level. One step up from this is a scientific theory, which is a well confirmed hypothesis. It is not mere guess, like a hypothesis may be, but is a contention supported by experimental evidence. When Newton proposed his account of gravity, he accompanied it with a wealth of observational evidence, which quickly elevated it to the status of a theory. This, though, is where the theories of gravity and intelligent falling part company: there’s no scientific evidence in support of intelligent falling, and thus it fails as a theory. Lastly, there is a scientific law which is a theory that has a great amount of evidence in its support. Indeed, laws are confirmed by such a strong history of evidence that they cannot be overturned by any individual piece of evidence to the contrary; rather, we assume instead that that single piece of contrary evidence is flawed. As compelling as Newton’s theory of gravity was, it took well over 100 years before it was confirmed to the point that it gained status as a law.

             We see that confirmation is the critical component in establishing a scientific claim: it is what elevates a hypothesis to a theory, and a theory to a law. The process of scientific confirmation is a complex one that typically involves evaluating observational data within a statistical tool that measures its probability. Our concern here, though, is with general rules of scientific methodology that help confirm one hypothesis over a potential rival. The first such general rule of scientific confirmation is that the notion exhibits simplicity; that is, when evaluating two rival theories, the simplest theory is the one most likely to be true. This doesn’t guarantee that it’s true, but, all things being equal, it’s the one that we should prefer. Compare, for example, universal gravity and intelligent falling. Universal gravity involves a single gravitational force that is inherent to all physical bodies. Intelligent falling, on the other hand, involves countless divine actions that guide individual bodies downwards. We should thus prefer universal gravity as the correct explanation since it is not burdened by such an abundance of distinct divine actions.

             A second rule of scientific confirmation is unification, that is, the ability to explain a wide range of phenomena. The rule of thumb here is that the more information explained by a theory, the better. Science is an immense interrelated system of facts, laws, and theories, and scientific contentions gain extra weight when they contribute to the scheme of unification. It is unification that gave an initial boost to Newton’s theory of universal gravitation. Prior to Newton, Astronomers assumed that planets and other celestial objects followed their own unique set of laws that were distinct from those on earth. However, Newton showed how the motions of the planets were governed by precisely the same rules of gravity and motion that physical bodies on earth obey.

             A third rule of scientific confirmation is successful prediction. Good scientific theories should not simply organize collections of facts, but should be able to reach out and predict new phenomena. This is what bumped Newton’s theory of gravity up to the status of a law. Astronomers in the early nineteenth-century noticed strange movements in the orbital pattern of the planet Uranus, and they hypothesized that the irregularities were caused by the gravitational tugging of an undiscovered eighth planet. Applying Newton’s formulas of gravity and motion, they pinpointed a location in space where the large body must be. Then, pointing their telescopes at the spot, they discovered the mystery planet, which was subsequently named “Neptune.” Major scientific predictions like these do not happen too often, but when they do they do much to confirm a theory. Einstein's theory of relativity, for example, was confirmed with the prediction of bent star light during a solar eclipse.

             A fourth rule of scientific confirmation is falsifiability: it must be theoretically possible for a scientific claim to be shown false by an observation or a physical experiment. This doesn’t mean that the scientific claim is actually false, but only that it is capable of being disproved. The criterion of falsifiability is important for distinguishing between genuine scientific claims that rest on tests and experimentation, and pseudo-scientific claims that are completely disconnected with testing. Take, for example, the views of Heaven’s Gate believers that we examined at the outset of this chapter. According to them, aliens come down to earth in the form of teachers, but once becoming human they are stripped of their previous memories and knowledge. Thus, we can’t test the claims of these teachers about their previous alien lives, since they can’t remember anything about them. “So tell me a little about your home planet” I might ask one alleged alien. He replies “Sorry, I can’t remember anything about it, but I’m still an Alien.” To make things worse, Heaven’s Gate believers claim that the aliens purposefully imposed this knowledge restriction on themselves since "too much knowledge too soon could potentially be an interference and liability to their plan." In short, their claims about the aliens are completely resistant to refutation. Fortune tellers are another example of this, as the philosopher Karl Popper explains here:


[B]y making their interpretations and prophesies sufficiently vague they were able to explain away anything that might have been a refutation of the theory had the theory and the prophesies been more precise. In order to escape falsification they destroyed the testability of their theory. It is a typical soothsayer’s trick to predict things so vaguely that the predictions can hardly fail: that they become irrefutable. [Conjectures and Refutations]


A common prediction made by fortune tellers is that “a major event will happen in your life” or “you will soon make a very important acquaintance”. Both of these so-called “predictions” will take place no matter what happens in your life, and no experience you have will refute those predictions. Legitimate scientific theories, by contrast, always hold open the possibility of being refuted by new data or a new experiment. By putting forth their theories, scientists take a risk that what they’re proposing might be disproved by the facts. Even universal gravitation is vulnerable to refutation if some future experiments produce compelling evidence against it. Thus, a good theory is always potentially falsifiable, although it has not been actually falsified.


Scientific Revolutions

Scientists continually push the boundaries of knowledge, and on a daily basis we see new theories about the spread of diseases, healthy eating habits, or the environment. We also read about new studies that challenge previously accepted scientific views. For example, in contrast to earlier claims by scientists, the accepted wisdom now is that vitamin C does not help prevent colds, and fiber in our diets does not help prevent colon cancer. Science thus moves ahead in baby steps, occasionally taking a step backwards to correct an erroneous theory. All the while, though, the larger body of scientific knowledge seems secure and well established. But then sometimes a new scientific theory comes along that is so radical and far reaching in its consequences that it forces scientists to throw out many of their underlying assumptions about the world and set things on a dramatically new course. These are scientific revolutions. The most dramatic example is the shift from the earth-centered view of the heavens, championed by the ancient Greek astronomer Ptolemy, to a sun-centered system which was defended by Copernicus in the 1500s. This “Copernican Revolution,” as it is now called, did more than simply swap the earth with the sun in the model of celestial objects. It also had the effect of overturning medieval theories about matter and motion, and ultimately replacing them with Newton’s laws of motion. Other important revolutions were sparked by Darwin’s account of evolution, Einstein’s account of general relativity, genetics, and tectonic plates.

             The most probing philosophical analysis of scientific revolutions was offered by American historian of science Thomas Kuhn (1922-1996). Kuhn argued that scientific revolutions are the result of changing paradigms, that is, the web of scientific beliefs held in common by members of the scientific community. When major paradigms are overthrown and replaced with new ones, such as replacing the Ptolemaic with the Copernican paradigm, we have a scientific revolution. What triggers the paradigm shift, according to Kuhn, is that scientists run into inconsistencies with the old paradigm that cannot easily be explained away. Scientific theories will always face some irregularities, such as with an experiment that seems to contradict an accepted theory. If the theory is well established, a few irregularities here or there won’t matter; in fact, scientists often chalk these up to an acceptable level of error that’s built into the enterprise of scientific investigation. However, sometimes irregularities pile up to such a degree that it throws science into a condition of crisis. Seeking resolution to the crisis, scientists then replace an old scientific paradigm with a new one that better resolves the irregularities.

             Kuhn argues that scientific revolutions have much in common with political revolutions: rebel groups think that the ruling institution is inadequate, which they then overthrow and replace with a new one:


Political revolutions are inaugurated by a growing sense, often restricted to a segment of the political community, that existing institutions have ceased adequately to meet the problems posed by an environment that they have in part created. In much the same way, scientific revolutions are inaugurated by a growing sense, again often restricted to a narrow subdivision of the scientific community, that an existing paradigm has ceased to function adequately in the exploration of an aspect of nature to which that paradigm itself had previously led the way. In both political and scientific development the sense of malfunction that can lead to crisis is prerequisite to revolution.


Kuhn warns that the transition from the old to new paradigm is not a smooth one. Many scientists will hold fast to the old paradigm, and the new paradigm needs to attract an ever-growing number of supporters before it can finally overthrow the old one. The upshot of Kuhn’s theory is that science is not cumulative: our present theories are not built upon the secure foundation of past theories. Instead, scientific knowledge shifts according to our current paradigms, and, once again, the issue of relativism arises. These paradigms are webs of belief that are held by the scientific community at the time. Truths in science, then, are relative to these shifting webs of belief.

             Kuhn’s account of scientific revolutions has its critics, particularly among those who believe that science, when done properly, is grounded in objective truth, and not in shifting belief paradigms of the scientific community. One criticism is that Kuhn has over dramatized the sweeping nature of most scientific revolutions. Sure, the Copernican revolution was indeed a major one that resulted in overthrowing old scientific models that were rooted in superstitious conceptions of the world and sloppy experimentation. In fact, the older models were so engrained with religious mythology and metaphysics, it’s overly generous to even call them “scientific.” Since the time of Copernicus, however, we’ve not seen any scientific revolutions that “overthrow” entire paradigms. Rather, new mini-revolutions seek to encompass much of the theory and data of previous scientific investigations while at the same time setting a new direction for future investigation. For example, Newton’s laws of motion were not overthrown by Einstein’s theory of relativity; instead scientists try to incorporate both into a larger scientific vision of reality that unifies all of nature’s forces. Other mini-revolutions, such as Darwinian evolution, quickly put an end to rival theories of biological development, such as Lamarck’s, that had little or no supporting evidence to begin with. Thus, contrary to Kuhn’s position, when science is done properly, our knowledge of the natural world is cemented into a fixed and objective reference point. The contest between Kuhn and his critic, then, ultimately rests on how we view the various revolutions that take place within science. It is a question of whether they dramatically overturn previous scientific conceptions, or just contribute to an ever-accumulating body of fixed scientific knowledge.

             This chapter began with a discussion of radical skepticism, and while that may not be the most cheery way of investigating the nature of knowledge, in many ways it sets the right tone. No matter what we say to clarify the characteristics of knowledge, warning flags immediately go up. All our sources of knowledge have serious limitations. The very definition of knowledge can be picked apart by an endless variety of Gettier-type counter examples. Theories of truth and justification seem to be either naively optimistic, or they lean towards relativism. While scientific knowledge attempts to move progressively towards unchanging truth, it is always cradled by a potentially changing web of beliefs held by scientists. Achieving genuine knowledge is in some ways like playing a video game where the winning score is infinitely high: no matter how close we move towards it, it remains at a distance. If the human effort to gain knowledge was merely a leisure activity like playing an impossible game, we’d certainly give it up for a more attainable diversion. But the pursuit of knowledge is a matter of human survival that we can’t casually set aside. Philosophical discussions of knowledge are an important reality check as we routinely gather facts and construct theories about how the world operates. The hope of acquiring a fixed body of knowledge is very seductive, and the problems of knowledge that we’ve covered in this chapter help us resist that temptation.




Information about the Heaven’s Gate cult can be found at

Sextus Empiricus, Outlines of Pyrrhonism (c. 200 CE). A recent translation is by J. Annas and J. Barnes (Cambridge: Cambridge University Press, 1994).

Hume, David, Treatise of Human Nature (1739-1740). The standard edition is by David Fate Norton, Mary J. Norton (Oxford: Clarendon Press, 2000).

Descartes, René, Meditations (1641). A recent translation by J. Cottingham is in The Philosophical Writings of Descartes (Cambridge: Cambridge University Press, 1984).

Locke, John, Essay Concerning Human Understanding (1689). The standard edition is by P.H. Nidditch (Oxford: Oxford University Press, 1975).

Kant, Immanuel, Critique of Pure Reason (1781). A recent translation of this is by P. Gruyer and A.W. Wood, Cambridge: Cambridge University Press, 1998.

Gettier, Edmund, “Is Justified True Belief Knowledge?” Analysis (1963), Vol. 23, pp. 121-123.

Nietzsche, Friedrich, Will to Power (1906). A standard translation of this is by Walter Kaufmann, (New York: Viking, 1967).

Beattie, James, “The Castle of Skepticism” (1767). Manuscript transcribed in James Fieser, ed., Early Responses to Hume’s Life and Reputation (Bristol: Thoemmes Press, 2004).

"Evangelical Scientists Refute Gravity with New 'Intelligent Falling' Theory," The Onion, August 17, 2005, Issue 41.33.

Popper, Karl, Conjectures and Refutations (London: Routledge, 1963).

Kuhn, Thomas, The Structure of Scientific Revolutions (Chicago: University of Chicago Press, 1962), Ch. 9.




Please answer all 21 of the following questions.


1. Define procedural knowledge, propositional knowledge, philosophical skepticism, limited skepticism and radical skepticism.

2. Explain Pyrrhonian, Humean and Cartesian radical skepticism.

3. What are the four main criticisms of radical skepticism?

4. What are the four main sources of experiential knowledge?

5. What is the problem with ESP as a reliable source of knowledge?

6. What are the key features of non-experiential knowledge?

7. What are the key features of rationalism and empiricism?

8. What is Kant’s solution to the rationalism vs. empiricism debate?

9. Describe the "JTB" definition of knowledge and the point of the Gettier problem.

10. What is the correspondence theory of truth and what is the criticism of it?

11. What is the coherence theory of truth and what is the criticism of it?

12. What is the deflationary theory of truth and what is the criticism of it?

13. What is the foundationalist theory of justification and what is the criticism of it?

14. What is the coherence theory of justification and what is the criticism of it?

15. What is the reliabilist theory of justification and what is the criticism of it?

16. Define etiquette relativism, aesthetic relativism and perceptual relativism.

17. Define truth relativism and justification relativism.

18. Define scientific hypothesis, scientific theory and scientific law.

19. Explain simplicity, unification, successful prediction, and falsifiability.

20. What is Kuhn’s view of scientific revolutions and paradigm shifts?

[Short Essay]

21. Short essay: pick any one of the following views in this chapter and criticize it in a minimum of 150 words. Pyrrhonian, Humean or Cartesian radical skepticism; the Getter problem or the no-defeater solution; the correspondence, coherence, or deflationary theory of truth or one of their criticisms; the foundationalist, coherentist, or reliabilist theory of justification or one of its criticisms; Beattie’s criticism of truth and justification relativism, or Nietzsche’s defense of them; scientific confirmation through simplicity, unification, successful prediction, or falsifiability; Kuhn’s view of scientific revolutions.