In this section of our series we'll consider epistemology, starting with an outline of what we mean by the term and moving on to look at some of the ideas proposed to answer the questions that arise in this area of philosophy. There has been much recent work in the subject so we'll have a lot to cover.
What is epistemology?
The word itself derives from the Greek epistéme ("knowledge" or "science") and logos ("speech" or "discourse"). In English, then, we render it as the theory of knowledge, concerned with the nature, sources and scope of knowledge. Interestingly enough (or not, as the case may be), in both French and Italian epistémologie and epistemologia respectively actually mean the philosophy of science, but we'll concern ourselves here with the general subject of ertkenntnistherorie and move on to scientific epistemology later; not so far down the line, though, as to disappoint those readers looking to jump on your humble narrator if he should mention Feyerabend.
Epistemology, then, is concerned with the following:
- The nature of knowledge: what is knowledge? What do we mean when we say that we know something?
- The sources of knowledge: where do we get knowledge from? How do we know if it's reliable? When are we justified in saying we know something?
- The scope of knowledge: what are the limits of knowledge? Are there any in the first place?
We'll look at each of these in turn and try to get a grip on them via the efforts of philosophers or the past and present. First, though, we'll take a glance at some historical developments that led to these problems in the first place.
The beginnings of epistemology
The early Greeks were studying and trying to determine the properties of reality. They were initially, generally speaking, quite sure of themselves and pronounced in one way or another on its nature. Heraclitus, for example, thought that change was constant; Parmenides, on the other hand, allowed that reality could not change. His argument ran that if Being were to change, it could only become not-Being, which is absurd. Such speculations involved a certainty as to what could or couldn't be said about reality that led others to questions whether we can be so sure of ourselves to make such declarations, hence leading to the birth of skepticism
with the Sophists
. However, Socrates, Plato and Aristotle still thought that the mind could reach truth and certitude but they considered instead the conditions under which we can say something is valid. Plato thought that knowledge was the attempt to get at universal ideas or forms, and, while Aristotle placed more emphasis on logical and empirical ways of so doing, he still accepted the basic idea that universal principles were to be sought.
It was recognised early that our senses are sometimes unreliable: for example, most of us have experienced the strange phenomenon of seeing someone far away who looks just like a friend but upon closer inspection proves to be a stranger. Some thinkers nevertheless supposed that we could arrive at knowledge by using our intellect to go beyond the information given by our senses and arrive at general or universal
notions and ideas. The question then
was what we could say about these universals; do they exist, or are they just a product of our thinking? This led to a great deal of controversy during the Middle Ages between the realists
, who thought that these universals existed independently of whether anyone was around to think about them, and the nominalists
who held that they did not. (Like most philosophical disputes, the matter was actually not so clear-cut.) Another question brought up by the fallibility of our senses is to wonder why we nevertheless arrive at understandings that seem to work.
Most people have heard of Descartes and his famous cogito, ergo sum
. His method was to try to doubt everything and see what remained standing, believing this would lead to ideas that were certain. The empiricists
thought that knowledge could only be gained from experience
; the rationalists
considered instead that reason
could reveal it to us. These two give us the oft-heard phrases a posteriori
, meaning "after
the fact", and a priori
, meaning "before
the fact". Of course, it seems strange to insist that we choose one or the other.
Partly in response to Hume (whom we'll consider later), Kant proposed to provide a synthesis of these two and proposed the following categories for what we could find in the world: phenomena
, or things as they appear to us; and noumena
, or things as they are in themselves. He suggested that we could have knowledge of the former, but not the latter. Of course, it could be that the two are not so far apart—close enough for government work—and that this explains why our senses still give us useful information. With Kant we reach some of the modern problems that are still vexing epistemologists today, but before moving on to the specific areas we'll lastly consider the pragmatism of recent years.
is this context may be explained as considering something knowledge if it is useful
to some end. The attendant questions of whether an idea is true
or gets at reality are sometimes considered meaningless or at least not important; what matters is that a given model helps us to solve problems
. Does it matter where we get our knowledge from if it works
? This trend has led to some interesting epistemological avenues of investigation that we'll consider later on.
The nature of knowledge
The first thing we need to consider is the question of what knowledge is
. A popular classical account held that knowledge is justified true belief
(JTB): we say we know x
if, and only if,
- x is true (1);
- We believe that x (2);
- We are justified in believing that x (3).
Let's take a specific example: suppose we want to say that we know it's snowing outside. According to this account of knowledge, to know
that it's snowing we must have:
- It's true that it's snowing outside (1);
- We believe that it's snowing (2);
- We're justified in believing that it's snowing (3).
In the first case, it could be the case that it is
snowing; in the second, we need only believe
it; and in the third, we could be justified in so believing because we can see the snow falling and children are praying (religious or not) that school will be cancelled. This seems straightforward enough.
Unfortunately a paper by Edmund Gettier pulled the rug out from under this conception. He suggested that we could infer a true circumstance from a false one and then find (2) and (3) satisfied. In that case, we would've made our way to knowledge from a false claim, making the road to this truth something of a fluke. This would be a strange kind of knowledge.
Looking at our example, then, it could be that we say "the thermostat went up, so it must be snowing". In fact, it may have been that the temperature didn't change and we in fact imagined it; then we would've arrived at justified true belief that couldn't be considered knowledge. The so-called "Gettier Problem" frustrated attempts to counter it and many epistemologists gave up on this account of knowledge.
Sources of knowledge
Let us consider, then, some other ways of arriving at knowledge. We could use our senses, our memory, testimony or reasoning. However, any or all of these could be flawed, so there are several theories for how we use them to get at knowledge.
One alternative to the JTB account is reliabilism
, in which we say that we are justified in believing something if, and only if, we arrived at it via a reliable
cognitive process. For example, in the past our senses have been very reliable (barring the few exceptions we touched on before), so we may be justified in saying "I know there's a computer in front of me". The difference between this account and JTB is the removal of condition (3). Unfortunately the Gettier problem is a difficulty here too and there is the issue of what makes a process reliable in the first place.
One way of explaining where we get our knowledge from is to start with certain ideas or statements as our foundations
and build on them, which is called foundationalism. This is much the same way as we can proceed in mathematics—starting with some axioms and building up a system based on what we can get from them. An obvious question to ask, of course, is where we get our foundations from in the first place: are they justified? Some thinkers propose that the choice of such foundational concepts doesn't need to be explained because they're so obvious; perhaps denying them makes no sense at all, for instance, like questioning if the universe really exists. In more recent times some philosophers have called such fundamental assumptions properly basic
, meaning that they require no justification at all and can be held by any reasonable person without argument. The trouble then
, though, is debating which ideas can be so called.
Some people don't think there are any foundational beliefs; instead, they think that our ideas hang together (or cohere
, hence the name) and support each other, like bricks in a building. In that case, for example, we might reject the claim "Hugo is interesting" because it doesn't fit with all we already know about how dull he is; on the other hand, if someone said "Hugo is as mad as a clown's trousers" we might be inclined to accept it because we already have a long list of instances that have given an inkling of that theory.
The difficulties with coherentism are several: firstly, how do we know which
ideas a new one has to agree with? Secondly, how do we tell the difference between a true idea and a false one, given that the latter may still agree with a lot of what we already think we know? Lastly, it seems that some of our thinking is more
certain than the rest and therefore has greater importance, like the foundations in the brick building; can we account for this feeling?
As we noted above, empiricism proposes that our knowledge comes from experience
. One of the attractions of this idea is that we might be able to avoid false superstitions with no basis in the world, but there are problems all the same. Many people—not just philosophers—have wondered how logic, mathematics or even ethics could be based on experience: isn't twice two four whether our earth exists or not? If so, how can we account for these (and other) areas from experience alone? On the other hand, we do seem to take a lot of our ideas from our experiences in the first place and when we come on to consider scientific epistemology
later we'll see that empiricism has a large part to play with some further (theoretical) difficulties as well.
On the other side of the coin, we have the notion that much of our knowledge comes from reason
, or the act of reasoning
. In particular, we might be looking for knowledge that must hold, irrespective of circumstances, like mathematics or logic (again). The overuse of reasoning, though, can lead to being accused of piloting an armchair; no matter how well you can fly it, you won't leave the lounge and see what the world outside can tell you. It makes sense to think that empiricism and rationalism can tell us a good deal together
In epistemology, naturalism refers to the idea that knowledge can be studied as a science and involves a relationship between sense inputs and cognitive (or mental) outputs. In that case, psychology, sociology and biology can tells us a lot about how we come to our beliefs and further investigation may show how our experiences influence what we end up thinking. We can also apply evolutionary ideas to our questions above.
There are more possibilities to consider here, but they make sense best in a specifically scientific context. To that end, we'll save them for the later discussion.
The scope of knowledge
Perhaps the many and varied difficulties we have seen so far could lead to the suggestion that no knowledge at all is possible, or that at the very least it's a tricky business? Can we know anything
Some thinkers have suggested that no matter how hard we try or how successful our efforts, there always remains the possibility of error. Even when we feel
absolutely sure of something, we could still be wrong. Fallibilism is the idea that all
knowledge is provisional and could have to be revised at any instant. Essentially, then, it involves the notion that perfect certainty is impossible.
Being traced back to the Greeks (again), skepticism has a long history. We've seen that the conditions required for knowledge are strict and perhaps they may never be satisfied. Some skeptics take this "alas, it doesn't seem that it can be done" attitude, while others are quite sure that knowledge is impossible. (Another usage of the term refers to withholding judgement.) Generally speaking, fallibilism can lead to skepticism.
An argument heard often against skepticism is given by the question "how do you know
that nothing can be known?" The implication here is that the skeptic contradicts him- or herself, knowingly (pun intended) or otherwise. In fact, the skeptic can get around this using an interesting idea from Bertrand Russell, wherein the claim "nothing can be known" is rewritten as, for example, "there is no x
such that x
matches the description of knowledge". This means that the skeptical challenge is a powerful one but the impossibility of certain
knowledge may only mean that we have to be satisfied with what we can get.
As noted, we'll move on to study some of these notions in more depth when we consider scientific
Some common topics
We'll finish this discussion by looking at a few subjects that come up frequently and investigate them a little.
The tree in the woods
Many of us have heard the question "if a tree falls in the woods and no-one is there, does it make a sound?" (Indeed, Bart Simpson employed it to his benefit while playing golf.) This conundrum here is based on the idea of verificationism
: a claim is justified only if it can be verified in some way. In the case of our tree, then, we seem to know
from past experience that falling trees make a noise; however, if there's no-one there
to hear it, it can't be verified
—hence the confusion. In fact, if we leave a tape recorder behind we can soon answer the question.
The problem can be extended by disallowing any such monitoring device but this shows—quite clearly—that the epistemological ideas we adopt influence what we can say.
In his writings, Hume (amongst other things) expressed what is now called the problem of induction
: we ask whether a finite number of particular cases can ever justify a general conclusion. For example, suppose that we visit a farming district and see very many sheep that are all white. Can we then assert from the hundreds of sheep we saw that all
sheep are white? It turns out that we would be wrong to do so, or else there'd be no nursery rhyme. How many white sheep would we need to see before we can justify saying that all sheep are white, though? To look at this further, let's lay out the information we have in a logical form using what we learned in the last article from Gabriel Dumont:
Premise 1: The first sheep was white.
P2: The second sheep was white.
P501: The five hundredth and first sheep was white.
(And so on.)
Conclusion: All sheep are white.
The problem is that none of the premises contain the conclusion and all of them are moreover the same in form, so we're relying on a kind of brute force of numbers. Suppose we saw another thousand sheep, all of which were also white. Are we justified then? Again, apparently we aren't because in fact some sheep aren't
white? Is there any way around this difficulty?
One thing we can do is recast the problem: instead of asking when we're justified in saying that all sheep are white, we can wonder instead when it would be reasonable
to assume it. In this light, the matter takes on a much different hue. If we see but one
white sheep, it seems unreasonable to insist that all sheep are white. However, once we've seen several hundred of the walking sweaters we may be reasonable in supposing that they all are. The subsequent finding that some sheep are black doesn't change the fact that we were justified in supposing them to be white.
In recent times, Goodman has posed what is called the "new riddle of induction". Rather than using his example, let's stay with the sheep and reconsider what we have. Our observations of hundreds of white sheep has led us to propose the theory that all sheep are white. Suppose now that we offer another theory: all sheep until some time T in the future will be white, but the next one will be purple. The evidence we have to date supports our theory, but it also fits this new (but silly-sounding) theory too. How do we decide which is the more reasonable, given that both are equally well grounded? Moreover, we could create plenty of other theories of the same form, involving sheep of all the colours of the rainbow and more besides. We can't say that we have to go with the white theory because we've only seen white so far because that is assuming what is to be proven; i.e. that we expect to see new sheep that are also white. This interesting and perplexing problem is notoriously difficult to escape. In scientific terms it's called the underdetermination of theories
and we'll see it again when we come on to scientific epistemology
and the philosophy of science
. Look for these topics next. Later in our series we'll return to epistemology again for a further discussion.
Dialogue the Third
The Scene: Our intrepid philosophical adventurers have moved on to a public house of ill repute called The Drunken Bishop, having acquired its moniker following an incident involving an atheist in fancy dress who, unable to hold his liquor, passed out while the owner was discussing a change of image for the establishment. A hastily painted picture of the unfortunate fellow adorns the sign outside. Our friends are, as if by coincidence, discussing Hume.
So tell me about this so-called problem that we're all supposed to be worried about. I read about it in one of these "science is just another tool" polemics. (He is drinking a real ale.
What did it say?
The gist of it, my dear, is that those who don't think we can justify a claim that's been found to hold time and time again are not so full of themselves as to test it by throwing themselves out of a window instead of using the lift.
(Anna looks to Trystyn, who is himself counting the number of occasions that the phrase "you know?" comes up in the conversation of the couple on the next table.
If I had a pound for every time I heard this you'd still be buying the drinks but I'd have a warmer coat.
If you want another you can tell me where I'm missing the point. Why are philosophers so poor anyway? Your talents are under-appreciated, no doubt.
Not by her, apparently. (She nods in the direction of a girl who is trying to get Trystyn's attention.
(Smartening himself up...
) Be quick, Socrates, before I make my move.
Okay, Don Juan. The general point we want to make is that on previous experimenting we've found that throwing something out of a window has resulted in it falling to the ground. If it was a grand piano, we of course had comedy potential below, especially if we had an anvil to follow it. Even more generally, we posit the existence of a force—gravity—that explains why it should happen that way. In particular, we noted that unfortunate or deluded pseudo-supermen who opted to fly downstairs also failed to fight crime another day.
(He pauses to drink some of the red wine he is sharing with Anna.
We want to conclude from this finite
number of specific cases that throwing ourselves out of a window is not such a good idea, the more so if we have business to attend to at ground level.
Where's the problem, then? (He is looking at the girl, who is still looking at Trystyn, who is looking at his wine while Anna alternated between glaring at him and the girl.) The moral of the story is: don't throw yourself out the damned window.
I think the point is that no number of cases is enough to justify the inference...
So if I launch myself from my cousin's slide head-first into the garden each day for a year and twice on Sundays, you'll say I can't justify my cousin supposing I'm an idiot?
) You must know I'll disagree with you whatever you say, but at least this time we have an interesting point to draw out.
... that my four-year old cousin is smarter than all of us, you mean?
You may be right. Okay—here we go:
(He clears his throat in dramatic fashion. Anna is glaring at the girl now and the latter has caught on.
) Try to imagine a situation in which we throw ourselves out of the window and in fact don't
fall; perhaps we fly all the way to Krypton instead. Ask yourself if there's anything that smacks of the impossible about the idea
of it. You can't really refer to gravity making the suggestion ridiculous because we could ask the same thing: what, before the fact, is so impossible about the idea that gravity might not apply at some particular time?
The obvious response, of course, is to say "well, it always has
; why should today be any different?" Even so, that hasn't answered the point at all: how are we to go from the fact that everyone to so far don the red cape has failed to the assertion that everyone must do so, given that
the idea doesn't seem impossible to conceive?
Now take the fellow not bothered by such matters who challenges the philosopher to jump. In the first place, the philosopher who takes the bet is probably doing the gene pool a favour, but suppose that he does indeed dent the pavement—what then? What has the other fellow shown? That the trainee Reeve has to fall? No, since we already noted that the divers of the past weren't enough for that; all he has is one more case to add to the stockpile that wasn't enough to begin with regardless of how many unfortunate people litter the street below. So by demanding that the consistent philosopher jumps if he doubts the inductive arguments that Hume wrote of he in fact shows only two things: one, that he didn't understand the problem at all; and two, that there are some dumb and flightless philosophers in our world.
I see your point, Lex, but what kind of idiot really worries about this kind of thing? Aside from philosopher superheroes, I mean. (Anna is struggling to pay attention now.
That's a good point, and one that's been made before. Sure, we can't justify an assertion about always falling, but who's going to bet against it? It seems unreasonable
to assume otherwise, so Hume's problem is not something to lose sleep over.
Right. (He begins to focus his attention on the girl again.
Not so fast, Errol. Suppose I offer a theory that everyone will fall until
tomorrow at noon when the bastard son of superman flies off into the sunset. You, on the contrary, say everyone will fall. Which is the more reasonable theory? Unfortunately for both
of us, all the facts collected to date about plummeting people supports both theories equally well. Yours may seem
more reasonable but the evidence collected applies to both our ideas just the same. That, my friend, is a far more difficult problem, but not as important as pretty girls, I'll wager. (He looks at Anna but she doesn't notice.
If you'll excuse me for a moment, I'm must have a word with her... (She leaves in the direction of the girl.
I think she likes you, you know. (He winks.
I think I'm thirsty, Cilla.
By Paul Newall (2004)