Lecture 12 - A Problem for Arguments

Now we've learned how to identify an
argument and put it in standard form.

We've also learned from the definition of
argument, that the premises are intended

to be reasons for the conclusion. Great.
Intended to be a reason for the

conclusion. Well intentions are nice, but
success is better. What we need to figure

out is when the person succeeds in giving
premises that really are reasons for the

conclusion. For simplicity, let's focus on
argument whose purpose is justification.

Then the question is, do the premises
justify you in believing the conclusion?

Well, imagine that you don't know whether
there's any life on Mars. You have no

evidence one way or the other. Then you
ask a friend and the friend says ha, I

know there's life on Mars. I can prove it
to you. Here, this argument will show you

that there's life on Mars. There is at
least one bacterium on Mars. Therefore

there is life on Mars. Now notice that if
the premise is true the conclusion has got

to be true. And if you and your friend are
justified in believing the premise then

you and you are friend are also justified
in believing the conclusion. So this

argument looks pretty good so far but of
course you have to ask your friend, well

how do you know that there's at least one
bacterium on Mars? And suppose your friend

says, well I'm just guessing. Then the
argument is clearly no good. If there's no

reason to believe the premise because your
friend is just guessing then you're not

justified in believing that premise. And
if you're not justified in believing the

premise then how can that premise make you
justified in believing the conclusion.

More generally an argument cannot justify
you in believing the conclusion unless you

are justified in accepting the premises of
that argument. Now, suppose your friends

says oh, but I do have a reason for the
premise I do then we have to ask what kind

of reason is it? And, at that point your
friend needs to express reason. And, how

do we express reasons? In arguments. So,
your friend has to give another argument

for th e premise. Where the premise of the
first argument is the conclusion of the

second argument. Wait a minute, we got a
problem. Because that second argument is

itself going to have premises. And you
have to be justified in believing those.

So the premises of the second argument
have to be the conclusion of a third

argument. And so on and so on because the
third argument needs premises. They have

to be justified. So they have to be the
conclusion of another argument. Which also

has to have premises. And those have to be
the conclusion of another argument. And so

on and so on. It looks like we've got a
real problem here. In order for the

premises to be justified they have to be
backed up by an argument, but the argument

has premises of its own that have to be
backed up by another argument and so on,

and so on. This problem is called the
problem of the skeptical regress because

you regress back to one argument after
another, after another, after another,

after another, after another, after
another. And it's hard to see how that

regress, is ever, going to come to an end.
There seem to be only three ways to get

around this skeptical regress. The first
is to start with a premise that's

unjustified. If it's unjustified, then it
doesn't need an argument to back it up.

And that means that you're not going to
have this chain of arguments going back

and back and back and back and back and
back and back. The second possibility is

to have a structure where the arguments
move in a circle. One claim is justified

by another which is justified by another
which is justified by another which is

justified by the first claim. And they
just move in a circle. The third

possibility, is that the chain of
arguments goes back infinitely. It never

stops. Every claim has an argument to back
it up. And there's no end, so you never

have a prentice which doesn't have an
argue to back it up, cuz it's infinite.

Those seem to be the three main options
here to avoid the skeptical regress. The

first possibility then, is to start with a
premise that's unjustified. And that seems

pretty neat if you can get away with it
but we already saw why that won't work. We

saw your friend arguing that there's life
on Mars because there's at least one

bacterium on Mars, and he was just
guessing. If you just guess at your

premises, you have no reason to believe
them. Then, an argument that uses those

premises cannot justify you in believing
the conclusion. But in addition, just

think about it this way. You could prove
anything if we let you start with

unjustified premises. If you can just make
up your premises for no reason then

there's no stopping you from believing
whatever, including things that are

obviously false. So it seems to be a real
problem to start with premises that are

unjustified. Next, the second way to
respond to the skeptical regress is to use

a circular structure. and it's kind of
neat, if you think about it, because, if

you want to prove one claim, you prove it
on the basis of another claim. And then

you prove that second claim on the basis
of the third, and the third on the basis

of the fourth and the fourth on the fifth,
and the fifth on the basis of the first.

And now you've got this circle, and the
arguments go in a circle. But that means

that every premise has an argument to back
it up. Cause you can keep going around the

circle forever. You can think about it a
little bit, it'll be obvious that that's

no good. And that can be shown by looking
at the smallest circle there is. So

suppose your friend says, I can prove
there's life on Mars. Here's my argument.

There's life on Mars, therefore there's
life on Mars. Clearly, that's no good. And

the reason why it's no good is that if he
didn't know whether there was life on Mars

to begin with, you wouldn't know whether
the premise was true. Cuz if you don't

know the conclusion you can't know the
premise, since the premise is the

conclusion. So if you're not justified in
believing the conclusion to begin with,

you're not justified in believing the
premise. And that means that the argument

didn't really get you anywhere. It just
ends up where it started. And in addition,

it has the same problem we saw in the
first approach involving unjustified

premises. Because you can use circular
arguments to prove anything. You can prove

there's life on Mars there's life on Mars
therefore there's life on Mars. You can

prove there's no life on Mars therefore
there's no life on Mars. You can do it

either way and the fact that an argument
can be used either way to prove either

conclusion suggests there's a big problem
with that kind of argument. So now we are

down to the third and final way to get
around the skeptical regress, and that is

to use an infinite chain of arguments. If
you think about it, in a concrete case,

you'll see why that's a problem as well.
Suppose your friend says there's life on

Mars and I can prove it. And you say fine,
give me a reason? Well there's a least one

bacterium on Mars, therefore there's life
on Mars. And you go okay, fine but how do

you know there's a least one bacterium on
Mars? I've got another argument he says,

there are at least two bacterium on Mars,
therefore there's at least one bacterium

on Mars. But how do you know there's at
least two? Well, there are at least three

bacteria on Mars. Therefore, there are at
least two bacteria on Mars. But how do you

know there are at least three? Well, there
are at least four so there are at least

three. Well, there are at least five so
there are at least four. Well, there are

at least six so there are at least five.
And so on and so on and so on. You could

go on infinitely. So an infinite chain of
arguments. Would allow you to prove that

there's life on Mars even if you have no
evidence whatsoever of any bacteria

because you're going to have an argument
but if the premise that you are arguing

from doesn't have an independent
justification then the infinite chain is

going to be no good at all in justifying
the conclusion of that argument. So many

people see this skeptical regress as a
deep and serious philosophical issue. If

the unjustified premise appr oach doesn't
work, and the circular argument structure

doesn't work, and the infinite chain of
arguments doesn't work, then it's hard to

see how we can get around the problem.
Which is to say, it's hard to see how any

kind of argument could ever justify us in
believing anything. Philosophers really

scratched their heads about that for a
long time. And worry about it. It keeps

them up at night. But we're going to have
look at how practical people solve a

similar problem in everyday life. So how
we solve this skeptical regress problem in

everyday life well there various tricks
which you can use. For example, one way is

to just start from assumptions that
everybody shares. So if I say, well you

really ought to buy a Honda because Hondas
are very reliable cars. Then I'm assuming

that you want your car to be reliable. You
don't like to have to take it in to the

mechanic all the time. You don't want it
to break down on the road. And if you want

reliability and I want reliability, then
we can start from the assumption that

reliability's a good thing and that that's
a reason to buy a car that is reliable.

But of course you might say, well but our
Honda's reliable. And then I might. Appeal

to an authority. Well, it's obvious that
they are. Or, Consumer Reports has done a

study that shows that they're reliable.
And I can appeal to an authority, and if

you accept that authority, go, Consumer
Reports, we can trust them, then, my

argument's going to work. You're going to
have a reason to believe the conclusion,

and it might persuade you, and make you
come to believe the conclusion. But,

suppose that someone is going to raise an
objection. They say, well, Consumer

Reports has been wrong before. They might
be wrong this time. Well, then, I need to

discount that objection. I need to respond
to it, and say, well, maybe they have been

wrong sometimes, but this time, you know,
they've got a good study and it was

careful, or whatever. And the but means,
I'm discounting the objection that you

have raised, and I might even discount i t
in advance. Well I might just guard my

client I might say, well. They might be
right, in this case. Or they're probably

right, without claiming that they
definitely are right. So I can assure you

by signing as some kind of authority. I
can discount objections and I can guard my

premises by saying, well it's probably
right. Instead of saying that it's

definitely right. And those are three ways
of solving. The skeptical regress problem

in everyday life that we're going to look
at in much more detail in the next three

lectures. But the point of this lecture
has been more general. In order to solve

the skeptical regress problem, you have to
find some assumptions that you and your

audience share. They might be assumptions
about the premises of your arguments. They

might be assumptions about authorities
that supposedly support your premises and

that they accept as authorities or
whatever. But there have to be some

assumptions that you share with your
audience in order to get the argument

going. And that's kind of tricky cuz it's
going to depend on the context. If you are

dealing with an audience that shares a lot
of your assumptions that argument is going

to be relatively easy. But if you are
dealing with an audience that doesn't

share any of your assumptions it's going
to be impossible. And, in areas where

there's a lot of disagreement it's going
to be hard to get your argument going

because your premises are going to be
questioned and denied or rejected by the

people in the audience. So, what these
tricks do, is they give you ways to get

the argument going. But they're not going
to work in every case. And we'll have to

look at that, as we're looking at these
three different ways to solve the

skeptical regress problem in the next
three lectures.