Private

Students have no access

Assign  Present 

Reason Better: An Interdisciplinary Guide to Critical Thinking by David Manley

Copy URL Rate This Content Edit Delete

3. Clarity
Contains 10 items Worth 10 points Adjust Points

Manley, David. 2021. Reason Better: An Interdisciplinary Guide to

Critical Thinking. Tophatmonacle Corp. Stable URL: https://tophat
Exported for David Manley on Sat, 14 Mar 2020 19:04:12 GMT
.com/marketplace/arts-&-humanities/philosophy/full-
course/reason-better-an-interdisciplinary-guide-to-critical-thinking
-david-manley/3425/117616/. Accessed: 27 March 2021.
Reason Better
An interdisciplinary guide to critical thinking 

Chapter 3. Clarity

Introduction
In the first two chapters, we looked at some influences on our reasoning that operate below the level of our
awareness and can lead us to make systematic errors. These things affect us most when our minds are cloudy.
For example, when we're stressed or tired, it's harder to reason carefully. This means we're more likely to form
beliefs or make decisions without really understanding why we did so.
Mental clarity helps us resist these effects. The purpose of this chapter is to introduce three kinds of clarity that
are crucial to good reasoning:

Clear inferences: how exactly do our reasons support our beliefs? 

Clear interpretation: how should we understand arguments presented by others? 
Clear language: what exactly are we claiming in our arguments? 

The task of philosophy is to make thoughts clear and to give them sharp
boundaries.
            —Ludwig Wittgenstein

Learning Objectives
By the end of this chapter, you should understand:

the components of inferences and arguments

the two criteria for a good argument
what makes an inference or argument suppositionally strong
the difference between a deductive and inductive inference or argument
the trade-off between suppositional strength and well-supported premises
what sorts of beliefs are found at the "ground floor" of a structure of beliefs
how to identify premises and conclusions of arguments
the need for interpretive charity in reconstructing arguments and identifying implicit premises
the nature of ambiguity, generality, and vagueness, and the differences between them
why vagueness o en can't be avoided
why it's wrong to assume that real and important categories can't have borderline cases 

3.1 Clear inferences

When our reasoning is muddled, we're not sure exactly why we have the beliefs we do, or make the decisions
that we make. When our reasoning is crystal clear, we can identify how our beliefs are supported, and which
facts we take to support them.
Suppose I come to believe that a certain star in the sky is the North Star because I have two other beliefs:

Alice said so
Alice is an expert on stars

When I arrive at a belief because I take it to be supported by other beliefs, I've made an inference.

In this case, neither of my supporting beliefs provides much

support for the new belief all by itself. Alice's opinion is not very strong support unless her views on stars are
reliable. And her expertise is irrelevant unless she said something about that particular star. It is only taken
together that these two beliefs provide strong support for my new belief that the star I have in mind is the
North Star.

If someone asks me why I think it's the North Star, I can express my inference in the form of an argument—
which, as we saw in the last chapter, is a series of claims presented as support for a conclusion. I could say,
"Alice said that's the North Star. And she's an expert on stars. So that's the North Star." The statements that
express my supporting beliefs are called premises, and the statement that expresses the belief they are
supporting is called the conclusion. To state an argument clearly requires making the premises and
conclusion clear, and also making it clear how the premises are supposed to support the conclusion (at least, if
it's not already obvious).

Note that the word "argument" is often used to describe a heated discussion, but an "argument" in our sense
need not even be an attempt to persuade anyone. If we want to get clear about one of our own inferences, it
can help to write down our conclusion along with the support we take it to have. In that case, we will have
written down an argument for the purposes of clarity rather than persuasion.

Of course, if we have a controversial belief, we may want to explain why we believe what we do. This may
involve expressing our reasons in the form of an argument, and others may not think that it is a very good
argument. But with the right mindset, this shouldn't make us defensive. It should be genuinely interesting why
other people think the argument is not good. If they can show us a weakness in the argument, it means that
our reasons for believing something were not as strong as we thought. They might be saving us from being
wrong!

The two elements

Once we are clear about the premises and conclusion of an argument, we can begin to assess how good the
argument is: that is, how well does it support its conclusion?

It's useful to think of an argument as similar to a portico like the one  on  the left, in which the weight of a roof
is resting on pillars.  The pillars must support the roof, but this requires two things: (1) solid pillars, and (2) a
solid connection between the pillars and the roof.  If the pillars aren't solid, they could collapse under the
weight of the building. But it's not enough to have solid pillars; they also need to be properly arranged and
connected so that they can bear the weight of the roof. Otherwise, no matter how solid they are, the roof could
slide off or collapse round them. 

Similarly, we are looking for two things in a good argument:

premises that are true

premises that are properly connected to the conclusion 

How well the premises are connected  to the conclusion—how much they would support it if they were true—
we will call the argument's suppositional strength. We'll get to the truth of the premises later.

Suppositional strength
In a suppositionally strong argument, the conclusion is well-connected to its premises. This means that if we
could be sure that the premises were true, we'd be fairly safe in accepting the conclusion.

In other words, to assess an argument's suppositional strength, we ask:

how much evidence do its premises provide for its conclusion if we suppose they are true?

As we will see in later chapters, evaluating the suppositional strength of an argument helps us figure out how
much to adjust our confidence in the conclusion if we learn that the premises actually are true.

So a suppositionally strong argument is like a portico whose pillars are strongly connected to the roof. This
means that if we can assume the pillars are solid, we can be confident that the structure itself will hold up. But
it doesn't mean that we can assume the pillars themselves are solid. The roof might be perfectly connected to
the pillars even if they are sinking into the ground, or made of a material that is too weak to bear its weight. 

Likewise, an argument can be suppositionally strong even if its premises are questionable—indeed, even if
they are obviously false. Here is an example of an argument like that:

         All giraffes can fly.

         I am a giraffe.
         So: I can fly.

This argument is suppositionally as strong as any argument can be—if we suppose the premises are true, we
have to accept the conclusion as well. There would be no way to avoid the conclusion if the premises were
true. Still, this is not a good argument, because we know the premises are false. An argument can have
premises that are perfectly connected to its conclusion, but if the premises are obviously false, the argument
will be no good. 

There are also arguments that have clearly true premises, but the premises are not well-connected to the
conclusion. That's like having pillars that are solid but aren't well-connected to the roof, so they can't support
it properly and it might just collapse around them.

For example: 

        All cats are mammals.

        Some mammals bark.
        So: some cats bark.

In this case, it's easy to see that the argument is suppositionally weak. We know that the premises are true, and
yet they don't give us any reason to think that some cats bark. Even if we had no idea whether cats bark,
learning that cats are mammals and some mammals bark wouldn't give us much reason to think that cats
bark.

Now consider this similar argument with a true conclusion:

        All cats are mammals.

        Some mammals bite.
        So: some cats bite.

It's easy to get distracted by the fact that this time the conclusion happens to be true. But actually that's
completely irrelevant to whether this argument is suppositionally strong. The question is whether the
argument gives us a reason to accept the conclusion. And it doesn't. If we had no idea whether any cats bite,
learning that cats are mammals and that some mammals bite wouldn't give us much reason to think that cats
do. In fact, the "bite argument" is suppositionally weak for exactly the same reason that the "bark argument" is.
(You'll see why if you compare the arguments!)

In short, the truth of the premises and/or conclusion is irrelevant to an argument's suppositional strength. If we
happen to know the premises of an argument are false, we have to set that aside and remember that the
argument can still be suppositionally strong, as long as if the premises were true, they would provide good
reason to accept the conclusion. On the other hand, if we know the premises of an argument are true, we have
to set that aside and remember that the argument can still be suppositionally weak, if the premises don't
actually support the conclusion.

Implicit premises
Sometimes the suppositional strength of an argument depends on information that is taken for granted but
not stated. For example, consider this one-premise argument:

         Fido is a dog.
         So: Fido has a tail.

Learning that Fido is a dog gives us a reason to think that Fido probably has a tail, but that's only because we
already know that most dogs have tails. In a conversation where we can assume that everyone knows that
most dogs have tails, someone presenting this argument can just leave that fact unspoken. When some claim
is being taken for granted in this way, we can treat it as part of the argument for purposes of assessing the
argument's suppositional strength.

Things do not always work out so well. We might take

something for granted in making an argument without realizing that it's not known by everyone in the
intended audience. Or we might skip over a premise because it's questionable and we're hoping no one
notices! In any case, when we leave a claim unspoken but take it for granted in making an argument, it's called
an implicit premise.

Of course, the exact content of an implicit premise is not entirely clear. For example, in the argument above,
the implicit premise could be "Most dogs have tails", or "Almost all dogs have tails", or "Dogs typically have
tails". For that reason, and because we are often wrong about whether other people share our assumptions, it's
best to be as explicit as possible when presenting an argument. Spelling out all of our premises explicitly might
feel long-winded, but it can be revealing for ourselves as well as for those we're communicating with. It's
common to discover that something we were taking for granted seems much less obvious when we have to
state it clearly and explicitly.

Deductive vs. inductive

So suppositional strength comes in degrees. The highest degree of suppositional strength is when the
connection between the premises and conclusion is so tight that there is no wiggle room at all.

We say that the premises of an argument entail the conclusion when they guarantee it completely: if the
premises were true, the conclusion would have to be true. No exceptions: there is no possible way that the
conclusion could be false and the premises true. 

So if you know that the premises of an argument entail its conclusion, then the argument is as suppositionally
strong as it can be—learning the premises would make the conclusion inescapable.

But again, even arguments with the highest level of suppositional strength can be terrible arguments. Recall
the argument from above:
         All giraffes can fly.
         I am a giraffe.
         So, I can fly.

There's no way for the premises to be true and the conclusion false: the premises entail the conclusion. But the
argument is still not going to convince anyone, because its premises are obviously false.

The process of arriving at a conclusion because we accept some premises that entail the conclusion, we call
that deduction. When an argument is presented as entailing its conclusion, we say the argument is
deductive. (We call it that even if the premises don't actually entail the conclusion—for example if the
speaker is making a logical mistake.) And when an argument's premises actually do entail its conclusion, we'll
say that the argument is deductively valid.

However, most arguments that people give in real life are not intended as deductive. Usually, the truth of the
premises is not supposed to guarantee the truth of the conclusion. Instead, the premises are just intended to
provide support for the conclusion—perhaps a great deal of support, but not so much that the conclusion
follows with the certainty of mathematical proof. When we arrive at a conclusion because we accept some
premises that merely support the conclusion, we call that induction. And arguments presented as merely
supporting their conclusions are called inductive arguments. (The Fido argument is one example.)

The tradeoff 

In many cases, we could offer either a deductive or inductive argument for a conclusion. For example, consider
these two arguments:

        All ravens are black.

        So, if there is a raven in that tree, it's black.

        All ravens I have observed so far have been black.

        So, if there is a raven in that tree, it's black.

The first argument is deductively valid; the second argument is not. However, the second argument may have
been intended as an inductive argument rather than a deductive one.

So which argument is better? It might be tempting to think that it's always better to have an argument whose
premises entail its conclusion. But, on the other hand, the first argument's premise is extremely hard to
establish as true. If someone were to challenge it, we would probably have to provide an additional argument
to back it up—perhaps using the very premise that's used in the second argument!
In real life, most of us have only observed black ravens but are also aware that most species have members
that are albino. So neither argument will convince us with certainty, but for different reasons. One has a
doubtful premise that would guarantee its conclusion if it were true; the other has a well-supported premise
that does not guarantee the conclusion. This comparison illustrates the fact that, when we are deciding how to
formulate an argument, there is often a trade-off between suppositional strength and how much support the
premises have.

Navigating this tradeoff requires careful attention to the context of the argument. For example, suppose the
speaker thinks that all ravens are black on the basis of a limited number of observations. In that case, the
second argument is more informative because it provides the reason for the speaker's generalization. An
argument should articulate the structure of reasoning in the speaker's mind, and the observational premise is
deeper in that structure than the premise that all ravens are black. So the second argument gets to the real
reason why the speaker expects the raven in the tree to be black. However, even the second argument is
presented in an overconfident manner. It would be better to add "probably" to the conclusion, to make it clear
that the argument is not intended to be deductively valid.

The ground floor 

As we have seen, evaluating an argument involves asking:

whether the premises are true

whether the conclusion is well-connected to the premises

We've been focusing on the second requirement, but haven't said much about the first. How do we know the
premises are true? If our premises support our conclusion, what supports our premises?

For example, consider again the argument that a certain star is the North Star. The premises were: Alice said so;
and she's an expert on stars. But why should I think the premises are true? Maybe she wasn't  talking about the
same star. Or maybe she's only pretending to be an expert. It seems that my two supporting beliefs also require
support.

Many of our beliefs are supported by multiple tiers of more

fundamental beliefs. Picture a building with several floors, where the top floor is directly supported by the floor
below, and that floor is in turn supported by additional floors. But, of course, the lowest floor of such a
structure is supported directly by the ground. This raises the question: what would correspond to the ground
floor of a well-built structure of beliefs? These would be beliefs that don't require support from other beliefs, or
maybe don't require any support at all. Let's consider two sorts of beliefs that seem to fall in this category.

First, there are beliefs that are supported directly by perception, rather than by other beliefs. These are
directly perceptual beliefs. If I look up and see an apple on the table in front of me, I will form the
belief that there is an apple there. If someone asks, "Why do you think there is an apple there?", I will say
"Because I see it." My belief that there's an apple on the table doesn't seem to be based on any other beliefs—
except maybe the belief that apples look like this and I seem to see this. Of course, it's possible that there's no
apple at all. Maybe someone spiked my breakfast with LSD, and I'm hallucinating. In that case, I seem to see an
apple but there is no apple at all. But I would have to be pretty unlucky for that to happen. If I have a history of
reliable sense-perception, and no reason to think anything strange is going on, then just accepting what I
clearly perceive with my senses is a fairly reliable method of forming beliefs.

There are hard questions about what counts as a directly perceptual belief. Is the belief that there is an apple
right there directly perceptual, or is it supported by the belief that I seem to see an apple? Either way, the key
point is that my belief that there is an apple ultimately rests on a perceptual process, so the reliability of that
belief depends on the reliability of that perceptual process.

Another hard question has to do with exactly what sorts of

cognitive processes count as directly perceptual, in the sense that they don't require further support from
other beliefs. For example, would a strong feeling that someone is watching me count as a directly perceptual
belief? Is religious experience a kind of direct perception, offering a foundation of support for religious beliefs?
And if so, how can we assess the reliability of these ways of forming beliefs? We will set such questions aside in
this text.

Another traditional category of beliefs that can be treated as on the ground floor of our belief structures are
self-evident beliefs. These are beliefs that are so obviously true that we don't know how we would
even go about supporting them with evidence. For example:

1 + 1 = 2.
Green is a color
If someone was murdered, they died. 
If x is taller than y, and y is taller than z, then x is taller than z.

What would evidence for these things even look like? Suppose we want to check whether everyone who was
murdered has died. This would involve identifying all the murders and seeing if the victims died. But death is
one of the criteria we would use to identify cases of murder in the first place!

There are hard philosophical questions about how we know self-evident things. One view is that we know
some of them simply by understanding certain concepts. According to this view, understanding the concepts
green and color is all it takes to know that green is a color.

There are also hard questions about which beliefs are really
self-evident. For example, consider a clear moral truth like inflicting pain for no reason is bad. Some
philosophers hold that obvious moral truths are self-evident, while others think they are known through a kind
of moral sense that is more like perception.

Whatever the answers to these questions, we should be very careful when we treat a claim as self-evident,
because as we will see, we have a tendency to think things are far more clear than they really are. In fact, we
humans have a very bad track record when it comes to things we consider to be completely obvious. For
example, there have been many studies in which people are asked how confident they are about various
simple and verifiable factual claims. Sadly, when we say we know something with 100% certainty, we are
correct only about 80% of the time; and when we say we are 90% confident, we are correct only about 70% of
the time [1].

Section Questions

3-1 Show Correct Answer Show Responses

If I believe something based on the support of other beliefs, then...

A I have made an inference

B I have made a claim

C I have made a deductive argument

D I have made an inductive argument

3-2 Show Correct Answer Show Responses

If the truth of the premises in an argument does not guarantee the truth of the conclusion...

A the argument is not an acceptable argument

 B the argument is not suppositionally strong enough to make an inference

C the argument should not be presented as deductive but may still be a good argument

D there is no reason to accept the conclusion even if you know the premises are true

3-3 Show Correct Answer Show Responses

For an argument to be suppositionally strong means...

A that its premises are convincing and would give us good reason to accept the conclusion

B that its premises are well-supported and provide strong evidence for the conclusion

C that its premises would give us good reason to accept the conclusion if they were true

D that it is a good inductive argument but should not be presented as deductive

3-4 Show Correct Answer Show Responses

If we have support for a claim that we can present either as an inductive or a deductive argument...

A the inductive version is likely to have better-supported premises

B the deductive version will be suppositionally stronger and therefore a better argument

C there can't be a deductive version whose premises entail its conclusion

D it never makes any difference which type of argument we choose because both can be good arguments
 3-5 Show Correct Answer Show Responses

Which of the following is not true?

A Directly perceptual beliefs need not be supported by other beliefs

B It is reasonable to treat any claim as self-evident as long as we are clear in our argument that we are doing so

C Usually, believing what our senses clearly present to us is a fairly reliable way of forming beliefs

D If we think we know something with certainty, there is o en a decent chance we are wrong.

3.2 Clear interpretation

Getting clear about our own reasoning requires understanding exactly what it is that we believe, and what the
structure of our beliefs is. Likewise, we should be clear when we express our reasoning in the form of
arguments. This means we should: 

state every premise and the conclusion in clear language

make the structure of the argument clear

If you are reading or hearing an argument, it can take some effort to figure out exactly what the premises and
conclusion are, and what the structure of the argument is supposed to be. This can be equally hard if you are
the one putting forward the argument!

It's often useful to begin by identifying the conclusion. Ask yourself: what is the argument driving at? Which
claim is supposed to be the thing that is supported, and what claims are supposed to be supporting that
claim? What, if anything, does the author or speaker want to convince me of?

Next, try to find all of the claims that are being made in support of the conclusion. Sometimes a premise
provides support for the conclusion all on its own; sometimes it only provides support for the conclusion once
it is linked up with other premises. And, in a bad argument, a premise might provide no support for the
conclusion at all. What makes it a premise is that it is intended as support for the conclusion.