IDENTIFYING FLAWS IN REASONING

Some arguments give either no support, or such weak support, to their conclusions that it
is reasonable to regard them as having a flaw. This may be because a mistake in logic is
made in moving from the reasons to the conclusion, or it may be because the reasons
support the conclusion only if they are accompanied by an implausible assumption. The
skill of identifying flaws in reasoning is being able to see that the conclusion does
not follow from the reasons or evidence, and being able to say why it does not follow. We
illustrate this with the following examples.
Example 1: Violence on television
In Chapter 1, when discussing assumptions, we presented the following example of an
argument:
Some people say that the depiction of violence on television has no effect on viewers’
behaviour. However, if what was shown on television did not affect behaviour,
television advertising would never influence viewers to buy certain products. But
we know that it does. So it cannot be true that television violence does not affect
behaviour.
One way of summarising this piece of reasoning is:
Reason: Television advertising affects viewers’ behaviour.
Intermediate conclusion: So, what is shown on television affects viewers’ behaviour.
Main conclusion: So, violence shown on television must affect viewers’ behaviour.
If we take the intermediate conclusion as meaning that some of what is shown on television
affects behaviour, then it does follow from the reason given, because television advertising
is some of what is shown on television. However, the intermediate conclusion, interpreted
in this way does not support the main conclusion, as it is intended to, because violence
might be one of the things shown on television which does not affect behaviour. If, on the
other hand, we interpret the intermediate conclusion as meaning that everything shown on
television affects behaviour, then it does not follow from the reason, because from the fact
that one thing shown on television affects behaviour, it does not follow that everything else
shown on television will do the same. So, whichever way we interpret the intermediate
conclusion, this is not a good piece of reasoning, because it does not give good grounds for
the conclusion it draws.
If we are asked to say what the flaw in the reasoning is, we could express it as follows:
The fact that some things which are shown on television affect viewers’ behaviour is
not a good reason for thinking that violence shown on television must affect viewers’
behaviour,
or
The fact that advertising shown on television affects viewers’ behaviour is not a
good reason for accepting that everything shown on television affects viewers’
behaviour.
The ability to state flaws in this way is an important skill to develop, because it can be an
effective way of showing other people that there is something wrong with their reasoning.
Note that we have stated this flaw without ever considering whether the basic reason –
that television advertising affects viewers’ behaviour – is true. If we can identify flaws in
reasoning, then we can often be satisfied that a particular piece of reasoning does not
establish its conclusion, without needing to dispute the truth of the claims upon which the
conclusion is based.
We noted in our earlier discussion of the above example that another way of interpreting
the argument was to see it as assuming, unjustifiably, that television advertising and
violence shown on television were comparable, or analogous, in all relevant or important
respects. When assessing arguments, it is useful to look out for analogies or comparisons,
and to consider whether the two things which are being compared really are alike in ways
which are relevant to the conclusion which is being drawn. This was evident in
our discussion on pp.40–41 of the argument about written tests for learner drivers in
Portugal and the UK.
Example 2: Affluence and health
Let us consider another example:
If people became healthier as the affluence of the country increased, we would
expect the population to be healthier now than it was thirty years ago. But over the
last thirty years new illnesses, such as chronic fatigue syndrome, have appeared, and
we have become more vulnerable to old diseases such as heart disease, strokes and
cancer. So the increased wealth of the country has not produced improvements in
the health of the population.
The first thing to do when we want to assess whether an argument is flawed is to sort out
what the conclusion is, and what evidence or reasons are offered for it. Before reading on,
identify the conclusion and the reasons in this passage.
The conclusion, signalled by the word ‘So’ which introduces the last sentence, is:
the increased wealth of the country has not produced improvements in the health of
the population.
The evidence offered for this is that over a period during which the wealth of the country
has increased, new diseases have appeared, and certain old diseases have become more
common. Here is a more detailed analysis of the reasoning. There are two strands. First:
Basic Reason 1: Over the last thirty years new illnesses, such as chronic fatigue
syndrome, have appeared, and we have become more vulnerable to old diseases such
as heart disease, strokes and cancer.
This is intended to support an unstated:
Intermediate conclusion: There have been no improvements in the health of the
population over the last thirty years.
The second strand is as follows:
Assumption (unstated): The affluence of the country has increased over the last thirty
years.
This gives support to:
Basic Reason 2: If people became healthier as the affluence of the country increased,
we would expect the population to be healthier now than it was thirty years ago.
The intermediate conclusion and basic Reason 2 are then taken together to support the
main conclusion. Before reading on, ask yourself whether any of the moves in this
reasoning are flawed. Do you accept that the intermediate conclusion follows from
basic Reason 1, that basic Reason 2 follows from the unstated assumption, and that
the main conclusion follows from the intermediate conclusion together with basic
Reason 2?
Remember that when we are looking for flaws, we are not considering whether the
reasons are true. So, we do not ask, ‘Is it true that the wealth of the country has
increased over the last thirty years?’ and ‘Is it true that new diseases have appeared, and
certain old ones have become more common?’. We say instead, ‘Even if these claims are
true, do they give adequate support to the conclusion that the increased wealth of the
country has not produced improvements in the health of the population?’ It is clear that
they do not give adequate support, because we have not been given much information
about the general health of the population. It may be true that there is more vulnerability
to heart disease, strokes and cancer, but perhaps some ‘old’ diseases, for example
tuberculosis and bronchitis, are much less common. Perhaps people have longer lives
than was the case thirty years ago, and perhaps they are relatively healthy for long
periods of their lives, before succumbing in old age to heart disease, strokes or cancer.
There is a problem of interpretation here – what exactly is meant by ‘the health of the
population’? If we assume that it refers to the percentage of people’s lives during which
they are free from illness, then we have insufficient information upon which to base the
conclusion.
Now we must state concisely what the flaw is:
Even if some new diseases have appeared and some old diseases have become more
common during the last thirty years, it does not follow that the population is less
healthy than it was thirty years ago, because people may have long periods of good
health before suffering from these diseases.
Note that the flaw occurs in the move from basic Reason 1 (the claim about prevalence of
diseases) to the unstated intermediate conclusion (that the population is less healthy now
than thirty years ago). Note also that, in establishing that this is a flawed argument, we
have not established that the main conclusion is false. It may be true that the increased
affluence of the country has not produced improvements in the health of the population.
This could be true if, as the argument tries to suggest, there have been no improvements
in the health of the population. But it could be true even if there have been improvements
in the health of the population, because those improvements might have occurred
even if the country had not become more affluent. So someone aiming to counter the
original conclusion in the way set out in Example 2 would also be producing a flawed
argument.
Example 3: Affluence and health – a connection?

Making a connection between health and affluence, someone might reason:
There have been improvements in the health of the population over the past thirty
years, a period during which there has been an increase in the affluence of the
country. So the increased affluence of the country has produced the improvements
in the health of the population.
The question as to whether increased affluence has or has not produced improvements in
the health of the population cannot be settled without more evidence – evidence both
about the incidence of all illnesses in the population, and about whether any improvements
could not have occurred if there had not been greater affluence. The argument
simply assumes, without producing any evidence for it, that because two things have
occurred together, one of them must have caused the other.
This unwarranted assumption of a causal connection often occurs when someone discovers
a correlation – that is, a connection between x and y such that whenever you find x,
you are likely to find y, or such that whenever a person or a population has characteristic x,
they are likely to have characteristic y. For example, suppose you find that children who
frequently watch violent videos are likely to be aggressive; this may be because watching
violent videos causes children to be aggressive, or it may be because having a natural
tendency to aggressive behaviour causes children to enjoy watching violent videos. Or
suppose you find that people who have a great deal of tooth decay are likely to be overweight.
This may be because a third factor – perhaps eating large amounts of sugary foods
– causes both these conditions. All that you have found when you have discovered a
correlation is that two things occur together. This may be because x causes y, or because
y causes x, or because x and y are both caused by something else, or it may be simply
a coincidence. You are guilty of flawed reasoning if you just assume, without further
evidence that x causes y.
Nevertheless it is important to note that discovering correlations is not a pointless exercise.
It is often the first step in the attempt to investigate whether there is a causal connection
between two phenomena.
Example 4: Exhaustion of mineral resources
Here is our last example in this section.
It has always been the case in the past that new discoveries of mineral resources have
kept pace with demand. For example, bauxite reserves have tripled in the last ten
years, while demand has doubled over the same period. At no time have the known
reserves of minerals been as great as the total mineral resources of the world. Therefore,
even though at any given time we know of only a limited supply of any mineral,
we can be confident that there is no imminent danger of our running out of mineral
resources.
Before reading on, identify the conclusion and the reasons in this argument, and try to
state for yourself what is going wrong in moving from the reasons to the conclusion.
The main conclusion, clearly signalled by ‘Therefore’, is the final sentence. The argument
can be regarded as having the following structure:
Reason 1: It has always been the case in the past that new discoveries of mineral
reserves have kept pace with demand.
Reason 2: At no time have the known reserves of minerals been as great as the total
mineral resources of the world.
These two reasons, taken together, are intended to support:
Main conclusion: Therefore, even though at any given time we know of only a limited
supply of any mineral, we can be confident that there is no imminent danger of our
running out of mineral resources.
Note that the example presented in the second sentence is being used to give some support
to Reason 1. But we have not shown this as a reason from which Reason 1 follows, because
one example could not be sufficient to establish a general claim such as Reason 1, nor is it
likely that the author of the argument thinks that the example does establish the general
claim. It is being used in an illustrative way. The second sentence could be included in the
argument structure by simply treating it as a part of Reason 1.
Since this argument is clearly relying on past experience, it may be tempting to describe
the flaw as an assumption that what has been true in the past will continue to be true in
the future. But this assumption underlies many arguments, particularly those relying
on laws of science (for example that at sea level, water boils at 100 degrees Celsius), and in
many such contexts, it is not an unreasonable assumption to make.
So we need to state the flaw more specifically. Why should we not conclude that there is no
imminent danger of running out of mineral resources, based on evidence that at any given
time in the past the known reserves of minerals have not been as great as the total mineral
resources of the world? The answer is that it is reasonable to assume that the mineral
resources of the world are finite, and thus that if they continue to be used they will run out
at some time in the future, and we cannot know when that point will be reached. We can
state the flaw as follows:
Assuming that the mineral resources of the world are finite, at present (and at any
given time in the future) the total mineral resources of the world may be no greater
than the known reserves of minerals.
A further point could be made about this argument. It makes no mention of evidence
which may be relevant – i.e. a possible acceleration in the rate at which mineral resources
are being used.