98 SMART THINKING: SKILLS FOR CRITICAL UNDERSTANDING & WRITING
example, I had 'unluckily' chosen a sample of 1000 Australians who were unusually
pro-republic, then it is unlikely that my broad conclusions would be correct. So,
good reasoning from specific cases requires some consideration of the degree to
which the cases selected represent the general category as a whole.
This question of 'representativeness' is precisely why reasoning from specific
cases needs to proceed on a sound base of'specifics'. If I were to argue, on the basis
of one bad meal of Italian food, that ^//Italian food was bad, I would be relying on
far too small a sample for my argument to be effective. Equally, we should not trust
surveys that rely on large numbers of responses from an unrepresentative group. For
example, television stations have taken to conducting 'polls' in which people ring
in to answer 'yes' or 'no' to a particular question (for example, 'Should the death
penalty be reintroduced?'). The answer is then represented as a good generalisation
of all Australians' attitudes when, in fact, it is only a generalisation of the views of
those viewers of that particular television station who were able and willing to ring
in.Reasoning from analogy
An analogy is a special form of reasoning, which has some similarities with
reasoning from specific cases. Reasoning by analogy involves drawing an equally
specific conclusion from specific premises via a comparison of like aspects. Good
analogies avoid comparisons between items that have too many dissimilarities. For
example:
Imagine a friend gave you a guinea pig to look after but forgot to tell you
anything about what to feed it. You might say to yourself, 'I have a guinea
pig and do not know what to feed it; but I do know that my rabbit eats
carrots, and that rabbits and guinea pigs are similar. Hence, I can prob-
ably feed my guinea pig carrots as well'.
Such arguments take the following general form:
An analogy between X and Y (in the premises) supports a conclusion about
Y by showing that the conclusion is true of X; and X and Y are similar in
sufficient relevant respects and are not relevantly dissimilar.
You need to be careful to make sure that you are comparing things that are
similar in a relevant way. Take the following example of reasoning:
Shaving cream is clearly similar in colour, texture, moistness, and body to
whipped cream, and I know that whipped cream is delicious on fruit salad.
Hence, shaving cream is delicious on fruit salad.
Do you see what is wrong? The two types of cream are similar, but they are
definitely not similar in respect of the one main characteristic involved in fruit salad
eating: how they taste. This question of relevance has been explored in more detail
in chapter 6.