WHAT KINDS OF REASONING ARE THERE? 97When we are confident about our generalisations, and a specific case does fit a
particular general category, then reasoning from generalisations is very easy and
effective. The trick is in making that initial judgment about the relationship
between the specific case and the generalisation, as expressed in the premises, so
that our conclusion (also about the specific case) is well supported. At the same
time, we should recognise that many items fit a number of generalisations—that
often an 'item' has many conflicting qualities or components that make it hard to
judge the appropriate generalisation. For example, doctors are often confronted
with a series of symptoms in a patient that could mean any one of a number of
illnesses. The tests that doctors perform are designed to work out exactly which
'generalisation' to apply to the specific patient and thus make sure that the correct
treatment is prescribed.
Reasoning from specific cases
Where do these generalisations come from? Do we just make them up? No, in most
cases they have been established via reasoning—in this instance, from specific cases
to a generalisation. The difference in reasoning from specific cases is that, although
a general statement is involved, it is not used as a premise but as the conclusion.
We routinely find such reasoning in, for example, opinion polls, statistical analyses,
or any other surveys in which the reasoning supports conclusions that generalise
beyond the specific scope of those premises. For example, I might argue from
specific cases in this manner:
I have surveyed 1000 Australians, from all social classes and ethnic back-
grounds, and 70 per cent of them tell me that they favour changing from
a monarchy to a republic. Hence, I would conclude that most Australians
also support this change. Ninety per cent of.the respondents who were
born overseas or whose parents were born overseas were positive about
Australia becoming a republic. Hence, I further conclude that republican
sentiment will be strongest among the newer members of the Australian
community.There are two conclusions here; each is a general statement about what all
Australians think, based on a sample of 1000 specific cases. The premises provide
a summary of the many specific cases. Nevertheless, relative to the premises, the
conclusions are much broader assertions ('most Australians' and 'newer members').
The general form of such arguments is as follows:
Specific cases (x) of the general category X show the common property A;
hence, generally speaking, we can expect all members of the category X to
have the common property A.
Reasoning from specific cases depends on the same sorts of judgments about
the underlying relationship between the cases and the general category that we
encountered in the previous section on reasoning from generalisations. If, for