Transcript The Problem of Induction

Science and induction
Science and we assume causation (cause and
effect relationships)
 For empiricists, all the evidence there is for
empirical knowledge, including science,
concerning “matters of fact” is sensory
experience
 For some, we move from individual
experiences/singular statements to
generalizations/universal statements using
induction (and we certainly often do).
The Problem of Induction
David Hume
Empirical generalizations
1. Millions of ravens have been observed and all are
black.
2. A non-black raven has never been observed.
--------------------------------------------------------3. All ravens are black
 Are, like other forms of inductive arguments,
ampliative
 Reasoning moves from the past and present to the
future
 From what has been experienced to what has not
Hume’s question
What justifies our use of induction?
There are, he imagines, two possibilities:
Experience, which concerns matters of fact
Reason, which concerns relations of ideas
And he proposes that we explore each to see
if the justification lies there
The inference
What does my past or present knowledge about
some kind of object (e.g., that bread has always
been nourishing) suggest about my next
encounter with that kind of object?
 Put another way, propositions of the form:
What we have all experienced that X causes Y, and
X will always cause Y
Are very different
What justifies such the inference from the first to
the second?
Reason (aka Demonstrative
Knowledge)
There is no necessary connection between
I’ve always experienced that X causes Y.
I foresee that the next X I encounter will cause Y.
It is logically possible that however many my
experiences of X causing Y, it won’t next time,
or next week, or next July…
 Inductive arguments are not deductively valid.
Experience
 Can we appeal to our past experience using
induction to justify our use of it?
 After all, while not deductively valid, many
inductive arguments seem strong and have
proven helpful.
1. Induction has worked in the past.
So, our use of induction is justified.
 Appealing to an inductive argument (that
induction has worked in the past) to justify
induction is circular.
The “nature” of Nature:
an added premise
1.
Induction has worked in
Induction will the past.
2. Nature is uniform.
----------------------------------work in the future.
1. The sun rose today.
2. The sun rose yesterday.
n. The rose on n.
3. Nature is uniform
----------------------------------The sun will rise tomorrow.
The “nature” of nature
What justifies the premise:
Nature is uniform
It was uniform in the past and it is uniform
in the present.
That is, it is also the conclusion of an
inductive argument.
Is this all about “secret powers”?
Is the problem just lack of knowledge?
Say, on Hume’s part, about why bread
nourishes?
Say, about why the law of the conservation
of energy holds?
The argument: No matter how much we
learn, the problem will apply to that
knowledge as well.
Can evolutionary theory help?
Induction assumes that the universe contains
“kinds” and causation
It also assumes the uniformity of nature
Reasoning this way is useful because it
underlies predictions.
Other species also use induction.
Is it possible that the capacity and/or
disposition is the product of natural selection?
Can evolutionary theory help?
Is it possible that the capacity and/or
disposition is the product of natural selection?
Yes and cognitive scientists maintain it likely
is.
If this is the case, does it justify induction?
No: the problem remains.
And the past success of induction does not
guarantee its success in the future.