The Fallacy Of Inductive Reasoning

If you've heard of inductive reasoning getting distinguished from deductive reasoning, then you know that the two are not identical.  For the purposes of this post, by inductive reasoning I mean beliefs that because some of something are a certain way, all of those things must be the same way.  By deductive reasoning I mean the use of premises to reach conclusions (example: 1) If X, then Y; 2) X; 3) Therefore Y).  Inductive reasoning can be somewhat more complex than the definition provided here, but it is almost always used as the basis of probability judgments, whereas deductive reasoning is used to form infallible proofs (at least infallible proofs that certain conclusions follow from certain propositions).

I will show two things in this post: 1) that inductive reasoning is fallacious and 2) that recognizing the invalidity of inductive reasoning can help Christians defend Christianity.  Inductive reasoning represents a manifestation of the fallacy of composition.  For instance, past precedent does not necessitate future conformity to what has always happened or seems to have happened.

I will list some examples of such inductive reasoning:

Just because I woke up every day in my past without missing an arm does not mean that when (or if) I wake up tomorrow morning I will not be missing an arm.

Just because I have always fallen back to the ground after jumping does not mean I will fall back to the ground when I next jump.

Just because my car's fuel gauge has always gauged the car's gas level at predictable rates does not mean that my car will use gas or that the gauge will report it at the same rate.

With several short examples presented, entertain the following hypothetical scenario with me.  A man named Nathan gets abducted by terrorists and held by them for 300 days without being killed, and Nathan begins to think by the 300th day that he will survive the next day as well.  But on the 301st day, something he did not expect happens--his captors shoot him.  Nathan fallaciously assumed that because he had not been shot on any of the prior days of his captivity that he would not be shot on the 301st day.  The similarities of some days--even the majority of days during his captivity--did not mean all of the days must be alike.

Here I have adapted Bertrand Russell's example of an "inductivist turkey" waking up and being fed at 9:00 in the morning long enough to feel secure that he would always be fed at 9:00.  But on a particular morning by which time he has grown content with his inductive reasoning, his throat is sliced open and he is killed, in direct contradiction to his inductivist expectations.

I am not saying that all inductive conclusions are false, only that inductive reasoning as defined here is unreliable!  In fact, pointing this out can actually help show the logical possibility of things some may otherwise consider totally impossible, such as, say, the resurrection of Christ.  Some people who erroneously rely on science as their primary epistemological foundation may realize that it seems that people, after biologically dying, do not return to life.  Suppose a scientist named Susan understands that bodily resurrection is not a natural biological phenomenon and on such grounds rejects the resurrection of Christ as impossible.  She has succumbed to the inductive fallacy on the assumption that Christ cannot have resurrected because all other dead people she has heard of have not resurrected.  What is true of the part is NOT necessarily true of the whole when dealing with matters of science and everyday experience.

Be careful what predictions you make of the future, for a being with my limitations has no basis by which to judge it necessary that everything about the future will mirror the way the past has unfolded (infallible axioms and self-evident truths will remain true by necessity, so it is not as if nothing at all is perfectly predictable).  Inductive reasoning of this type commits the fallacy of composition by concluding that what is true of a part (or many parts) must be true of the whole of something.  Deductive reasoning, not inductive reasoning, is inherently reliable.