I'm interested in knowing whether all ravens are black.  Armed with my trusty binoculars, I've traveled to the ends of the earth, viewing every raven I could catch sight of.  Every raven I've seen has been black, and I've seen a lot of ravens.  It follows that all ravens are black.

Is this a good argument, is it truth-preserving?  Let's put it to the following test: Can we imagine the premises of the argument being true and the conclusion false?  I can!  Perhaps when my back was turned a non-black raven sneaked by.  The truth of the premises of my argument doesn't guarantee the truth of the conclusion.  So is this a bad argument?  If our standard to determining the goodness or badness of an argument is truth-preservation, then we have to conclude that this is a bad argument. But that clearly won't do.

What we need is a distinction between two kinds of argument. The ravens argument isn't intended to establish a guarantee that the conclusion is true, given the evidence. Rather, the evidence presented makes it more likely that the conclusion is true than it would be without the evidence.  We call arguments where the intended support for the conclusion is of this sort inductive arguments.

Inductive inferences are very common.  I believe that my car is in the parking lot. What's my argument? I know I parked it there this morning, and that cars tend to stay put when they've been parked.  So I'm comfortable in concluding that my car is still in the lot. My evidence makes it likely that the conclusion is true, but it doesn't provide an ironclad guarantee.  I might have left it in neutral, and it rolled down the hill or it might have been stolen.

Deductive arguments are arguments whose premises purport to provide a guarantee of the truth of the conclusion, given the truth of the premises.  I say "purport" because not all deductive arguments actually fulfill that guarantee.  This course in formal logic deals exclusively with deductive inference. There are many reasons for this limitation.  Logicians understand deductive inference better than they understand inductive inference.  In particular, we stand a much better chance of characterizing deductive inferences through the form of deductive arguments than we do when we look at inductive reasoning.

To determine whether an argument is deductive or inductive, you often need to assess what the purported type of support is posited for the conclusion.  Sometimes the text makes this explicit. If the conclusion indicating words contains "probable" or "likely", then the argument is probably (!) inductive.

Exercise: Inductive or deductive inference?

There's a great deal of work in inductive logic, and to give you just a hint of it, I've included an appendix on Inductive Logic.  (The material covered in this appendix is not part of the material for the course.)