You Can't Really Punt on Abduction, Stef, Part II

On Sunday, I said something about Stefan Molyneux's new book. He purports to discuss the art of argument, and runs through certain basics about logic but ... punts on the Peircean principle of abduction. I'll address Stefan directly hereafter. The rest of you may listen in.

---------------------------

This is how you define abduction:

"There is another category [of inference] called abductive reasoning that draws a tentative hypothesis from disparate data, but which is related to some sort of testable hypothesis, rather than the reaching of a specific conclusion."

Um ... no. You seem to have simply thrown together some words here without thinking about why or whether they even make any sense together. An abduction won't get you to a certain or dogmatic conclusion, but it doesn't follow that the conclusion must fail to be "specific." Also, what does it mean for data to be "disparate," why would the data on which an abduction be based be expected to be "disparate," what does disparateness have to do with it?

Again: the "some sort of" language is just the jelly you're pumping into a word donut here, isn't it?

Let's start from the beginning. What IS an abduction? You seem not to know. It isn't hard to find out though. Search engines are helpful. So for that matter is wikipedia.

An abduction is the inference from data to the simplest plausible explanation of that data. It is that simple. Skip the donut manufacturing exercise.

Here's an equally simple example: If I'm walking down the street on a wintery day I might turn a corner  and see Sally with snow on her face, and Jane a few feet away, laughing at Sally, I might infer that Jane has just thrown a snowball at Sally, and is amused by her own success. That is plausible. It explains the facts. It doesn't require extraordinary supportive hypotheses. And so forth.

Note that there is nothing "disparate" about the data, and the conclusion is perfectly "specific." Yes, an abductive hypothesis is usually in principle testable. There may be surveillance footage of the scene that I could examine. But whether a specific hypothesis is as a practical matter testable is very circumstance sensitive.

Still, as Charles Peirce wrote back in 1902: "The surprising fact, C, is observed. But if A were true, C would be a matter of course. Hence, there is reason to suspect that A is true."

A lot more might be said about abduction. Some philosophers have contended that the whole idea is confused, though that view seems to have died out of late and AI research proceeds largely on Peircean assumptions about how inference works. At any rate: you seem not to think that IT is confused, but you seem to be confused ABOUT it.

Well, do some research the next time you try writing a nonfiction book, okay?