Tuesday, July 03, 2012

Sorting the Argument from the Conclusion

I've mentioned this before, but one thing a bit of training in analytical philosophy does is teach you to differentiate a good argument from an argument that reaches a conclusion that you like. Once you learn to do that, you become amazed at how often people judge arguments by their conclusions: The argument from the idea of universal human rights makes the case against slavery, therefore it must be a good argument! Hoppe's argumentation ethics advocates libertarianism, therefore it's a good argument!

Furthermore, it seems remarkably easy for people to think that if you reject an argument, you reject the argument's conclusion. But that does not follow at all: For instance, Bob Murphy and I wrote our critique of Hoppe's argumentation ethics at a point we both heartily endorsed his conclusion: We just didn't think the argument for it was any good.

I think that some readers experienced similar difficulties with my recent critique of Daniel Dennett. One commentator, in fact, said that by analogy with my argument about the thermostat, it seems one could prove cars can't move, since they are built of non-moving parts.

What a muddle! Nowhere in my three pieces did I ever claim that:
1) Computers don't think; or
2) That I had proven that computers don't think.

In fact, I have no idea if my computer has any thoughts at all. No, that's not quite true: I feel strongly that when I run "Shut down," it mulls over whether or not it really wants to comply for quite some time.

What I do know is that Dennett's argument is bad: He shows us a circuit that, he himself claims, lacks all understanding. Then to explain how understanding comes about, he waves his hands around and says "Just connect a whole bunch of these circuits together!"

My critique never claims anything remotely like "A machine built out of parts that don't think can't think!" or "A machine built out of parts that don't move can't move!" Instead, what I do claim is that, if I ask you how a car is able to be "automotive," it is a terrible explanation for you to say, "See this carburetor that is not automotive? And this spark plug that is not automotive? And this piston that is not automotive? Well, put enough parts like that together, and you have an automobile!"

Obviously, this has explained nothing at all. And so with Dennett's argument: his conclusion may be correct, but his argument is terrible.


  1. Phew, it's good thing I never said anything like that about your argument!

    1. Silas, I imagine you intended to be offensive, but given I have no idea of what you speak, you didn't succeed.

  2. I don't know whether I'm one of the commenters you had in mind, but my comment wasn't really meant to be a critique of your argument. I was just clarifying that introducing a set of psychophysical laws (as opposed to sticking with reductive materialism) would remove the difficulty.

    1. No, Huff, in fact I made a similar point: the problem was not the general idea of building up mind from mind components of some sort -- I mentioned it in the context of panpsychism -- but in the fact that all Dennett offered about *how* this is done is hand waving.


Euphemisms II

It's not just abortion where these euphemisms bug me either. I think it is sometimes OK to euthanize your pet, but... you did not "...