July 29, 2008

3 arguments.... 1 logic error.

But can you explain the error?


Michael E. Lopez said...

It's not really the same type of error in all three posts, unless we're going to generalize to the point where there are really only three or four types of error.

The first one contains two errors: a potentially problematic extraction of principle from the antecedent, and an application of that principle without regard to potentially counterveiling but relevant principles (such as the effect of an action on the actor, not only the victim).

The second one contains an error that could be seen in two different ways. Viewed one way, it's the universalization of a principle without regard to the inherent scope limitations of that principle (i.e., is there something in the principle of domestic redistribution that allows it to be universalized, or is there something that prevents it such as a relationship requirement?). Viewed another way, the error is once again the dismissing of counterveiling and potentially relevant principles.

Finally, the third one is obviously ignoring the counterveiling interests that the subject of the argument might have -- in other words, it is ignoring the fact that corporations may have interests other than profit (such as obeying the laws that make their own existence possible).

Viewed VERY broadly, these could all be seen as "ignoring counterveiling interests and principles," but that's so vague as to hardly warrant pointing out the similarity in the first place.

KLDAVIS said...

It's sort of fallacy of idealism...that's an ill-fitting word for the first and third cases, but it is idealism nonetheless.

The third case is close to the plot of a very enjoyable novel, Jennifer Government.

KLDAVIS said...

Michael's post makes me think that JAC might be getting at the idea that actors undervalue the costs of their actions as applied to themselves. It seems to apply to all three scenarios...the guilt of the torturer/murder, the cost of such a bureaucracy.

Joe M. said...

I'm not certain that a corporation does have any interest other than profit. It seems that the basic function of a corporation (in the abstract, at least) is to make money by any means necessary, and that obedience to law is really only playing along to a governmental protection racket.

(P.S. I'm on the side of society here, so please forgive me for sounding like some of my more scarily libertarian friends ...)

blake said...

The logical fallacy of believing people care about what gets argued on the Internet?

Heh. No, they all seem to exhibit scaling problems. I don't know if that's a logical fallacy or an engineering problem, though.

Ralph said...

IIRC, Italian Renaissance-era merchants were literally cut-throat competitive. Commerce was tied even closer to power politics than it is now, partly because Adam Smith hadn't explained to them that it wasn't a zero-sum game.

It looks like all three arguments take a widely accepted idea: punishment, redistribution, and competition, to ridiculous extremes.

Ralph said...

OK, lefties don't think #2 is ridiculous, but I do.

Joe M. said...

The second post isn't ridiculous: the author is arguing against luck-egalitarianism by showing the ridiculous extremes that result from its basic premises and pointing out that even (most) luck-egalitarians don't want to carry their professed beliefs that far.

Scott Colom said...

The logical error is the assumption that 1) torturing prisoners 2) redistributing wealth around the world and 3) having a lot of corporations kill competitors would work. We may not torture criminals because it wouldn't deter crime; or redistribute all wealth because it wouldn't the unluck inherent in people's birth; and killing competitors would not stop competition.

Are you going to give your explanation of the logical error?

John Althouse Cohen said...

Thanks for the analysis, everyone!

I plan to do another post to go into the arguments in detail, but I don't have time now.

Michael E. Lopez's "ignoring countervailing interests" is pretty close. Ralph also made a good point: they all "take a widely accepted idea ... to ridiculous extremes." Put these two together, and that's pretty much what I was thinking: they all take a principle and stretch them out to the extreme, without considering that there are concrete obstacles that get in the way of the abstract principles.

John Althouse Cohen said...

The second post isn't ridiculous: the author is arguing against luck-egalitarianism by showing the ridiculous extremes that result from its basic premises and pointing out that even (most) luck-egalitarians don't want to carry their professed beliefs that far.

Are you just assuming that he's correctly extending the egalitarians' views? I wouldn't assume that. If he's incorrect in how he applies liberals' views, then he's still making an error, even if though he himself is not a liberal (which he's not).

Joe M. said...

I'm not assuming that he's correctly extending the views of liberals. I think he is correct, but that's based on evaluation of his argument, not blind assumption.

If you don't think he's correct, I'd love to hear an explanation.

demian said...

The shared logic error is:

If modest pursuit of a goal is a commendable, then extreme pursuit is doubly virtuous.

Each proposal raises new problems.

* Torture: An extreme focus on suffering -- one of incarceration's aims -- may obscure or derail co-existing goals, such as rehabilitation.

* Global income redistribution: Extreme redistribution requires a world government. What form will it take: Democracy? Dictatorship? Communism? Many forms of governance actually suppress income generation.

* Assassinating CEOs:
An extreme focus on the balance sheet ignores real expenses. The cost of a hitman includes not just his fee but the risk of jail time and fatal damage to a company's reputation.

Paul Brinkley said...

I think multiple people grasped the motif here. My wording of the fallacy: if a factor is overwhelming in one case, then it is overwhelming in all cases.

It's like the equation x + 1/x. If x is bigger then 1, then the first factor dominates. But that doesn't mean it always does - x could be between 0 and 1.