How to disagree about how to disagree

A few people read this blog to hear about my thoughts on the Orioles or various rock shows, but let’s face it, most of you are just here because you want to know: what is the right philosophy lecture with which to kick off my weekend? Well, you’re in luck! This semester’s Math And… lecture will be delivered by philosopher Adam Elga, this Friday at 4pm. Elga is a fascinating guy who thinks a lot about quantitative problems arising from the careful study of decisions, knowledge, and belief. His papers include “I Can’t Believe I’m Stupid,” Defeating Dr. Evil with Self-Locating Belief, and “Bayesianism, Infinite Decisions, and Binding.” This last paper features a puzzle I like a lot: A dead man is in hell. On his first day, God comes to him and says, “There’s been a mistake, you actually belong in Purgatory. To make it up to you, after today, you can go to Heaven for two days and then spend the rest of eternity in Purgatory — or if you like, you can go straight to Purgatory now.” The dead man figures two days in Heaven is worth one more day in Hell, so he takes the deal. Next day, God comes back: “I’ve got another deal for you. If you spend one more day in Hell, you can have two extra days in Heaven afterwards, so you get four days in all.” Well, two days in Heaven is worth one day in Hell, so the man accepts again. You see where this is going: every day the dead man gets the offer, every day he does the rational thing and accepts, and as a consequence he spends all eternity in Hell. What went wrong?

On a not completely unrelated note, Elga was a participant in the Big Number Duel, in which he went head-to-head with a colleague from MIT to see who could write the biggest number on the board. (He lost.)

Anyway, here’s the information for the talk:

LOCATION: Helen C. White Hall, room 4281

TIME: 4pm, Friday, Oct 26

TITLE: How to disagree about how to disagree

ABSTRACT: When one encounters disagreement about the truth of a
factual claim from a trusted advisor who has access to all of one’s
evidence, should that move one in the direction of the advisor’s view?
Conciliatory views on disagreement say “yes, at least a little”.
Such views are extremely natural, but they give bad advice when the
issue under dispute is disagreement itself. So conciliatory views
stand refuted. But despite first appearances, this makes no trouble
for partly conciliatory views: views that recommend giving
ground in the face of disagreement about many matters, but not about
disagreement itself.

Tagged , , ,

3 thoughts on “How to disagree about how to disagree

  1. Dirty Davey says:

    Couldn’t he have won the large number duel by writing, “The number on the other guy’s board, plus one”?

    Also, what about the “Colin Powell” reaction to the leading premise of the abstract? “When one encounters disagreement about the truth of a factual claim from a trusted advisor…”, maybe one’s trust in the advisor is misplaced and should be reevaluated?

  2. John Cowan says:

    The answer to the puzzle is simple. Purgatory, as everybody knows, is not an eternal state at all, but a temporal one — after your sins are purged, you go from Purgatory to Heaven. Therefore, this so-called “God” is actually a Satanic deception, and the man will spend eternity in Hell, exactly as the real God foreordained.

  3. [...] who is that most powerful of creatures, a philosopher who knows Bayes’ Theorem. (See also: Adam Elga, K. Anthony Appiah.) Sober’s title is TBA, but he may well talk about (or, more likely, [...]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Follow

Get every new post delivered to your Inbox.

Join 546 other followers

%d bloggers like this: