SilasBarta comments on Open Thread June 2010, Part 3 - Less Wrong

6 Post author: Kevin 14 June 2010 06:14AM

You are viewing a comment permalink. View the original post to see all comments and the full post content.

Comments (606)

You are viewing a single comment's thread. Show more comments above.

Comment author: JoshuaZ 14 July 2010 12:45:35AM *  4 points [-]

Sometimes I almost think this is a signaling issue, where mathematicians want to display that they don't need such crutches.

This isn't really a signaling issue so much as a response to the fact that mathematicians have had centuries of experience where apparent theorems turned out to be not proven or not even true and the failings were due to too much reliance on intuition. Classical examples of this include how in the 19th century there was about a decade long period where people thought that the Four Color Theorem was proven. Also, a lot of these sorts of issues happened in calculus before it was put on a rigorous setting in the 1850s.

There may be a signaling aspect but it is likely a small one. I'd expect more likely that mathematicians err on the side of rigor.

ETA: Another data point that suggests this isn't about signaling; I've been too a fair number of talks in which people in the audience get annoyed because they think there's too much formalism hiding some basic idea in which case they'll ask questions sometimes of the form "what's the idea behind the proof" or "what's the moral of this result?"

Comment author: SilasBarta 14 July 2010 12:53:42AM 2 points [-]

Over 99% of students learning math aren't going to be expected to contribute to cutting-edge proofs, so I don't regard this as a good reason not to use "plain English" methods.

In any case, a plain English understanding can allow you to bootstrap to a rigorous understanding, so more hardcore mathematicians should be able to overcome any problem introduced this way.

Comment author: JoshuaZ 14 July 2010 12:56:55AM 1 point [-]

I agree that this is likely often suboptimal when teaching math. The argument I was presenting was that this approach was not due to signaling. I'm not arguing that this is at all optimal.