Non-communicable Evidence

In this video, Dou­glas Crock­ford (JavaScript MASTER) says:

So I think pro­gram­ming would not be pos­si­ble with­out Sys­tem I; with­out the gut. Now, I have ab­solutely no ev­i­dence to sup­port that state­ment, but my gut tells me it’s true, so I be­lieve it.


I don’t think he has “ab­solutely no ev­i­dence”. In wor­lds where DOUGLAS CROCKFORD has a gut feel­ing about some­thing re­lated to pro­gram­ming, how of­ten does that gut feel­ing end up be­ing cor­rect? Prob­a­bly a lot more than 50% of the time. So ac­cord­ing to Bayes, his gut feel­ing is definitely ev­i­dence.

The prob­lem isn’t that he lacks ev­i­dence. It’s that he lacks com­mu­ni­ca­ble ev­i­dence. He can’t say “I be­lieve A be­cause X, Y and Z.” The best he could do is say, “just trust me, I have a feel­ing about this”.

Well, “just trust me, I have a feel­ing about this” does qual­ify as ev­i­dence if you have a good track record, but my point is that he can’t com­mu­ni­cate the rest of the ev­i­dence his brain used to pro­duce the re­sult­ing be­lief.


How do you han­dle a situ­a­tion where you’re hav­ing a con­ver­sa­tion with some­one and they say, “I can’t ex­plain why I be­lieve X; I just do.”

Well, as far as up­dat­ing be­liefs, I think the best you could do is up­date on the track record of the per­son. I don’t see any way around it. For ex­am­ple, you should up­date your be­liefs when you hear Dou­glas Crock­ford say that he has a gut feel­ing about some­thing re­lated to pro­gram­ming. But I don’t see how you could do any fur­ther up­dat­ing of your be­liefs. You can’t ac­tu­ally see the ev­i­dence he used, so you can’t use it to up­date your be­liefs. If you do, the Bayes Po­lice will come find you.

Per­haps it’s also worth try­ing to dig the ev­i­dence out of the other per­sons sub­con­scious.

  • If the per­son has a good track record, maybe you could say, “Hmm, you have a good track record so I’m sad to hear that you’re strug­gling to re­call why it is you be­lieve what you do. I’d be happy to wait for you to spend some time try­ing to dig it up.”

  • Maybe there are some tech­niques that can be used to “dig ev­i­dence out of one’s sub­con­scious”. I don’t know of any, but maybe they ex­ist.


    Ok, now let’s talk about what you shouldn’t do. You shouldn’t say, “Well if you can’t provide any ev­i­dence, you shouldn’t be­lieve what you do.” The prob­lem with that state­ment is that it as­sumes that the per­son has “no ev­i­dence”. This was ad­dressed in Sec­tion 1. It’s akin to say­ing, “Well Dou­glas Crock­ford, you’re tel­ling me that you be­lieve X and you have a fan­tas­tic track record, but I don’t know any­thing about why you be­lieve it, so I’m not go­ing to up­date my be­liefs at all, and you shouldn’t ei­ther.”

    Brains are weird and fan­tas­tic thingys. They pro­cess in­for­ma­tion and pro­duce out­puts in the form of be­liefs (amongst other things). Some­times they’re nice and they say, “Ok Adam—here is what you be­lieve, and here is why you be­lieve it”. Other times they’re not so nice and the con­ver­sa­tion goes like this:

    Brain: Ok Adam, here is what you think.

    Adam: Awe­some, thanks! But wait—why do I think that?

    Brain: Fuck you, I’m not tel­ling.

    Adam: Fuck me? Fuck you!

    Brain: Who the fuck do you think you’re talk­ing to?!!!

    Just be­cause brains could be mean doesn’t mean they should be dis­counted.