Less Wrong is a community blog devoted to refining the art of human rationality. Please visit our About page for more information.

theguyfromoverthere comments on Words as Hidden Inferences - Less Wrong

40 Post author: Eliezer_Yudkowsky 03 February 2008 11:36PM

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

Comments (22)

Sort By: Old

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

Comment author: theguyfromoverthere 20 June 2014 05:26:59PM 0 points [-]

I think this is in the context of somebody insisting that Socrates is human so he must be mortal.

If you are trying to prove mortality by claiming he's human, then all humans must be mortal for you to assume this.

I agree, though, that, perhaps the statement was a little vague.

Comment author: CBHacking 04 January 2016 01:52:31AM 0 points [-]

Replying loooong after the fact (as you did, for that matter) but I think that's exactly the problem that the post is talking about. In logical terms, one can define a category "human" such that it carries an implication "mortal", but if one does that, one can't add things to this category until determining that they conform to the implication.

The problem is, the vast majority of people don't think that way. They automatically recognize "natural" categories (including, sometimes, of unnatural things that appear similar), and they assign properties to the members of those categories, and then they assume things about objects purely on the bases of appearing to belong to that category.

Suppose you encountered a divine manifestation, or a android with a fully-redundant remote copy of its "brain", or a really excellent hologram, or some other entity that presented as human but was by no conventional definition of the word "mortal". You would expect that, if shot in the head with a high-caliber rifle, it would die; that's what happens to humans. You would even, after seeing it get shot, fall over, stop breathing, cease to have a visible pulse, and so forth, conclude that it is dead.. You probably wouldn't ask this seeming corpse "are you dead?", nor would you attempt to scan its head for brain activity (medically defining "dead" today is a little tricky, but "no brain activity at all" seems like a reasonable bar).

All of this is reasonable; you have no reason to expect immortal beings walking among us, or non-breathing headshot victims to be capable of speech, or anything else of that nature. These assumptions go so deep that it is hard to even say where they come from, other than "I've never heard of that outside of fiction" (which is an imperfect heurisitic; I learn of things I'd never heard about every day, and I even encountered some of the concepts in fiction before learning they really exist). Nobody acknowledges that it's a heuristic, though, and that can lead to making incorrect assumptions that should be consciously avoided when there's time to consider the situation.

@Caledonian2 said "If Socrates meets all the necessary criteria for identification as human, we do not need to observe his mortality to conclude that he is mortal.", but this statement is self-contradictory unless the implication "human" -> "mortal" is logically false. Otherwise, mortality itself is part of "the necessary criteria for identification as human".