Today's post, Fallacies of Compression was originally published on 17 February 2008. A summary (taken from the LW wiki):
You have only one word, but there are two or more different things-in-reality, so that all the facts about them get dumped into a single undifferentiated mental bucket. It's part of a detective's ordinary work to observe that Carol wore red last night, or that she has black hair; and it's part of a detective's ordinary work to wonder if maybe Carol dyes her hair. But it takes a subtler detective to wonder if there are two Carols, so that the Carol who wore red is not the same as the Carol who had black hair.
Discuss the post here (rather than in the comments to the original post).
This post is part of the Rerunning the Sequences series, where we'll be going through Eliezer Yudkowsky's old posts in order so that people who are interested can (re-)read and discuss them. The previous post was Replace the Symbol with the Substance, and you can use the sequence_reruns tag or rss feed to follow the rest of the series.
Sequence reruns are a community-driven effort. You can participate by re-reading the sequence post, discussing it here, posting the next day's sequence reruns post, or summarizing forthcoming articles on the wiki. Go here for more details, or to have meta discussions about the Rerunning the Sequences series.
I think speaking more than one language (from distant language groups) does help a fair lot. You see how a continuum can be bucketed differently, among other things.
Other big issue is that great many people actually see a form of compression as some amazing virtue of "Abstract thought".
That is itself an example of compression - mixing together the human capacity to create useful abstractions, to be able to make predictions about the real world by using very little computational power, with the human brain limitations that disallow fine grained modelling and which in fact greatly impede our ability to make predictions about the real world.
While it is awesome that we can describe weather in terms of different cloud types, the most accurate weather simulation is the one working at the lowest abstraction level that we have the computational power to work at - cells with air properties, the smaller the better - the division into cloud types only helps when you are pretty much incapable of modelling clouds to any extent.
I propose that one should try to think of things as non-abstractly as one can (except for very special cases where abstractions are surprisingly close to the territory). The less stuff you abstract out the less likely you are to make dumb mistake or over the chain of reasoning slowly diverge from correctness.