[Previous] John Locke's Politics | Home | [Next] Stop Saying Lies and Other People's Ideas

Teachable Subjects

in some sense, people teach chess, math, Spanish, grammar, history, programming, formal and symbolic logic, chemistry, physics, biology, how our government works, how to cook.

they have school classes on these things. they have books explaining it. i can remember being taught about these things in the past.

by contrast, i cannot remember anyone ever teaching me:

  • how to understand what a sentence means
  • how to figure out if an argument is true or false
  • what an argument IS (and how to decide if X is an argument for Y)
  • how to decide if something is a non sequitur
  • how to decide if X implies Y
  • how to read sentences that don't not have double, triple or quadruple negatives.

people do teach relevant things. reading what the words are is relevant to understanding a written sentence. understanding !!X is relevant to double negatives. but using it is kinda up to you. people also do teach you to avoid using double negatives, and warn they are confusing.

people teach syllogisms. but that's a bad way to think about most arguments. it can make things worse!

people get examples. lots. kids hear many sentences. and they get information about what the sentence was about. like the parent says something including "pancakes" and then 20 minutes later breakfast is served. repeatedly.

i think a lot of how people understand sentences is actually like that. they map the sentence to: "blah blah something about pancakes" and think " cool i understand him". and when you're 3, hey, that's pretty good! success! but when you're 13 or 33, that's bad. but the 33 yo actually interprets tons of stuff that way and muddles through life.

and one of the things that makes it hard is the people talking are as dumb as you. so they say some actual specific words. but they don't know what words mean all that well either. so all they even MEANT is "blah blah something about pancakes". or maybe they had something more complicated in their head, but it didn't match the words they used anyway. that's not universal (even with dumb people) but it's common. a lot of times, "blah blah something about pancakes" is about as good as you can do because the speaker actually didn't use the right words to communicate more. if you try to listen to the details of what he said, and interpret them, you'll just misunderstand him!

people are really bad at explaining or teaching anything they find "self-evident" or super super obvious. that's part of the issue.

Rand, Popper and DD are great at this stuff, but their books don't teach it. they explain more advanced stuff.


anyway, i was thinking there are these subjects people know how to teach (often quite badly). and subjects they basically don't teach at all. and i don't think the other ones are impossible to teach btw, people just don't know how much or don't do it or really really suck at it.

some of the stuff they don't teach is basically the kinda stuff that IQ is about. people don't know how to teach IQ. (btw there's also other stuff they don't like teaching like swear words or sex stuff).

in more Popperian terms people need to do steps like:

1) understand, conceptually, what the problem they are trying to solve is.

and ppl get that wrong all the time. and no one really teaches it as a general skill. or knows how to. it is taught in specific ways, like they'll teach you about a particular category of chemistry problem and how to think about it and what to do with it.

2) brainstorm solutions

how do you brainstorm? there's infinite things you could come up with. which ones should you? how do you know? this isn't really taught. isn't not very important though, just don't get stuck here. the initial brainstorming can be shitty ideas, and that gets fixed in step 4.

3) criticize the brainstormed stuff

here is where you have to actually figure out what kinda stuff to target. you both think of attributes the ideas/examples/solutions/whatever should not have for some reason ("let's not use an animal example b/c ppl are confused about how animals differ from humans. an inanimate object will be clearer") and also attributes it should have ("i want a solution that leaves me with at least as much money as i had before, so i'm not gonna do X or Y").

4) judgement, like which criticisms apply to which ideas.

this is partly hard just in a basic way. if trait Z is bad, ok, well, which ideas have trait Z and which don't? how do you figure that out if you don't already know it?

and it's partly hard in a more complicated way b/c you don't wanna just throw out a bunch of ideas b/c of a dumb criticism. you also need to be judging the criticisms too and making counter-criticisms. that's so complicated it actually kinda ruins my attempt to make this a linear step-by-step process. i just threw it in here.

5) brainstorm variants

so the initial brainstorming can just be rather random crap. that's fine. i don't think any healthy adult actually has much trouble with that, even if they can't really say how they do it.

but this part is harder. where you're coming up with ideas that meet criteria you had from (3). it's like, how do i change the solution to leave me with more money? what are some inanimate objects? (ok that one is easy, but some are harder).

one reason people get stuck on (2) is they already know some criteria of criticism. so they skip (2) and do (3) first, and then move on. that's fine. no problem at all. it just doesn't count as being stuck on (2) if they are really stuck on (5).

6) judge when to stop

when is the idea good enough? how much more should you think of criticism and brainstorm better solutions? this isn't taught. people guess wildly and sometimes make corrections to their policies (like they realize to spend more time than normal for important stuff).


btw, i have not taught it to you here. this is summarizing and describing it. it doesn't actually teach you how to actually do it. it gives you some hints from which you might figure it out yourself. and in some sense that's all we ever give students. but there are lots of topics where the hints are way better and include actual explanations of how to do something. my 6 point list is not what i'd consider teaching it in the usual sense. it's talking about it and it may be helpful, but it really leaves a lot up to you to figure out how actually do the things i mention.

and there's other stuff besides the 6 things on my list and the points earlier like about non sequiturs and figuring out what's an argument for what and how that works and what would and wouldn't be a counter argument. but like, these are basic things people are bad at, and it's the kinda thing that matters to IQ, and people don't teach it.

it's also hard to get ppl to try to learn it. people on FI want to do things like learn Objectivism when they can barely read sentences, you know? they don't sit there and go over the basics.

part of the issue is people can read sentences. 90% of the time! but it's like, even a little idea involves using some basic skills a bunch of times. 50 tiny little basic things might go into 1 idea. so even if you're 99% reliable and doing the basic stuff right, tons of your ideas will be wrong! you need to be able to do the basic stuff with a VERY good success rate or you get totally overwhelmed with errors when you try to build complex mental structures out of millions of basic components. but people generally don't like practicing stuff they get right over 80% of the time. they don't like trying to go from 99% to 99.999%. and besides being HIGHLY reliable at the basics, people also need to be FAST at them. if you're going to build complex mental structures out of a million little pieces, you better be able to do most of the little pieces in well under a second. but people also don't like practicing to get faster at stuff they are already fast at. they don't like trying to go from 2 seconds to .2 seconds to .02 seconds.


morality is another thing people are bad at teaching. there are people who are good at explaining it like Ayn Rand. but like, i watched Pinocchio (the old disney movie) and the cricket (his conscience for some reason) gives Pinocchio a lecture on morality. it's something about avoiding temptation and doing the right thing. it's completely incoherent and Pinocchio doesn't understand. part of the actual plot is this incoherent moralizing that is not understood. and then Pinocchio is immediately thrust out into the world to face temptations to do other stuff besides go to school. and the movie illustrates, in a magical, exaggerated way with 2 unrealistic examples, how nice sounding things can be dangerous and he should have resisted temptation and gone to school.

i think this fits children's actual experiences pretty well: incoherent verbal moral advice they don't understand at home, followed by being thrust into the world totally unprepared.

a lot of how the Bible teaches morality is with stories, too. and there are other old stories with moral content, like fairytales.


Elliot Temple on June 2, 2017

Comments (12)

plz reply if you think you were taught any of the unteachables, or know how they're taught. or have more to add to the list. or know something about what the difference is. or have experiences about trying to learn the basics to share. or have insight about how to deal with ppl not wanting to practice from good to great (or not knowing how), but when it's tiny components which get used thousands of times in a project then they need to be at a much higher success rate than the higher level ideas one is used to dealing with.

curi at 11:08 AM on June 2, 2017 | #8697
>by contrast, i cannot remember anyone ever teaching me:

[...]

>how to figure out if an argument is true or false

>what an argument IS (and how to decide if X is an argument for Y)

i did some logic stuff in a class.

stuff in the nature of

>Arguments involve reasoning from PREMISES to a CONCLUSION.
>So consider the following:

>1. Democrats are evil
>2. You shouldn't vote for evil people
>*THEREFORE*
>3. You shouldn't vote for Democrats.

>1 and 2 are premises and 3 is the conclusion. If we ASSUME 1 and 2 are true, then 3 follows.

but that's pretty different than IRL stuff where you are dealing with lots of messy disputed facts.

>a lot of times, "blah blah something about pancakes" is about as good as you can do because the speaker actually didn't use the right words to communicate more. if you try to listen to the details of what he said, and interpret them, you'll just misunderstand him!

yeah. you kinda have to think on the level of a typical person somewhat in order to figure out what they're saying. (kinda like with Poker ... if you attribute too much skill to a person, you'll mess up and lose. You gotta figure out their skill level to some decent approximation)

>one reason people get stuck on (2) is they already know some criteria of criticism. so they skip (2) and do (3) first, and then move on. that's fine. no problem at all. it just doesn't count as being stuck on (2) if they are really stuck on (5).

i'm kinda confused here. are you saying people who think they are stuck on 2 actually aren't, oftentimes?

BTW I've seen a little material in a couple different places on what might be IQ kinda stuff (I mean places besides FI). Like I did a computer science course once on Udacity and they added a whole chapter (after people were struggling tons) about How to Solve Problems. And it had advice like "break up a big problem into smaller problems and solve the small problems" which probably seems very obvious to many people but which I found helpful.

that seems like a basic kinda thinking skill that isn't really taught much.

Anonymous at 5:24 PM on June 2, 2017 | #8698
> Arguments involve reasoning from PREMISES to a CONCLUSION.

those are called "syllogisms", i mentioned them.

> And it had advice like "break up a big problem into smaller problems and solve the small problems" which probably seems very obvious to many people but which I found helpful.

that is higher level stuff. Popper and Rand have lots of advice like that. that particular one (breaking into pieces) is also really common.

the sort of thing that's harder to teach is: how do you decide what smaller pieces to break it up into?

you can give examples. you can give some tips for a category. like if you're butchering an animal (category), usually you break it up into pieces where the tendons are. (no idea if that's true, just made it up.) or if you're trying to plan a wedding (category) there are various typical pieces to break it up into (flowers, dresses, cake, food, music, dance floor, alter, inviting guests, etc, etc). but how do you do it *in general*? i can do it but if someone had no idea how to do it and was lost, it's hard to figure out what to tell them besides walking them through my thought process for some examples.

Anonymous at 5:32 PM on June 2, 2017 | #8699
> how to read sentences that don't not have double, triple or quadruple negatives.

lol

i didn't catch it right away. i read it 2 times before i was like "OH DOUBLE NEGATIVE!"

Anonymous at 3:45 AM on June 3, 2017 | #8701
> plz reply if you think you were taught any of the unteachables, or know how they're taught. or have more to add to the list. or know something about what the difference is. or have experiences about trying to learn the basics to share. or have insight about how to deal with ppl not wanting to practice from good to great (or not knowing how), but when it's tiny components which get used thousands of times in a project then they need to be at a much higher success rate than the higher level ideas one is used to dealing with.

i'd rather not do that. i'd rather get help on my particular flaws.

elliot, do u know which of the things in the OP are things i have trouble with (my guess is yes), and will you explain them to me in detail, like with real life examples from my FI posts? i mean enough so that i can recognize that they are problems i have. (then after than i can work towards solving them.)

if you don't think it's beneficial for you, i'm open to considering making it beneficial enough with cash.

though this makes me think of another issue that i haven't communicated yet. i wonder whether it's a good decision to spend money getting you to help me with my problems when instead i could be spending that same money on getting you to help my kids. one idea is that spending money on them should come first. but i also have the competing idea that if i spend money on myself first, and fix some of my problems, then i'd be in a better position to help my kids.

GISTE at 4:04 AM on June 3, 2017 | #8702
> how to figure out if an argument is true or false

I got taught some (wrong) stuff about this in school. I remember it cuz it contradicted faith which I was taught at home.

What I was taught (wrong!) was basically to presume a claim is false until it's proven true. A claim can be proven true if the person making the claim provides more evidence which supports it than the person denying the claim can counter with their own evidence.

PAS at 5:44 AM on June 3, 2017 | #8703

Taught?

Lately I have been noticing things that seem like a shift in either Elliot's thinking or way of talking about stuff.

This blog is one example - previously I thought Elliot was quite against the idea of teaching stuff. The impression I had was something like you can help people learn stuff they're trying to learn, but you shouldn't try to teach them stuff. But this post seems kinda positive on the teaching idea. Am I misreading?

Another example is IQ, which Elliot recently posted about on FI. To the best I recall, Elliot was against the idea that you could measure intelligence with a test and describe it with a number. But now he seems more positive on the general idea of IQ while still criticizing specific aspects of IQ tests.

Elliot, have you changed your mind about something significant recently related to these topics, or is it just my imagination?

PAS at 5:54 AM on June 3, 2017 | #8704
> What I was taught (wrong!) was basically to presume a claim is false until it's proven true.

That's a *partial* method, not a full method. I don't think it actually addresses the hard part of figuring out what's true or not. In other words, I don't think it counts.

> This blog is one example - previously I thought Elliot was quite against the idea of teaching stuff. The impression I had was something like you can help people learn stuff they're trying to learn, but you shouldn't try to teach them stuff. But this post seems kinda positive on the teaching idea. Am I misreading?

By "teaching" I had in mind something like creating essays and videos, which tell people stuff, just like my prior essays and videos.

I was thinking about other stuff rather than how to reword "teaching". I don't think it matters much.

> Another example is IQ, which Elliot recently posted about on FI. To the best I recall, Elliot was against the idea that you could measure intelligence with a test and describe it with a number. But now he seems more positive on the general idea of IQ while still criticizing specific aspects of IQ tests.

I clearly stated that IQ tests are terribly designed (many of the questions are badly written) and don't measure intelligence.

They manage to correlate with some things anyway, which means something. To a loose approximation they try to measure a certain skillset which matters.

I wrote about IQ recently because Jordan Peterson talked about it and I found some of his comments interesting enough to write some of my own comments. Specifically he said something along the lines of: never mind if it measures intelligence, the test results correlate to some things, and that means something. I thought that was worth some further thought. (I haven't checked how good the correlations are, btw, because I found the correlation claim totally plausible. Stuff like a cultural bias on the test won't stop it from correlating with some measures of success in your culture!)

*I don't think I changed my mind about anything significant.*

curi at 10:56 AM on June 3, 2017 | #8706
http://emilkirkegaard.dk/en/wp-content/uploads/Intelligence-and-socioeconomic-success-A-meta-analytic-review-of-longitudinal-research.pdf

> The relationship between intelligence and socioeconomic success has been the source of numerous controversies. The present paper conducted a meta-analysis of the longitudinal studies that have investigated intelligence as a predictor of success (as measured by education, occupation, and income). In order to better evaluate the predictive power of intelligence, the paper also includes meta- analyses of parental socioeconomic status (SES) and academic performance (school grades) as predictors of success. The results demonstrate that **intelligence is a powerful predictor of success but, on the whole, not an overwhelmingly better predictor than parental SES [socioeconomic status] or grades**. Moderator analyses showed that the relationship between intelligence and success is dependent on the age of the sample but there is little evidence of any historical trend in the relationship.

Emphasis added. I find this totally plausible and also they are basically doing math which is the kind of thing they might actually get right. I think everything they're dealing with, besides IQ, is reasonably easy to measure. (It's different than actual life success. It's just some loose proxies for conventional success.)

Reading on:

> One of the central and personally most relevant desirable outcomes is socioeconomic success (or career success), which is usually measured by the educational level, occupational prestige, and income of an individual in adulthood.

All of those are easy enough to measure. You just have to make a culturally normal list ranking the occupations, and have some procedure for dealing with multi-occupation people which isn't so bad it ruins your study (not hard). Sure it's a bit biased, but we're just trying to loosely correlate IQ to biased, culturally-normal concepts of success. So it's good enough.

> 5.2. Collection of data

You can read their criterion for what studies this meta-study included. And it's reasonable. They rule out lots of dumb stuff. It's not perfect but it looks decently alright to me. About as good as I would have hoped for.

> It is rather common for published studies not to report the information necessary for meta-analysis

For some studies, the study reporting was too shit to use (e.g. they don't actually clearly state the direct correlation between IQ and X). But the raw data was available. So he calculated the correlation information from the raw data and used it in the meta study rather than waste the data. OK, cool.

----

Does anyone think IQ doesn't correlate with SES and some other things, given various background assumptions like you stay in the same country and culture, there isn't some big upheaval, etc,? Is this paper wrong and bad? Anyone want to dispute this?

curi at 11:17 AM on June 3, 2017 | #8707
> i'd rather not do that.

you'd rather not do what?

> if you don't think it's beneficial for you, i'm open to considering making it beneficial enough with cash.

no i don't want to write tons of stuff, for you personally, for free.

i posted a rate to hire me. you've seen it.

> though this makes me think of another issue that i haven't communicated yet. i wonder whether it's a good decision to spend money getting you to help me with my problems when instead i could be spending that same money on getting you to help my kids. one idea is that spending money on them should come first. but i also have the competing idea that if i spend money on myself first, and fix some of my problems, then i'd be in a better position to help my kids.

this is too vague. if you have actual projects in mind, then you could compare better. is there something one of your kids wants help with?

curi at 11:38 AM on June 3, 2017 | #8708

Taught?

PAS #8704 writes:

> previously I thought Elliot was quite against the idea of teaching stuff. The impression I had was something like you can help people learn stuff they're trying to learn, but you shouldn't try to teach them stuff. But this post seems kinda positive on the teaching idea. Am I misreading?

For years, Elliot has wanted to find ways of helping people learn important ideas. He's tried or considered many approaches. I see no indication that this post represents a shift in his ideas on how knowledge is created.

> Another example is IQ, which Elliot recently posted about on FI. To the best I recall, Elliot was against the idea that you could measure intelligence with a test and describe it with a number. But now he seems more positive on the general idea of IQ while still criticizing specific aspects of IQ tests.

I think Elliot's position here has always been something like (my loose paraphrase): IQ measures someone's ability to perform well at a modern knowledge-worker desk job. Being good at that means you'll also be good at some other stuff.

This is consistent with his criticisms of IQ tests, including the problem of weighting the results to form a single number and cultural bias in the test questions.

There's also significant difference between Elliot's view of IQ and most everyone else's. The consensus view is that a large part of the differences in people's can be explained at the level of their genes (like differences in height or eye color, but in a more complicated, less-well-understood way), but that's not Elliot's position. AFAIK, he believes that a high IQ is a result of ideas that can be learned (in principle) like anything else.

Alisa at 1:19 PM on June 3, 2017 | #8710
*differences in people's IQ can be explained

Alisa at 1:24 PM on June 3, 2017 | #8711

What do you think?

(This is a free speech zone!)