Also, many people’s optimal learning style is quite concrete; .
I hasten to point out (well, actually I didn’t hasten, I waited a day or two, but...) that while this is true for many people, it isn’t true for all, and, in particular, it isn’t true for me. (See here.)
I believe your first experience with a subject should be example-based, to fix ideas. After all, when you were a kid you played around with numbers long before you defined the integers
I don’t think the way I learned mathematics as a young child (or indeed in school at any time, up to and including graduate school) was anywhere near optimal for the way my mind works.
The best way for me would have been to work through Bourbaki, chapter by chapter, book by book, in order. I’m dead serious. (If I were making an edition for my young self I would include plenty of colorful but abstract pictures/diagrams.)
I assumed there were some folks like you but I’d never met one
It’s not as stark as that. For example, Alicorn, whom I believe you’ve met, shares with me a psychological need for concepts to be presented in logical order.
In my case, if you’re curious, I think the reason I’m the way I am comes down to efficient memory. To remember something reliably I have to be able to mentally connect it to something I already know, and ultimately to something inherently simple. The reason I can’t stand ad-hoc presentations of mathematics is that remembering their contents (let alone being able to apply those contents to solve problems) is extremely cognitively burdensome. It requires me to create a new mental directory when I would prefer to file new material as a subdirectory under an existing directory. (I don’t mind having lots of nested layers, but strongly prefer to minimize the number of directories at any given level; I like to expand my tree vertically rather than horizontally.)
This explains why it took me forever to learn the meaning of “k-algebra”. The reason was that (for a long time) every time I encountered the term, the definition was always being presented in passing, on the way to explaining something else (usually some problem in algebraic geometry, no doubt), instead of being included among The Pantheon Of Algebraic Structures: Groups, Rings, Fields etc. -- so my brain didn’t know where to store it.
I find it takes much more effort to learn things when different sources don’t coordinate well on definitions, notation, and the material’s hierarchical structure. For example, if everyone agreed on how to present the Nine Great Laws of Information Theory, that would make them much easier for me to remember them. It’s as if, instead of learning the overlap between different presentations, my brain shuts down and doesn’t trust any of them. But it’s hard to settle on such cognitive coordination equilibria.
Well, I can see the need to have the concepts fit together. What I need on a first pass through a subject is something that can attach to the pre-abstract part of my brain. A picture, even a “real-world example.” Something to keep in mind while I later fill in the structure.
The way I see it (which I realize is more of a metaphor than an explanation) human brains evolved to help us operate in large social groups of other primates. We’re very good at understanding stories, socialization, human faces, sex and politics. I think for a lot of people (myself included) the farther we get from that core, the more help we need understanding concepts. I need to make concessions to human frailty by adding pictures and applications, if I want to learn as well as possible. (This is something that people in abstract fields rarely admit but I think LessWrong is a good place to be frank.)
In my case, if you’re curious, I think the reason I’m the way I am comes down to efficient memory. To remember something reliably I have to be able to mentally connect it to something I already know, and ultimately to something inherently simple.
Interesting. That sounds like my habit of making sure everything I learn plugs into my model for everything else, and how I’m bothered if it doesn’t (literature and history class, I’m looking in your general direction here). Likewise, how I don’t regard myself as understanding a subject until my model is working and plugged in (level 2 in my article).
This is why I’ve usually found it easy to explain “difficult” topics to people, at least in person: per my comment here, I just find the inferentially-nearest thing we both understand, and build out stepwise from there. And, in turn, why I’m bothered by those who can’t likewise explain—after all, what insights are they missing by having such a comparmentalized (level 1) understanding of the topic?
I hasten to point out (well, actually I didn’t hasten, I waited a day or two, but...) that while this is true for many people, it isn’t true for all, and, in particular, it isn’t true for me. (See here.)
I don’t think the way I learned mathematics as a young child (or indeed in school at any time, up to and including graduate school) was anywhere near optimal for the way my mind works.
The best way for me would have been to work through Bourbaki, chapter by chapter, book by book, in order. I’m dead serious. (If I were making an edition for my young self I would include plenty of colorful but abstract pictures/diagrams.)
I assumed there were some folks like you but I’d never met one. Shame on me for making too many assumptions.
It’s not as stark as that. For example, Alicorn, whom I believe you’ve met, shares with me a psychological need for concepts to be presented in logical order.
In my case, if you’re curious, I think the reason I’m the way I am comes down to efficient memory. To remember something reliably I have to be able to mentally connect it to something I already know, and ultimately to something inherently simple. The reason I can’t stand ad-hoc presentations of mathematics is that remembering their contents (let alone being able to apply those contents to solve problems) is extremely cognitively burdensome. It requires me to create a new mental directory when I would prefer to file new material as a subdirectory under an existing directory. (I don’t mind having lots of nested layers, but strongly prefer to minimize the number of directories at any given level; I like to expand my tree vertically rather than horizontally.)
This explains why it took me forever to learn the meaning of “k-algebra”. The reason was that (for a long time) every time I encountered the term, the definition was always being presented in passing, on the way to explaining something else (usually some problem in algebraic geometry, no doubt), instead of being included among The Pantheon Of Algebraic Structures: Groups, Rings, Fields etc. -- so my brain didn’t know where to store it.
I find it takes much more effort to learn things when different sources don’t coordinate well on definitions, notation, and the material’s hierarchical structure. For example, if everyone agreed on how to present the Nine Great Laws of Information Theory, that would make them much easier for me to remember them. It’s as if, instead of learning the overlap between different presentations, my brain shuts down and doesn’t trust any of them. But it’s hard to settle on such cognitive coordination equilibria.
Well, I can see the need to have the concepts fit together. What I need on a first pass through a subject is something that can attach to the pre-abstract part of my brain. A picture, even a “real-world example.” Something to keep in mind while I later fill in the structure.
The way I see it (which I realize is more of a metaphor than an explanation) human brains evolved to help us operate in large social groups of other primates. We’re very good at understanding stories, socialization, human faces, sex and politics. I think for a lot of people (myself included) the farther we get from that core, the more help we need understanding concepts. I need to make concessions to human frailty by adding pictures and applications, if I want to learn as well as possible. (This is something that people in abstract fields rarely admit but I think LessWrong is a good place to be frank.)
Interesting. That sounds like my habit of making sure everything I learn plugs into my model for everything else, and how I’m bothered if it doesn’t (literature and history class, I’m looking in your general direction here). Likewise, how I don’t regard myself as understanding a subject until my model is working and plugged in (level 2 in my article).
This is why I’ve usually found it easy to explain “difficult” topics to people, at least in person: per my comment here, I just find the inferentially-nearest thing we both understand, and build out stepwise from there. And, in turn, why I’m bothered by those who can’t likewise explain—after all, what insights are they missing by having such a comparmentalized (level 1) understanding of the topic?