4 ms·
Ask HN: Using plaintext argument maps to improve critical thinking?
Hi HN,
A year ago I came across the concept of argument maps and was intrigued by a meta-study showing that college students who used argument maps as part of philosophy courses improved scores on critical thinking tests by roughly one standard deviation over the course of a semester. Most representations of them I've found online have a big tree structure that takes up lots of space, but I realized I could use them in an indented bulleted list since that's also a tree structure. For example:
Plaintext argument maps are superior to just bullet points
- Support: prefixing each bullet's content with "support" or "oppose"
lets you easily see the logical relationships between concepts
- Support: since this just uses plaintext, it works anywhere you can
type
- Support: if you have to use "support" or "oppose" for each bulleted
item, then you naturally don't bother listing things that aren't
relevant, because they won't support or oppose
- Support: and if you're sure that something's relevant, but it both
supports and opposes, then clearly you need to either start with a
more fundamental assertion or break down the mixed support/oppose
item into separate concepts
- Oppose: writing "support" and "oppose" for everything requires more
writing
- Oppose: you can abbreviate them as "S" and "O" if you're not
illustrating the concept
- Support: This is kind of hard to read... single capital letters
at the start of lines stand out more from the prose
- Oppose: if I want to share my notes with people and I use this goofy
notation, I'll have to explain it to them and I'll feel like a huge
dork
- Support: this is accurate, it's pretty embarrassing
Anyway, I figured I'd share it if it might be useful to anyone else. I'm also curious if anyone else is doing a similar thing or has related ideas.
- jayajay 11y agoI like your format here. Humans tend to make snap judgements before they think. For example, when they read something, their mind jumps to a "agree" or "disagree" state much faster than they can actually sit and think about why they feel that way. If you're interested, check out Blink (Gladwell) and Thinking Fast and Slow (Kahneman). Agree/Disagree trees like this nicely map into this "fast" thinking style.
- djokkataja 11y agoThanks! I hadn't thought about how it maps to this "fast" thinking style, but that does help explain why I've found this method generally useful for breaking intuitions down and understanding them.
- joshschreuder 11y agoThere was a Show HN here recently called Arguman with a similar concept which I quite liked: http://en.arguman.org/the-universe-is-a-simulation http://en.arguman.org/the-universe-is-a-simulation https://news.ycombinator.com/item?id=10405288 https://news.ycombinator.com/item?id=10405288
- joshschreuder 11y agoNow I'm off mobile, I took another look at the site and it does in fact have a view that is basically what you presented above: http://en.arguman.org/the-universe-is-a-simulation?view=list http://en.arguman.org/the-universe-is-a-simulation?view=list Hope you like it!
- djokkataja 11y agoThanks! I remember seeing the Arguman Show HN; I'm hoping that these kinds of structures catch on more for general use.
- mbrock 11y agoI think understanding the basic idea of the Curry-Howard equivalence is great for critical thinking. A statement is a claim that premises and axioms lead to a conclusion, and a proof is an algorithm for how the axioms should be applied to get the conclusion from the premises. So you can write your claims as types and your arguments as functions.
- djokkataja 11y agoNow I just need to write a parser to turn my natural language plaintext argument maps into code :)