Hacker News new | past | comments | ask | show | jobs | submit login
The Collector’s Fallacy (zettelkasten.de)
134 points by ingve on Mar 17, 2018 | hide | past | web | favorite | 61 comments

It's absolutely true that to be most useful, knowledge needs to be absorbed and integrated directly into the mind. But there's a large constellation of useful information and data out there too, and you can't and shouldn't memorize all of it.

I collect a lot of links to articles that I've read using pinboard.in. However, I fairly frequently end up searching through those bookmarks to find the article that I remembered reading. I suppose once or a few times per month. It's especially useful for bookmarking specific scientific studies, or papers presenting engineering techniques, etc., where it's useful to remember "I recall reading about that" and be able to get back to the original.

> there's a large constellation of useful information and data out there too, and you can't and shouldn't memorize all of it.

Not only that, but also

> Just knowing about a thing is less than superficial since knowing about is merely to be certain of its existence, nothing more.

Sometimes just having a sketchy map of the territory is infinitely more useful to learning than knowing nothing at all. Of course we should strive for more than that, but when push comes to shove, a person with a rough map will waste less time in learning what they must, than someone who has no map at all.

I definitely have hoarding tendencies. I have 100s of vintage UNIX computers, 1000s of books on CS and EE topics. I mostly find the idea of them inspirational, even if I rarely get to explore them in the depth I wish to. But the hoarding tendencies don't seem to show up in my work product, i.e. I am very deliberate and keep a small stack of active projects/tickets/emails. I guess I am good at compartmentalizing?

OTOH people often seek me out for pointers to code/libraries/tickets/books etc at work and in my circle of professional friends. I'm kind of like a librarian.

This strikes a chord with me. I do the same with the hacker news posts. Every time, I open quite a few tabs with every intention of reading them. However, due to time constraints, I don't. Eventually I just "onetab"(firefox addon) them. This fear of missing out on information that might be valuable is troubling me a bit.

Much the same here. I have six windows open, while only a few of the tabs are Hacker News, one of them has more than 100 tabs in it and the others have about 40 tabs. That's just on this computer. My laptop has another dozen tabs open. Eventually, Firefox will crash and I will lose all my open tabs when "Restore Last Session" fails on me. Hundreds of PDF documents, about 25 on my reader, dozens of songs in sheet music some as much as half-learned.

I have a Commodore 64C, monitor, floppy drive, and a pile of disks, I've used that once. I've two ZX Spectrum machines (one in a dk'tronics keyboard, nonfunctional) and a Spectrum 128, with 50 or so tapes and one of those Microdrives plus some cartridges.

My father is the same with DVDs, he has a couple of thousand. The dad of one of my friends has double that. His lounge has a wooden shelf hacked together, that thing stands floor-to-ceiling along one side of the room, wraps around a corner, and is packed solid with DVDs. Babylon 5, Red Dwarf, Monkey, the Goodies, Paint Your Wagon ("Gonna paint that wagon, gonna paint it good, we ain't braggin', we're gonna coat that wood"), the list goes on.

To me, my collecting and never doing anything with the items seems almost like a sickness. I still have all my University texts - I justify it to myself by quoting a friend of mine who said "I don't believe in selling textbooks, they might come in handy some day." They won't, most of them are more than a decade old, but I still can't bring myself to get rid of these books that cost me hundreds. It was hard enough dumping those ISA cards that were no use to anybody - and I still have a stack of 500 megabyte IDE drives somewhere.

Kudos for 8-bit machines. And Microdrive!

I understand your tabs behaviour - I am the same, but once when I come to 90-ish, I start culling session to cut it to more reasonable number.

I just wish I had space to set them up permanently. I've had a brief play with the Microdrive, it's a lot faster than the old tape deck. I don't know much about the longevity of the media, though.

Now I want to get the Spectrum out for a play, but I expect I'll lie down until the urge passes!

Thinking I might do a re-cap on each of the machines, two of the capacitors on the Commodore are a little bulgy. It is a 30something year old machine, though!

How do you get any work done. My pretty fast Macbook, 3y old, becomes really slow after only 15 tabs open in 1 window.

Do you use chrome or something? Firefox is excellent with hundreds and hundreds of tabs (on my 2011 linux machine)

Sandy Bridge i7 with 16 gigs and (as Kliment mentioned) Firefox, although I do occasionally have slight slowdowns. Those could be related to my six BOINC threads as the slowdowns are in YouTube.

I send them to onetab too. After a while I do cleanup and most of the tabs that look useless or too late to read, I delete them.

One of the rationalizations I use for collecting books is that, by organizing them in a certain way, I can constantly reinforce my model of the structure of certain topics. On top of that, the physical presence of certain books is intermingled with what they contain in such a way that I think it_might help retention.

That said, I've given away a ton of books lately, and plan on keeping the books I own to a minimum until I have the means to construct a really strong library, which I do think is a positive endeavor.

Also, tangentially related, does anyone have any suggestions on a good org to donate books to? I've seen one too many awesome sets of books mangled in Goodwill bins, and am looking for a new option.

I believe public libraries are usually happy to accept book donations.

True, but rarely do they actually incorporate the donations into their collections. More typically, they combine the donations with books they've culled from their collections and sell them in "Friends of the Library" sales to raise money. Which is all good and fine, but not often what people think will happen.

Which are often immediately sold to the general public in a used book sale.

I have no problem with that. Selling used books as an individual is hard, since people will not pay a lot of money for them (since brand new books are cheap), so at least for me I find its not worth the time to sell them. At least then the public library benefits from the donation.

What's wrong with that? The library gets money and the book goes to someone that wants it. Win-win.

On the other hand, if you save something to read later, this is a way to avoid reading it now. Might this be better than getting too distracted? Maybe you won't need to read it at all?

Also, it's a way of collecting things you could read; a wishlist, if you like. Having a wishlist is fine for things you don't need to do.

A couple of things I do:

- I don't buy books on Kindle. I just send a sample to my tablet. If I finish the sample, maybe I'll buy it. Otherwise, I avoided paying for it.

- Similarly, for any purchase on Amazon that's not urgent, I just add it to the shopping cart. If I decide I don't want it later, I avoided buying it.

Ironically, Umberto Eco (quoted in the article) had a library of 30 000 books at his primary residence and 20 000 books at his vacation home. He probably hadn't read most of them :)

Careful with that. Alan Kay claimed that at 76 he had read between 16,000 and 20,000 books[1] and that he now reads not more than about two hours a day[2]. When I first heard that numbers I was skeptical, but given that he started reading at the age of three[3], it checks out at about 500 words per minute, which is quite realistic.

Now consider that Umberto Eco

1. was, in contrast to Alan Kay, a full-time writer, so reading books was basically his job

2. lived till the age of 84

3. probably required a solid library of reference works (which are not supposed to be read from front to back) for his kind of work

4. was a polyglot and his library possibly contained the same books in different languages

it is not too unreasonable that he had read most of the (non-reference) books in his library.

[1] https://news.ycombinator.com/item?id=11945100

[2] https://news.ycombinator.com/item?id=14436018

[3] https://news.ycombinator.com/item?id=11806477

And Nassim Taleb has a take on this fact that makes it a feature, not a bug.


But that "feature" was from an era before the internet and Amazon and Google Books, back when not-buying a book meant the information would likely be lost to you forever.

He famously hadn't, as discussed in "How to Travel with a Salmon"

Can you elaborate what is discussed in "How to Travel with a Salmon"? The text is not available online, unlike the text that claims he did [0].

[0] https://www.telegraph.co.uk/culture/books/3642577/Heavyweigh...

It's in the chapter "How to Justify a Private Library". I found it excerpted here: https://books.google.com/books?id=xzjkYlfHXL4C&lpg=PP1&pg=PA...

Ironic that the way you showed the justification is proof that the justification is no longer valid. You didn't have to buy the book to use it as a reference when needed.

Books, as traditionally written, are too big for the amount of information which they typically contain. We need knowledge representation which is capable of coherently and consistently storing information in a more direct, relational format.

I've seen this sentiment many times before, and I usually respond with a variation of the following point:

If I took a book and stripped it down to the bare information, as you'd like, and then read this new information-pure skeleton, I wouldn't remember it as I would if I just read the book. I might only remember a few facts about Thomas Jefferson after reading a bio about him, but I wouldn't remember any facts if I just read the facts.

My hypothesis as to why this is true (disclaimer: this comes almost entirely from introspection): My mind doesn't store memories like a hard drive, you have to tell it a story (so to speak) to get it to remember what you said. My mind doesn't integrate purely relational data, it integrates narrative data. Yes, an important part of a narrative is context and relational information, but another key part of it is the cadence, the spaces interleaved between the stuff we remember. Unless your point was that we actually need a new architecture for our brains, I strongly disagree, albeit on a purely anecdotal basis.

Human memory is also associative. To retain a fact we usually need to know how it relates to our existing knowledge, a "hook to hang it on", so to speak.

You don't have to do away with the storytelling structure; just make it significantly shorter and designed for information retention. Quite a few books contain a good deal of filler to reach the 250+ page threshold for marketing purposes.

A longer book is correlated, in the audience mind, with more content, more research done, more of an investment to read it hence more of a payoff ("it's a book you have to study"), and also more time spent reading it so potentially more pleasure, even for non-fiction.

Unfortunately, when you read a lot, these propositions start to become cons, and you'd often rather have just the meat--granted, not in a plastic plate, to run the culinary metaphor, but not with the three-hour ceremonial of the (would-be) fancy restaurant either.

I agree to an extent, but I've come to believe that the 'padding', often just examples that are variations of each other, is much more useful than it appears at first.

When I think back to books that I've read that felt like they had too much filler, I have to admit that many of the memories are the concepts via the anecdotes.

So perhaps the author shot a lot of anecdotes at you and a few stick, but another author could have done the same job with half as many pages. Again, I'm not against storytelling or anecdotes.

Yeah, I don't disagree with that. I've all but stopped reading the books that are mostly anecdotes with about an A4's worth of actual information.

I think we’re talking about two different ideas. It seems like you’re talking about a publishing problem, not a problem with books qua books. That said, I haven’t ran into this other problem either. If I thought an author was padding a book just to make a page threshold, I’d never read that author again.

Ah, maybe we're talking of different things indeed. As for the padding, in my experience it happens very often, not with random sentences of course, but with unnecessary chapters, paraphrasing, too many anecdotes etc. If you think about it, it's not that surprising: why every idea worth a book should need 250+ pages to discuss? What about the many blog posts or TED talks artificially turned into a book, for money and prestige?

I suppose it's also why teaching can be so effective for learning things. Instead of things staying murky in your head, you're forced to create a narrative that not only you understand (or think you do), but another person can too.

Quite often I just get partly there by explaining something to myself as if I'm teaching myself, but it seems to have some benefit (but also with the introspection disclaimer there).

Then you will love blinkist.com, they summarize nonfiction book so that you can read or listen to them in 15 minutes.

But you might find that as other commentators point out, without the original packaging, although the summarized book contains all of the insights of the original, it is hard to retain the information.

Speed readers usually say the same thing, you might be able to read and parse the text faster but your brain still needs the same time to process and assimilate the content, so you'll instead only end up with a buffer overflow of parsed but unprocessed text.

There’s no digital interface yet that beats flicking through a book and annotating the margins. Just as there’s no videoconferencing solution that beats a bunch of guys in a room around a whiteboard.

Meh. This is certainly something that can happen, but on the other hand a well-curated and indexed set of bookmarks can serve a purpose that even Google with all the personalization it can muster can't touch. As in all things, you need to be honest with yourself about how your habits serve your goals.

Absolutely - just recently on two separate occasions I spent half an hour to an hour searching Google for particular website that I was a fantastic resource on a topic.

Yesterday that was <https://www.homenethowto.com/>, where I eventually gave up on Google & DuckDuckGo and instead went to the home networking subreddit that I thought might just have it because it really is a great website. They did (and its probably where I found it originally).

How shallow are the books that the author of this piece reads, that he can keep their entire contents in his head? I am pretty sure I can't keep the entire contents of my personal library in my head. And I don't even have that many books. (Around 20, between CS and math books.)

I don’t keep them in my head at all :) In fact, my memory is quite bad except for procedual stuff like programming techniques. That’s why I process and store notes so extensively.

The fiction work I read that I cannot remember were particularly shallow. At one point I threw out everything where I couldn't even remember one or two plot elements.

I'm not sure about a lot of these suggestions. They always sound like they're aimed at a specific group of people who only collect things, and never use them. What about people who collect things, and then use them just fine?

And me reading an article is not too useful if I can't find it later.

While extremely guilty of this, I see it as a cache of good knowledge in different areas - I can generally remember what I’ve saved - their general titles and genres. If I need to dive deeper into something, I already have good starting points saved.

This looks interesting. I should bookmark it for later.

At one point I was buying/trying a lot of new stuff, shiny new hardware, gadgets, software, technologies, frameworks, etc. But now I don't care about learning/collecting new tools anymore unless I have a really good reason to do it.

For me, creativity seems to spur from mixing and matching a minimal amount of basic elements. Creativity is very close to mastering a computer programming language. In general, you only need to command a paintbrush well to compose great paintings. You only need to know 30 keywords to make amazing programs. It's better to make use of general purpose tools really well to my specific needs than to employ a whole bunch of specialized tools. Having specialized tools is the answer only when I need to copy and mass produce something.

One of the trends I have seen lately is some people have the tendency to branch out to make new gadgets and teach very specialized tools nowadays without merit. For example, the amount of javascript frameworks is absolutely mind-blowing, there seems to be a new javascript way to achieve the same damn result every minute. Someone recently bought me a sous-vide cooking device which both before and after having it, I still have a really hard time justifying having it in my home: What couldn't I do without it and with just a thermometer? I cook sous-vide exactly 0 times before I got it, and do it 3 times in 6 months since I had it. When I do it, I could as well enjoy and watch the thermometer instead of having download an app to my phone that I use once every 2 months to watch the temperature for me.

It's not just in tech -- I understand they need to sell shit. Even in education, there are courses of "data science in X" or "X statistics" offered at every university for every field. One of my friends (who is doing a PhD) recently asked me to help with one of her assignments for her course. Basically, it was just two hours of boring shit teaching her to use graphs with various examples in her field X, I just jumped in and helped with what I knew, which has nothing to do with her field X. Why can't it be just a general-purpose data science or statistics course?

With this rate, sometime in the future, we'll have sous-vide device just for 20-year-old blonde people who are married instead of just a thermometer that works for everyone. Rand's razor and all that.

With things like Hackernews and Reddit, there's a lot of stuff we tend to skim. I tried more often than not to actually read a paper if the full text is available, but that can be difficult to do too.

I like writing, and sourcing things I've read into new work seems the best way to learn, retain and think critically. Essays from high school and Universities need not be an end. Having a professional blog where you organize your ideas and opinions means you have to put some real research into them.


Party conversation opinions are backed up with real stuff you look up, and you can feel more confident in expressing a critical opinion instead of parroting, inaccurately, something you may have read

The author says we shouldn't keep books, we should assimilate books into our model of the world, and then dispense with the book. To an approximation, I agree. I also agree with the idea that sometimes collecting books and bookmarks can lead to a false sense of satiety with our intellectual progress.

I disagree with his point that notes actually do represent assimilated information in a way that the books themselves don't. Notes might guarantee a certain level of assimilation, but they don't qualify as "knowledge" in the way the author maintains. There's other good reasons to keep notes, but not because they're more a part of your knowledge base than the book itself.

I do struggle with the information/knowledge distinction as I never researched that field in detail. Would love to hear about some helpful alternative!

I can imagine a primary source that’s important in your life becomes part of your knowledge, too, because you ... well, got to know it so well. But that really isn’t the point, since mere collecting is the opposite of familiarizing yourself with the material.

My point isn't really about a distinction between information and knowledge. While high quality notes might witness the existence, or creation, of knowledge, that knowledge is not contingent on the notes' continued existence. Put another way: If the note-taking process results in the creation of knowledge inside my head, why should I keep the physical notes any more than I should keep the book?

Edit: I should say I thought your article was interesting and useful in general, and I don't think my disagreement takes away from the general message that owning things is pointless, what matters is engaging with the world.

Sure sounds nice to have a perfect long term memory and have no need to use a quality source to refresh on any topic you ever once actually understood.

I recurrently think about these topics and my opinions have adapted from:

No notes -> wiki -> searchable flat wiki(s) (TXT note on a text editor)

Collect most -> collect relevant -> collect relevant and recurrently digest it (manually throw away what became irrelevant)

Some extra thoughts: https://github.com/galfarragem/superfolder

I like that project status approach to filing!

Old problem. The book "Too Much to Know" by Ann Blair is about how intellectuals dealt with information overload in early modern Europe: https://yalebooks.yale.edu/book/9780300165395/too-much-know

Makes me think of a t shirt I received as a gift many years ago. It said:

“So many books, so little time”

That's a Frank Zappa quote.

It's also not far off Hippocrates:


To combat this, I’ve started going through all my old bookmarks and compiling my own repository of knowledge on a self-hosted Bookstack. I highly recommend doing this. I feel like I assimilate a lot more knowledge than before when I would just read something once on HN and then bookmark it for posterity.


Guidelines | FAQ | Support | API | Security | Lists | Bookmarklet | Legal | Apply to YC | Contact