💾 Archived View for idiomdrottning.org › nfs captured on 2024-08-18 at 19:26:41. Gemini links have been rewritten to link to archived content

View Raw

More Information

⬅️ Previous capture (2024-05-26)

-=-=-=-=-=-=-

The Noguchi Filing System

The Noguchi Filing System is for keeping track of physical paper. I keep paper in envelopes on a shelf and new envelopes go on the right (or on the left if you live in Japan) and if I use an envelope, it goes on the right as if it were new. Linux nerds can think of this as sorting the envelopes by atime. Envelopes always must have a date along the spine and a word or sentence describing what’s in ‘em. Optionally they can have a color, using markers or stickers. Noguchi even cut them off one inch so that the a4 documents in there stick up a bit.

I use the Noguchi filing system and have for decades and tonight I needed to find a contract from 11 years ago and I managed to find it pretty quickly, next to break-up–letters and roller derby team rosters and tickets and old gadget manuals. 🤷🏻‍♀️

I’ve got to say, I did flinch when putting it at the top of the stack instead of putting it back with the other December 2011 era papers. But I trust the system, it’s been good so far, so…

♥︎

I personally do use one envelope per document rather than the buckets variant (in that variant, which Forster briefly promoted, each envelope represents an entire topic, like a folder). This is much better for me, since I don’t have to dig up something old just to file something new. Some documents are many pages, it’s not one envelope per page, but separate (although related) documents go in separate envelopes. For example, back in the day before publishers allowed digital submissions, one draft of a novel would be like hundreds of sheets, all crammed into one envelope. Separate drafts of the same novel went into separate envelopes even though they’re the same “topic”.

I’ve got to add, please don’t let me convince you to use the Noguchi filing system if you have doubts. I’ve loved it for years and it’s worked exactly as intended and it’s a perfect fit for me and for how my brain works. I’m more likely to remember temporal context than a name, I would’ve been like “is it C for contracts, H for housing, A for apartment, B for broker, S for the seller’s signature, F for the floor, X for the spot I live in” etc.

Similarly, I keep folders in the computer sorted by mtime, not by name.

I just have a huge and clear preference for that way of working but if you’re on the fence, I don’t wanna be responsible for if five years down the line you’re looking for a file and cursing because you feel I talked you into a filing system that wasn’t your bag.

One problem with the NFS is that it’s difficult to apply it to an existing collection of paper. For the first year getting organized, I used an A–Z filing system and when I wanted to switch to Noguchi’s method, I had one year’s worth of paper that I didn’t know how to sort. I was like: I wish I had used Noguchi’s system from the start and if I then would’ve wanted to move that into alpha, I could. But now it’s been fifteen years of NFS and I don’t have a lot of the junk from that one first year🤷🏻‍♀️

But in your case you might have several months of paper, are they all gonna get the same date?

Small might be why it works for me. Mine is only one foot wide (that currently is enough for a decade and a half). When it gets full, I grab the rarely-used end which by then is crystal clear to sort between junk and long-term-storage. A decision which isn’t as easy to make on the front end.

I also never put actionable, “TODO” type stuff in there. Only “this might come in handy” stuff.

Bonus hack!

I also have another hack that helps keep papers few.

Outside of my NFS I have two “junk” trays. Stuff only goes in them if I’m OK with throwing it directly into the trash. Paid bills are a good example.

Once the top tray is full (which usually takes about a year), I take the bottom tray and destroy everything in it, then I rotate the trays so the bottom is full and the top is empty. This way, I always have access to around a year’s worth of paper trash. In case a bill is disputed or something. I spend no time on “filing” things there, it’s just literally a flat trash basket.

It makes me less prone to over-filing and hoarding. I can just put something in “the flat trash bin” if I’m in doubt about needing it again. The tray becomes a mini NFS.

And, the main NFS itself also helps because if I’m in doubt but leaning towards keep, that’s also fine because it’s gonna “drift towards the left” the more useless it is. Everything on the left is either a perfect untouchable treasure, or completely obsoleted junk, and it’s gonna be so clear which is which so I can archive the treasures and get rid of the trash.

Better stack

Over the years, I’ve seen many people compare the Noguchi filing system to a stack, a pile on a messy desk or even papers in a huge barrel.

I first saw this comparison in the book A Perfect Mess by Abrahamson and Freedman, where they write:

If something seems vaguely familiar about the [NFS] arrangement, perhaps it’s this: Turn the row of envelopes so that the envelopes are stacked vertically instead of horizontally, place the stack on your desktop, and get rid of the envelopes. Now you’ve got an ordinary pile of papers of the sort that you’d find on any messy desk, where the most recent and most used items tend to end up at the top. So the next time someone tells you your pile-covered desk is messy, you can point out that it’s just hyper-organized.

Blinkyshark writes:

This made me chuckle, as it reminded me of a professor at my old university. He was a clever chap, naturally, but I never approved of his sense of organization. He had a huge stack of papers on his desk. It led me to joke that he didn’t file things alphabetically, but by centimeters from the bottom.

OK, so, in all seriousness, I believe there is some merit to this jokey comparison.

Stacks might work for some people, the same way the NFS has worked for me since 2007.

And if people are using stacks and piles, that’s a good thing. It’s not that I’d benefit from people rushing out and getting envelopes and folders if they don’t wanna. Save the Earth!

But so often when we hear these stories about a messy, absent-minded genius who has stacks and piles and barrels and clothespins and racks and cans, but still manages to be brilliant and get work done, it’s from the outside. Scoffingly and admiringly at the same time. It’s rare we get to hear how their “system” actually works.

So in that spirit, here are three ways to make your stack better (i.e. turn it into an NFS). Best (most impact) advice first, but you can mix and match.

Consistently put things in there

Part of what makes this work is that I know that it’s in there. When I had to find that contract, I was sighing and groaning because I didn’t wanna go delving in there, but I wasn’t scared. I knew it was in there (and I did manage to find it pretty quickly). It’s optimized for putting things in there without thinking (you're talking to Miss Decision Fatigue over here) so do it. A stack where you know there are treasures is a precious thing.

Also, I never put actionable stuff in there, only reference:

There are two kinds of notes

Stick to the atime sorting

Even with a vertical stack, you can do the NFS sifting thing of putting stuff on top, and then when you replace something from further down below, put it on the top instead of where it was. This has the con of breaking your temporal association, but these pros:

(Also you’re gonna feel better because you Have A Consistent System and not just a mess. This pro is kind of illusory but it’s a nice bonus.)

The envelopes aren’t useless

The envelopes, which have a date and a name and sometimes a color dot aren’t just for show (although they do look a li'l neater). It’s metadata and sometimes it’s easier to find something by seeing the date and/or name than to leaf through a huge pile of paper. I remembered the year of the contract which made it easier to find, and, I only had to read through one sentence per envelope, not stare at page after page of small print text trying to figure out what each one was. It’s like looking at filenames as opposed to lessing through a whole cat of files. This is also sometimes bad, though, because sometimes you do give it a dumb name and you don’t remember when it’s from.

Also, sometimes documents come in all kinds of shapes and colors, or are multipage folders or printouts or leaflets. That can make ordering difficult.

The envelopes are great to leaf through compared to a bunch of cut-inducing paper.

The horizontality isn’t useless either

This li’l section has defended stacks, but I’m actually happy that my NFS is in a horizontal shelf and not in a stack in a barrel or box or drawer. I can flip through it from the side as opposed to lifting every item one by one. A kinda minor benefit but nice none-the-less.

Finding things

I always say that the Noguchi Filing System has a tradeoff of making files slightly harder to retrieve but a lot easier to file, and for someone who hardly ever retrieves files but keep resisting filing because of the friction and boredom, it’s worth it, and it becomes a thousand times more worth it since it’s easier to cull old dead useless files which is very difficult and cumbersome to do in an A–Z system but trivial in a Noguchi system.

I might wanna revise that “tradeoff” philosophy because today I needed to find a file that in an A–Z I would have no idea where to even begin to look because I had no idea what I would’ve named it. It was just so odd and weird. I would’ve have to look through every folder. But with the Noguchi filing system I found it instantly because I knew when it had to be from.

I guess I’m more of a temporal thinker than a verbal one (as y’all can see how I struggle with the English language).

Also on Unix when I have no idea what a file is named or what its greppable contents are, I can be “but I worked on this other, known file at around the same time” and sort by time stamps (ls -rt |less) and search for the known file and look near there and that way find the weirdly-named file. Works great. On Unix I sort by mtime rather than by atime but in my physical NFS I like the atime version better.