Cross posted from Overcoming Bias. Comments there.
When I was a teenager, I lived in a nice house with my mother, stepfather and three younger brothers. The contents of the house were what you would expect if you took a normal house, multiplied the number of things in it by ten, then shook it very hard. Almost – a greater proportion of the things were in boxes or containers of some kind than you would expect by chance, and also there were narrow trails cleared along the important thoroughfares. For instance there was a clear path to the first few chairs in the living room, from which the more athletic members of the household could jump to most of the other chairs.
This state of affairs interested me. From what little I had seen of other families’ houses, it was pretty unusual. Yet looking at the details of of the processes which produced it, I couldn’t see what was unusual. I don’t remember my exact thoughts, but I figured it had to be something that affected the relative inflow and outflow of stuff from the house. But it wasn’t that we had way more spending power than other families, or that we kept a lot of garbage. Most of the things in the house were useful, or would be if you had a non-negligible chance of finding them. It seemed like my family bought usual kinds of things for usual kinds of reasons. A set of lego for the children to play with, a blender because sometimes we wanted to blend things, a box of second hand books or two because they were only 50c.
The last one there looks a bit problematic, but is not that unusual. People often buy marginally valuable items because they are cheap. There were a few other things like that that looked a bit problematic – a tendency to keep drawings, an inclination to buy several shirts if you found one that was good. But nothing that should obviously cause this massive phase transition into chaos.
In the end I’m still not sure what the dominant problem was, or if there was one. But I can tell you about one kind of failure that I think contributed, which I also notice in other places.
Suppose you have a collection of things, for instance household items. You want to use one, for instance a pair of scissors. Depending on the organization of your collection of household items, it can be more or less tricky to find the scissors. At a certain level of trickiness, it is cheaper to just buy some new scissors than to find the old ones. So you buy the new scissors.
Once you have the new scissors, you add them to your collection of things. This is both the obvious thing to do with items you possess, and the obvious solution to scissors having apparently been too rare amongst your possessions.
Unfortunately adding more scissors also decreases the density of every other kind of thing in the collection. So next time you are looking for a pen it is just a little bit harder to find. If pens are near the threshold where it’s easier to get new pens than find your old pens, you buy some more pens. Which pushes a couple of other items past the threshold. On it goes, and slowly it again becomes hard to find scissors.
In short, a given amount of organization can only support being able to cheaply find so much stuff. You can respond to this constraint by only keeping that much stuff, for instance borrowing or buying then discarding items if they are beyond what your system can keep track of. Or you can respond by continually trying to push the ratios of different things to something impossible, which leads to a huge disorganized pile of stuff.
Another place I notice this is in writing. Suppose you write a blog post. Sadly it is a bit too long for the average reader to remember a key point in the second paragraph. You suspect they will forget it and just fill in what they would expect, consequently missing the whole point. To avoid this, you emphasize the point again in the second last paragraph. But now the post is even longer, and it is not clear whether they will also remember another key part. So you add some more about that point in the conclusion. But now it’s so long the whole argument is probably too hard to piece together, so you add a bit of an outline. Perhaps this eventually reaches an equilibrium in which all the points have been repeated and emphasized and exemplified so much that nobody can fail to understand. Often it would nonetheless have been better to just quit early on.
I think I had a better list of such examples, in a half written post which I put in my collection of blog drafts. Unfortunately my collection is so sprawling and poorly organized that it seemed easier to just write the post again than to find the old one. So here you have it. It’s tempting to add this post too to my blog draft collection and look for it again when I find some more things to add, but no good lies in this direction.