The Library

I had walked past those shelves hundreds of times before I actually saw them.

I was in my local public library, looking for a particular section. I don’t remember whether I was following an alphabetical listing or the Dewey Decimal system. That part apparently wasn’t important enough for my brain to keep.

But something else was.

Every shelf I passed had empty space at the far right.

Not one shelf. Not one section that happened to be waiting for more books. Every stack, and every shelf within that stack.

There was no exception to the rule.

One hundred percent of the shelves were less than one hundred percent full.

I stopped and stared at them for a minute.

Why?

I had seen this hundreds of times. Presumably, I had understood on some subconscious level that libraries do not pack every shelf from end to end. But until that moment, I had never actually thought about it.

And then I realized what it reminded me of.

Hard-drive partitioning.

I warned you this was From the Journal.

Years ago, for reasons that made perfect sense to me at the time, I decided I wanted to build my own computer from individual component parts. I learned about motherboards and hard drives and eventually how to partition a hard drive from scratch.

Somewhere in those lessons was the idea that you deliberately did not use every bit of available capacity. I no longer remember the exact technical rule, and it doesn’t really matter. What stayed with me was the idea of intentionally preserving unused space even though you knew the space was available.

That was what those library shelves reminded me of.

The analogy wasn’t perfect. Hard drives and books obviously aren’t organized in the same way. But that is the fun of mental tangents. You don’t actually have to tie everything back together in a neat little bow.

One thought just has to get you to the next one.

So I started thinking about why a library would deliberately leave all that space.

And suddenly it was obvious.

The books have to stay in order.

Imagine a shelf of mystery novels arranged alphabetically by author. The shelf is completely full, and the authors on it run from A through D.

Then tomorrow the library acquires an amazing new book by Mary Higgins Clark.

It belongs with the C authors.

That part is easy.

Except the shelf is full.

You can’t put Clark on some empty shelf farther down the aisle simply because there happens to be room there. The whole point of the system is that the book has a place where it belongs.

So you make room for it.

But now another book has to move to the next shelf. And if that shelf is completely full, one of its books has to move to the shelf after that. If every shelf is completely full, this could theoretically continue all the way to Z.

One new book could disrupt an entire library.

Suddenly all that empty space didn’t look very inefficient.

And that was the part I filed away.

Apparent inefficiency.

Leaving empty space—deliberately not using all of the available capacity—could actually be necessary for the efficiency of the larger system.

I loved the contradiction.

And then, years later, I needed it.

I was helping design a classification system for a long list of things sold to clients.

The first part was reasonably straightforward. Take the jumble of things and start grouping them into categories: things that are more like each other than they are like the other things in the list.

We do this everywhere. Supermarkets do it. Libraries do it. We do it in our homes.

But then the problem became more interesting.

How do we name those things in a standardized way so that the naming convention is intuitive and matches the classification system we’ve created?

And then came the question I cared about most:

What happens when the new thing arrives?

Not one of the things already on the list.

The thing that gets created tomorrow. Or next month. Or three years from now.

Where does that go?

A classification system can look beautifully organized when it contains only the things that existed on the day it was designed. But if adding one new item means changing the identifying information of everything that comes after it, we have recreated my completely full library shelf.

The new thing has nowhere to go.

And suddenly, standing nowhere near a library, I remembered those empty shelves.

I used the library to explain the problem to my colleagues.

Imagine the shelf is completely full, I told them. Now we need to insert a new book in the C section. What happens to everything after C?

And finally, they understood what I was trying to describe.

That may be my favorite thing about mental tangents.

An observation can sit quietly in your memory for years without having any obvious purpose. Then one day, something entirely different calls for it, and there it is.

I had thought the empty shelf was unused space.

It turned out the empty space was part of the design.

Sometimes the empty space is where the future goes.